Linda Marie Crowe - eprints.qut.edu.au · Linda Marie Crowe Bachelor of Nursing Graduate Diploma of...
Transcript of Linda Marie Crowe - eprints.qut.edu.au · Linda Marie Crowe Bachelor of Nursing Graduate Diploma of...
Assessment of Preterm Infants’ Readiness to
Commence Breastfeeding
Linda Marie Crowe
Bachelor of Nursing
Graduate Diploma of Midwifery
Master of Midwifery
Submitted to
Queensland University of Technology
School of Nursing and Midwifery
Institute of Health and Biomedical Innovation
for
Doctor of Philosophy
2011
ii
iii
Keywords: preterm infant, feeding readiness, breastfeeding, mixed methods,
instrument development
iv
v
Abstract
Preterm infants commence breastfeeding when health-care professionals deem them
to be ready. However, the optimal timing for commencement of breastfeeding is
unclear. Currently, there is little guidance for neonatal care providers to decide when
to initiate breastfeeding among preterm infants.
A mixed-methods study was conducted to develop and test the Preterm Sucking
Readiness (PTSR) scale in four phases. The first phase involved a chart audit to
explore the use of age as a criterion by investigating when preterm infants meet
feeding milestones as well as other factors that may affect an infant’s readiness to
engage in nutritive sucking behaviour. The second phase utilised focus groups to
explore and define how neonatal care providers decide when to commence
breastfeeding. To gain consensus on the criteria mentioned by the focus groups, a
Delphi survey was conducted in phase 3, involving neonatal providers across
Australia and New Zealand. Phase 4 of the study involved an observational study
that was used to test the six-item PTSR.
The age at which specific feeding milestones were reached was consistent with what
has been previously described in the literature. The chart audit showed that the time
taken to the first feeding attempt in the preterm infant population was affected by
gestational age at birth, birth weight, and specific interventions. Staff also considered
age along with other criteria when deciding when to initiate feeding. Consensus on
nine criteria for inclusion into the six-item PTSR was achieved using the Delphi
technique. Three items of PTSR showed significant differences between the preterm
and fullterm infant groups. Only two items, feeding-readiness behaviour and low
pulse oximetry during handling, explained the variance in breastfeeding behaviour.
The inter-rater variability ranged between moderate and very good for the PTSR
items.
vi
The results of this study indicate the importance of assessing behavioural cues as an
indication of breastfeeding readiness in the preterm infant population, once an infant
is deemed physiologically stable. Age continues to be a factor in some clinicians’
decisions to commence breastfeeding. However, age alone cannot be used to decide
if an infant is ready to engage in breastfeeding. Further research is needed to confirm
these findings.
vii
Contents
Abstract ........................................................................................................................ v
Contents ..................................................................................................................... vii
List of Tables............................................................................................................... xi
List of Figures ........................................................................................................... xiii
List of Abbreviations.................................................................................................. xv
List of Appendices ................................................................................................... xvii
Publications ............................................................................................................... xix
Statement of Original Authorship ............................................................................. xxi
Acknowledgements ................................................................................................. xxiii
Chapter 1: Introduction ................................................................................................ 1
1.1 Background ........................................................................................................ 1
1.2 Significance ........................................................................................................ 3
1.3 Instruments to Assess Breast- or Bottle-Feeding Readiness .............................. 5
1.4 Aims ................................................................................................................... 6
1.5 Thesis Structure .................................................................................................. 7
Chapter 2: Literature Review ....................................................................................... 9
2.1 Introduction ........................................................................................................ 9
2.2 General Development of Behaviour in Preterm Infants ..................................... 9
2.2.1 Synactive Theory of Development.............................................................. 9
2.2.2 Neurobehavioural Development and Feeding Ability .............................. 12
2.3 Development of Feeding Behaviour in Preterm Infants .................................. 13
2.3.1 Development of Oral Structures and Function ......................................... 14
2.3.2 Development of Sucking Behaviour ......................................................... 14
2.3.3 Development of NNS ................................................................................ 15
2.3.4 Development of Nutritive Sucking ........................................................... 16
2.3.5 Development of Breastfeeding Behaviour ................................................ 20
2.4 Infant Factors Related to Readiness to Commence Breast- or Bottle Feeding 21
2.4.1 Behavioural State ...................................................................................... 21
2.4.2 Feeding-Readiness Behaviour Cues .......................................................... 22
2.4.3 Severity of Illness ...................................................................................... 22
viii
2.4.4 Physiological Stability ............................................................................... 23
2.5 Maternal Factors Related to Breast- and Bottle-Feeding Initiation ................. 24
2.5.1 Factors Affecting Mothers’ Decision to Initiate and Maintain
Breastfeeding ...................................................................................................... 25
2.5.2 Role of Mothers’ Experience of Breastfeeding on Decision to Breastfeed
............................................................................................................................ 26
2.5.3 Decisions Surrounding Methods to Provide Breast Milk.......................... 27
2.6 Clinical Decision Making Surrounding Breast- or Bottle-Feeding Initiation .. 28
2.7 Limitations of Current Research ...................................................................... 29
2.8 Research Directions and Significance .............................................................. 30
2.9 Conclusion ........................................................................................................ 30
Chapter 3: Systematic Review ................................................................................... 33
3.1 Introduction ...................................................................................................... 33
3.2 Abstract ............................................................................................................ 33
3.2.1 Background ............................................................................................... 33
3.2.2 Objectives .................................................................................................. 33
3.2.3 Search Methods ......................................................................................... 34
3.2.4 Selection Criteria ....................................................................................... 34
3.2.5 Data Collection and Analysis .................................................................... 34
3.2.6 Results ....................................................................................................... 34
3.2.7 Author’s Conclusions ................................................................................ 34
3.3 Plain Language Summary ................................................................................ 34
3.3.1 Instruments for Assessing Readiness to Commence Suck Feeds in Preterm
Infants: Effects on Time to Establish Full Oral Feeding .................................... 35
Chapter 4: Development of the Preterm Infant Readiness Scale ............................... 49
4.1 Introduction ...................................................................................................... 49
4.2 Overview .......................................................................................................... 49
4.3 Phase 1: Chart Audit ........................................................................................ 52
4.3.1 Background ............................................................................................... 52
4.3.2 Aim ............................................................................................................ 52
4.3.3 Research Questions ................................................................................... 53
4.3.4 Design ........................................................................................................ 53
4.3.5 Sample ....................................................................................................... 53
4.3.6 Data Collection Tools................................................................................ 54
ix
4.3.7 Procedure................................................................................................... 54
4.3.8 Data Analysis ............................................................................................ 55
4.3.9 Ethical Considerations .............................................................................. 56
4.3.10 Results ..................................................................................................... 58
4.4 Phase 2: Focus Groups ..................................................................................... 71
4.4.1 Background ............................................................................................... 71
4.4.2 Aims .......................................................................................................... 71
4.4.3 Research Questions ................................................................................... 71
4.4.4 Design ....................................................................................................... 72
4.4.5 Sample ....................................................................................................... 72
4.4.6 Data Collection Tools ............................................................................... 72
4.4.7 Procedure................................................................................................... 73
4.4.8 Data Analysis ............................................................................................ 74
4.4.9 Ethical Considerations .............................................................................. 74
4.4.10 Results ..................................................................................................... 76
4.4.11 Summary ................................................................................................. 86
4.5 Phase 3: Delphi Survey .................................................................................... 86
4.5.1 Background ............................................................................................... 86
4.5.2 Aim ............................................................................................................ 87
4.5.3 Research Question ..................................................................................... 87
4.5.4 Design ....................................................................................................... 87
4.5.5 Sample ....................................................................................................... 87
4.5.6 Data Collection Methods .......................................................................... 88
4.5.7 Procedure................................................................................................... 89
4.5.8 Data Analysis ............................................................................................ 90
4.5.9 Ethical Considerations .............................................................................. 91
4.5.10 Results ..................................................................................................... 92
4.5.11 Summary ................................................................................................. 99
Chapter 5: Testing of the Instrument ....................................................................... 101
5.1 Introduction .................................................................................................... 101
5.2 Phase 4: Observational Study ........................................................................ 101
5.2.1 Background ............................................................................................. 101
5.2.2 Aim .......................................................................................................... 101
5.2.3 Research Questions ................................................................................. 102
x
5.2.4 Design ...................................................................................................... 102
5.2.5 Sample ..................................................................................................... 102
5.2.6 Data Collection Tools.............................................................................. 103
5.2.7 Procedure ................................................................................................. 106
5.2.8 Data Analysis .......................................................................................... 107
5.2.9 Ethical Considerations ............................................................................. 109
5.2.10 Results ................................................................................................... 111
5.2.11 Summary ............................................................................................... 126
Chapter 6: Discussion ............................................................................................... 129
6.1 Introduction .................................................................................................... 129
6.2 Main Findings ................................................................................................ 129
6.3 Development of the PTSR Scale .................................................................... 130
6.4 Criteria to Commence Nutritive Suck Feeds .................................................. 132
6.4.1 Gestational Age ....................................................................................... 132
6.4.2 Medical and physiological stability ......................................................... 133
6.4.3 Infant Behaviour ...................................................................................... 134
6.5 Testing of the PTSR Scale ............................................................................. 135
6.5.1 Content Validity ...................................................................................... 136
6.5.2 Construct Validity ................................................................................... 136
6.5.3 Criterion-Related Validity ....................................................................... 136
6.5.4 Inter-Rater Reliability ............................................................................. 137
6.6 Strengths and Limitations............................................................................... 138
6.7 Implications for Research and Practice .......................................................... 140
6.7.1 Theoretical Implications .......................................................................... 140
6.7.2 Clinical and Policy Implications ............................................................. 142
6.7.3 Recommendations ................................................................................... 144
6.8 Conclusion ...................................................................................................... 145
Bibliography ............................................................................................................. 147
Appendices ............................................................................................................... 161
xi
List of Tables
Table 1: Priorities of quantitative and qualitative methods in each phase ................. 52
Table 2: Frequency of feeding methods at discharge................................................. 59
Table 3: Continuous infant characteristics by GA at birth ......................................... 60
Table 4: Interventions required during admission by GA at birth ............................. 61
Table 5: Type of delivery for twin births ................................................................... 62
Table 6: Maternal characteristics .............................................................................. 63
Table 7: Presenting obstetric complications .............................................................. 63
Table 8: Median age (PCA) in reaching feeding milestones ..................................... 64
Table 9: Median age at feeding milestones by GA group .......................................... 65
Table 10: Median time taken to reach feeding milestones ......................................... 67
Table 11: Median time taken to reach feeding milestones by GA group................... 67
Table 12: Effect of GA at birth on time taken to reach first breastfeeding attempt ... 68
Table 13: Model of variables associated with time taken to first feeding attempt ... 69
Table 14: Age criteria for preterm infants to attempt breastfeeding mentioned by
focus group ................................................................................................................. 77
Table 15: Stated reason for minimum age criteria mentioned by focus groups......... 78
Table 16: List of criteria to decide feeding readiness mentioned by all groups ........ 80
Table 17: List of criteria to decide feeding readiness not mentioned by all groups .. 81
Table 18: List of criteria to decide an infant is not ready to feed mentioned by all
groups ....................................................................................................................... 83
Table 19: Delphi survey sample characteristics ........................................................ 94
Table 20: Item CVI for three rounds .......................................................................... 98
Table 21: Sample characteristics of the preterm and fullterm infants ..................... 114
Table 22: Type of delivery for twin births .............................................................. 114
Table 23: Obstetric complications leading to delivery ............................................ 115
Table 24: Frequency and quality of previous feeding attempts ............................... 115
Table 25: Preterm and fullterm infants’ behavioural state just prior to feed time ... 119
Table 26: Preterm and fullterm infants’ transition between behavioural state during
handling .................................................................................................................... 120
Table 27: Preterm and fullterm infants’ display of feeding-readiness behaviours .. 120
xii
Table 28: Multi-variable model of total Preterm Infant Breastfeeding Behaviour
(PIBBS) and PTSR variables ................................................................................... 122
Table 29: Sample characteristics .............................................................................. 124
Table 30: Method of twin deliveries ....................................................................... 124
xiii
List of Figures
Figure 1: Model of the synactive organisation of behavioural development. ............ 11
Figure 2: Time taken to first attempt at breastfeeding by GA at birth ....................... 68
Figure 3: Summary of study participants ................................................................. 112
xiv
xv
List of Abbreviations
ANOVA Analysis of Variance
CI Confidence Intervals
CINAHL Cumulative Index to Nursing and Allied Health Literature
CNLD Chronic Neonatal Lung Disease
CPAP Continuous Positive Airway Pressure
CVI Content Validity Index
EBM Expressed Breast Milk
EFS Early Feeding Skills
EMG Electromyography
FRAPPS Feeding Readiness and Progression in Preterms Scale
GA Gestational Age
HR Heart Rate
HREC Human Research Ethics Committee
IBCLC International Board Certified Lactation Consultant
ICN Intensive-Care Nursery
IV Intravenous
MMH Mater Mothers’ Hospital
N-CRS Neonatal-Clinical Reporting System
NHMRC National Health and Medical Research Council
NNS Non-Nutritive Sucking
NOMAS Neonatal Oral Motor Assessment Scale
PCA Post-Conceptual Age
PIBBS Preterm Infant Breastfeeding Behaviour Scale
PIH Pregnancy-Induced Hypertension
PINFRS Preterm Infant Nipple Feeding Readiness Scale
PMA Postmenstrual Age
PNA Postnatal Age
PPROM Preterm Pre-labour Rupture of Membranes
PTSR Preterm Sucking Readiness
QUT Queensland University of Technology
xvi
RBWH Royal Brisbane and Women’s Hospital
RD Risk Difference
RDS Respiratory Distress Syndrome
RR Relative Risk
SCN Special-Care Nursery
SD Standard Deviation
SSB Suck-Swallow-Breath
SVD Spontaneous Vaginal Delivery
TPN Total Parenteral Nutrition
WMD Weighted Mean Difference
xvii
List of Appendices
Appendix A: Summary of Studies in Literature Review ......................................... 161
Appendix B: Search Strategy for Systematic Review ............................................. 175
Appendix C: Characteristics of Studies Excluded from Systematic Review ........... 177
Appendix D: List of Variable Fields Exported from the N-CRS ............................. 181
Appendix E: Chart Audit Form ................................................................................ 184
Appendix F: Participant Information Letter for Focus Groups ................................ 189
Appendix G: Participant Information Letter for Delphi Survey .............................. 191
Appendix H: Staff Information Letters for Observational Study ............................ 193
Appendix I: Information Letters for Parents with Babies in SCN ........................... 199
Appendix J: Information Letter for Parents of Fullterm Infants in Postnatal Ward 205
Appendix K: Parental Consent Form ....................................................................... 209
Appendix L: Delphi Surveys .................................................................................... 211
Appendix M Demographic sheet for observational study........................................ 219
Appendix N: Preterm Infant Breastfeeding Behaviour Scale (PIBBS) ................... 221
Appendix O: Preterm Sucking Readiness (PTSR) Scale ......................................... 223
Appendix P: Modified Kappa Statistics Tables ....................................................... 225
Appendix Q: Comments from Delphi Survey .......................................................... 229
xviii
xix
Publications
Submitted for Publication
Systematic Review: Instruments for assessing readiness to commence suck feeds in
preterm infants: effects on time to establish full oral feeding and duration of
hospitalisation. Submitted for publication to The Cochrane Collaboration October,
2010.
Conference Presentation (Paper Accepted)
Crowe, LM, Chang, AM, Tudehope, D & Fraser, J. Questioning current practice:
Using a systematic review to identify safe and effective assessment of preterm
infant’s readiness to feed. Paper presented at 16th International Nursing Research
Congress, Hawaii, July 2005.
Crowe, LM, Chang, AM, Tudehope, D & Fraser, J. Factors that influence feeding
initiation in preterm infants Paper Presented at Australian College Neonatal Nurses
Association Inc Annual Conference ‘Nurturing Generation Z’ April 19–20, 2008
Gold Coast.
Crowe, LM, Chang, AM, Tudehope, D & Fraser, J. Deciding Feeding Readiness in
Preterm Infants. Paper presented at Perinatal Society of Australia and New Zealand
(PSANZ) Annual Congress ‘Celebrating 25 years: Reflections and Projections’ April
20–23, 2008 Gold Coast.
Conference Presentation (Invited Speaker)
Crowe, LM. Feeding readiness in the preterm infant. Paper presented at RBWH
12th Annual Neonatal Conference. February 2010, Brisbane.
Awarded Grants for PhD Studies
Development of an instrument to assess a preterm to assess feeding readiness in
preterm infants
Grant successful February 2008 for funding of $14,348 from the Mater Mothers
Research Centre Grant
Linda Crowe, Anne Chang, David Tudehope
xx
Testing of a newly developed instrument to assess a preterm infant’s readiness to
commence breast feeds
Grant successful October 2007 for funding of $14,999 from the Mater Children’s
Golden Casket Fund.
Linda Crowe, Anne Chang, David Tudehope, Jenny Fraser
xxi
Statement of Original Authorship
This work contained in this thesis has not been previously submitted to meet
requirements of an award at this or any other higher education institution. To the best
of my knowledge and belief, the thesis contains no material previously published or
written by another person except where due reference is made.
xxii
xxiii
Acknowledgements
Firstly I would like to thank my supervisory team, Prof Anne Chang, Assoc Prof Jenny
Fraser and Prof David Tudehope for all their support and guidance. You have taught me
many valuable skills that will assist me throughout my career as a researcher. Your
enthusiasm and encouragement for my research project even when things were not going to
plan will not be forgotten.
To the mothers and babies who kindly participated in the research at both the Mater
Mothers’ Hospital and Royal Brisbane Women’s Hospital. Thank you for giving up your
time and allowing me to share in your early breastfeeding experiences.
I would also like to thank all the neonatal care providers across Australia and New Zealand
who participated in the development and testing of the PreTerm Sucking Readiness Scale. I
would especially like to acknowledge the staff from the Mater Mothers’ Hospital and Royal
Brisbane and Women’s Hospital. Without your input this thesis would not be possible.
Many thanks go to the neonatal nurses Judy, Jenny, Lisa, Philippa, Annie and Bekki who
acted as my research assistant at one time during this project. Thanks also go to Maree
Crepinsek who assisted in the training of the research assistants.
I would like to acknowledge the support that I received through the Mater Foundation and
QUT. I am very appreciative of the financial assistance provided.
To my friend Karen, we have experienced many ups and down over the last few years and I
am so glad I have shared this journey with you. You were the perfect “PhD buddy”.
I wish to thank all my family and friends for your love and support. In particular I would
like to thank my parents Phil and Maggie Crowe. You have been incredible role models who
have instilled in us the importance of education, hard work and helping others. This thesis
is dedicated to my father who was the one who put me on the path of my PhD.
Finally to my husband Adam, I am eternally grateful for your love, understanding, endless
patience and encouragement.
xxiv
1
Chapter 1: Introduction
1.1 Background
Infants born prematurely suffer a range of medical and developmental sequelae
related to the interruption of normal development in utero (Flint, Davies, Collie, &
Palmer-Field, 2007; McGrath & Medoff-Cooper, 2002). Preterm infants often lag
behind fullterm infants with regard to both physical growth and development.
Preterm infants are more likely than those infants born at term to experience poor
weight gain and feeding difficulties in infancy and through early childhood (da Costa
& Bos, 2008; Hawdon, Beauregard, Slattery, & Kennedy, 2000; Rommel, De Meyer,
Feenstra, & Veereman-Wauters, 2003). Further, many of the interventions that
preterm infants are exposed to during their stay in neonatal nurseries are not
supportive of a positive oral experience or a nurturing relationship between the infant
and their parents, both of which are needed for a positive feeding experience
(Sundseth Ross & Browne, 2002). Improved clinical practices, with regard to the
identification and management of preterm infants at risk of feeding difficulties
during their stay in neonatal nurseries, has the potential to affect short- and long-term
outcomes in relation to both growth and development (Hawdon, et al., 2000).
Feeding is a highly organised behaviour requiring the infant to coordinate sucking
and swallowing, while balancing the demands of respiration (Mason-Wyckoff,
McGrath, Griffin, Malan, & White-Traut, 2003). Most healthy, fullterm infants are
able to do this with minimal difficulty soon after birth (Lau, Smith, & Schanler,
2003; Mason-Wyckoff, et al., 2003). However, this is not the case for preterm
infants, who at birth do not have the physical development to consistently engage in
safe and efficient breast or bottle-feeding behaviour (Dodrill, McMahon, Donovan,
& Cleghorn, 2008). As a result, preterm infants are often fed via a tube placed into
the stomach until they are able to safely consume enough calories via sucking to fuel
baseline energy expenditure and to support growth (Crosson & Pickler, 2004). The
transition time from tube feeding to all suckle feeds may take a few days to several
weeks depending on the individual infant (Medoff-Cooper, Weininger, & Zukowsky,
2
1989; Thoyre, 2003). However, the two main feeding milestones that all preterm
infants must reach prior to being discharged home are the commencement followed
by the achievement of exclusive breast or bottle feeds (Dodrill, Donovan, Cleghorn,
McMahon, & Davies, 2008).
The optimal timing of commencement of breast or bottle feeds is not clear in the
literature or practice. In many nurseries, it has been a common practice to delay
breast or bottle feeds until the infant has reached 34 weeks post-conceptual age
(PCA) as this was the time that the coordination of sucking, swallowing and
respiration was thought to be established (Breton & Steinwender, 2008). However,
other studies have demonstrated that there are differences in the ability of infants to
engage in feeding behaviour with some infants developing these skills earlier or later
than the expected norm (Hedberg Nygvist, Rubertsson, Ewald, & Sjoden, 1996;
Lemons, 2001; Medoff-Cooper, Verklan, & Carlson, 1993). Differences in the ability
of infants to engage in feeding behaviour at a particular gestational age (GA) have
been shown through studies of preterm infant sucking (Hedberg Nyqvist, Rubertsson,
Ewald, & Sjoden, 1996; Lemons, 2001). Studies have found that although GA is a
guide to expected maturity, there are differences in the rates at which infants mature
(Hedberg Nygvist & Ewald, 1999; Simpson, Schanler, & Lau, 2002). These
differences are highlighted by Hedberg Nyqvist, Sjoden and Ewald (1999), who in
their study of preterm infants’ breastfeeding found that there were a small number of
infants that demonstrated efficient rooting, areolar grasp and latching on as early as
28 weeks PCA and nutritive sucking at 30.6 weeks PCA. However, this is not the
normal course of events for a ‘healthy’ preterm infant. The general pattern of
nutritive sucking in these infants is commencement of a suck-pause pattern at 32
weeks, which is then followed by a regular sucking rhythm at 34 weeks, until at 37
weeks, they have the rhythm and rate that is sufficient to ensure growth and
development via exclusive breast or bottle feeds (McGrath & Braescu, 2004).
The reason for some infants to deviate from the expected timing of progression of
feeding skills is still not clearly understood. There are a number of known factors
that affect preterm infants’ ability to engage in feeding behaviour including reduced
gastric capacity, poor intestinal motility, neurobehavioural immaturity, physiological
instability and immaturity of the gastrointestinal structures (da Costa & Bos, 2008;
3
Hill, 2002; McGrath & Braescu, 2004). Successful coordination of feeding is also
dependent on the adequate development of structures of the upper airway including
the lips, palate, jaw, tongue, pharynx, larynx and oesophagus (Cahill & Wagner,
2002b). The above factors are generally reliant on the physical development of the
infant and therefore, may not be amenable to intervention to improve feeding
capabilities. However, there is an increasing interest in the role that other factors may
have to improve an infant’s readiness to commence and develop feeding skills. These
factors include infant’s previous experiences, method of suckling, rate of flow of
milk, the presence of a nasogastric or orogastric tube during feeding and support
given to infant by the caregiver (Breton & Steinwender, 2008; da Costa & Bos,
2008). Interventions utilising these factors may lead to earlier commencement of and
quicker transition to exclusive breast or bottle feeds.
1.2 Significance
Safe feeding skills are a major determinant for discharging home, consequently, an
infant’s inability to feed results in increased duration of hospital stay and increased
cost of care (Pickler & Reyna, 2003; Simpson, et al., 2002; White-Traut et al., 2002).
Delayed discharge, simply because the infant is unable to breast and/or bottle feed,
could lead the infant to an increased risk of iatrogenic injury or infection (Crosson &
Pickler, 2004), separation-related effects on the parental relationship (McGrath,
2004) and increased cost to the health-care system (Burklow, McGrath, Valerius, &
Rudolph, 2002; Pickler & Reyna, 2003; Simpson, et al., 2002; White-Traut, et al.,
2002).
A 2008 report by the National Perinatal Statistics Unit states that preterm infants
(less than 37 weeks gestation) account for 8.2 per cent of the 294,737 annual live
births in Australia with the number of preterm infants born in Queensland being
slightly higher (8.6 per cent) than the national proportion (Laws, Li, & Sullivan,
2010 ). Although preterm infants account for a small percentage of the total number
of live births, they place a huge burden on the health-care system (McGrath &
Braescu, 2004). In the United States, one study found that the small percentage of
preterm infants (nine per cent of live births) accounted for 57 per cent of the annual
4
cost of neonatal care (St John, Nelson, Cliver, Bishnoi, & Goldenberg, 2000).
However, while cost-effectiveness is important, it is essential to ensure that the
introduction of breast and/or bottle feeding is safe, and will result in positive
outcomes both in human and financial terms.
Introducing suck feeds as soon as the neurologic development and physical condition
of the infant permits has several advantages, including shorter transition time to all
suck feeds, greater maternal satisfaction and shorter hospital stay (Pridham, Sondel,
Chang, & Green, 1993; Simpson, et al., 2002). However, feeding infants who are
unable to safely commence feeding may lead to suboptimal respiration, growth and
nutrition with the infants being at a higher risk of 1) aspiration pneumonia, 2)
readmission to NICU, 3) fatigue, 4) increased energy expenditure, 5) hypoxia and 6)
bradycardia and deglutition apnoea (Cahill & Wagner, 2002a; Case-Smith, Cooper,
& Scala, 1989; Hanlon et al., 1997). Therefore, careful timing is vital in order to
ensure that the commencement is beneficial rather than detrimental to the health of
the infant (Cousins, 1999).
Delayed commencement of suck feeding may also have psychological and social
impacts on both the infant and family. Sucking is a behaviour that infants find
enjoyable as it can lead to satisfaction of hunger as well as opportunities for an infant
to explore and engage with their environment (Harding, 2009). A delayed
introduction may lead to long-term behavioural aversions to feeding and greater
episodes of distress, as studies have shown that sucking has a calming effect on
preterm infants (Harding, 2009; Pinelli & Symington Amanda, 2005). Breast or
bottle feeding also allows an opportunity for social interaction as well as
development of social bonds (Harding, 2009; Holditch-Davis, Miles, & Belyea,
2000). In the early postnatal period, most interactions between mother and infant are
focused on feeding (Holditch-Davis, et al., 2000) with feeding success having
implications on fostering parent–child bonding as it is intimately intertwined with
perceived good parenting (Hall, Shearer, Mogan, & Berkowitz, 2002; McGrath &
Braescu, 2004). For mothers of preterm infants in particular, breastfeeding fosters
her involvement in the infant’s care and provides an opportunity for the mother to
touch and hold her infant (Black & Hylander, 2000).
5
Nurses in neonatal nurseries, in particular, are charged with the responsibility of
establishing and maintaining successful breastfeeding in the preterm infant
population. Unfortunately, it is well documented that an ad hoc approach is taken in
determining feeding readiness. One explanation for these inconsistencies in practice
may be the fact that the decision to commence feeding is not always made solely on
the use of criteria. This explanation is supported by McCain (2003), who found that
decisions on when to commence suck feeds were primarily based on established
nursery routines, caregiver preferences and what is convenient for staff at the time
rather than on pre-determined criteria of readiness. Another explanation for variation
in feeding practices could be the lack of agreed guidelines for neonatal care providers
on when to initiate breast or bottle feeding. Studies from Australia and the United
States have shown that while there are common criteria used as indicators of feeding
readiness such as GA and non-nutritive sucking (NNS) ability, no consensus has
been achieved among neonatal care providers (Burklow, et al., 2002; Pamela Dodrill,
et al., 2008; Kinneer & Beachy, 1994).
1.3 Instruments to Assess Breast- or Bottle-Feeding Readiness
The use of a formal screening instrument that encompasses an individual infant’s
behaviour and development has been suggested by clinical researchers as a way to
improve the accuracy of determining when an infant is ready to commence feeding
(McGrath, 2003; Meyer-Palmer & VandenBerg, 1998). Although many believe that a
screening instrument would be the ‘best way’ to determine feeding readiness,
currently, there are no instruments that have undergone rigorous psychometric
testing. Debate exists in both clinical practice and the literature in what the
instrument should encompass. Different parameters such as behavioural state,
maturity and severity of illness have been associated with feeding readiness;
however, there is little research on what factors or characteristics can actually predict
feeding success with most studies being conducted primarily in the bottle-fed
population (Pickler, 2005).
The possible benefits of a screening instrument to assess feeding readiness must be
weighed against the possible detrimental effects, additional staff time required and
6
other costs. Many clinicians have had to rely on clinical wisdom alone to decide
when to commence breast or bottle feeds. This has led to great variations in practice
with suckle feeding commencing at ages ranging from 26 to 36 weeks postmenstrual
age (PMA) (McCain, Gartside, Greenberg, & Lott, 2001). The impact of these
variations in practice is not known. It can be reasoned that clinicians may be
inhibiting some infants, who are behaviourally and physiological ready to commence
suckle feeding, while pushing other infants, who are less developmentally mature,
thus, increasing stress and detracting from success (McGrath & Braescu, 2004). A
possible benefit of using an instrument would be to decrease the variation of feeding
initiation practices between clinicians and allow for a more systematic and consistent
approach to deciding readiness.
1.4 Aims
The overall aim of this research was to develop and test an instrument to assess
breastfeeding readiness in the preterm infant population. The specific primary aims
of this research were to:
1. explore how clinical staff determine if a preterm infant is ready to engage in
breastfeeding so that items for possible inclusion in a breastfeeding readiness
instrument could be generated
2. reach consensus on which items (criteria) should be included into the
instrument to assess feeding readiness
3. test the newly developed Preterm Sucking Readiness (PTSR) scale to
establish both construct and criterion-related validity as well as inter-rater
reliability
To assist in the development of the PTSR scale, the secondary aim of this study was
to gain a greater understanding of infant factors that may influence the decision that a
preterm infant is ready to engage suck feeding behaviour. The specific objectives
were to explore: 1) age as a criterion to determine feeding readiness 2) factors
associated with prolonged time taken to reach first attempt at nutritive sucking.
7
1.5 Thesis Structure
This chapter introduced the concept of feeding readiness and discussed the current
methods clinicians utilise to decide if a preterm infant is ready to engage in feeding
behaviour. Further, this chapter also briefly discusses the possible impacts of
mistimed introduction of feeding and how clinicians working with this vulnerable
population may improve their decision-making process through the use of a feeding
readiness assessment instrument.
Chapter 2 provided a discussion on how the synactive theory of development may be
used to explain the general development of behaviour in preterm infants. Moreover,
this chapter describes the development of suck feeding behaviour and what factors
may affect individual infant’s ability to feed. The discussion within this chapter
describes the current methods that are used to make a decision to determine readiness
to engage in feeding. Finally, limitations of current research as well as further
research directions were also explored.
Chapter 3 is a systematic review that searches for existing instruments or assessment
tools to decide feeding readiness in preterm infants. Any instruments found were to
be evaluated on the basis of their effect on the outcomes of time to establish feeding
and duration of hospitalisation. The results of this systematic review are discussed in
the context of the aims of this thesis.
Chapter 4 gives a brief overview of the methodology of Mixed Methods. It also
describes the methods and results of the first three phases of this research. These
three phases describe the development of the PTSR scale.
Chapter 5 incorporates the methods and results of the final phase of the study, the
testing of the newly developed PTSR scale.
Chapter 6 is the final chapter that provides an overview of the principal findings and
acknowledges the research study’s limitations and strengths. It also provides
recommendations for clinical practice and further research.
8
9
Chapter 2: Literature Review
2.1 Introduction
Chapter 2 provides a comprehensive literature review on the current knowledge of
development of nutritive feeding skills. Included in this chapter is also a discussion
on how the synactive theory of development has been used to underpin this study by
explaining how infants are able to communicate their readiness to commence
breastfeeding to caregivers. Further, this chapter describes what infant-related factors
may affect the initiation of nutritive sucking in the preterm infant population and
thus, may be used by neonatal care providers to decide when to initiate nutritive suck
feeding. A discussion regarding the maternal decision to initiate breast or bottle
feeding is also included. In addition, the neonatal care providers’ decision to initiate
nutritive suck feeds in the preterm infant population is placed within a clinical
decision-making context. Finally, limitations of current research as well as further
research directions were also explored.
2.2 General Development of Behaviour in Preterm Infants
2.2.1 Synactive Theory of Development
Als’ (1986) synactive theory of development provides a framework for
understanding the general development of behaviour in infants as well as
understanding the development of specific behaviours such as feeding readiness
(McGrath, 2004). According to this theory, infants are seen as social beings that are
in ‘continual interaction with their environment via five subsystems: autonomic,
motor, state, attentional, interactive and self-regularity’ (Holditch-Davis, Blackburn
Tucker, & VanderBerg, 2003). These subsystems mature simultaneously, and within
each subsystem a developmental sequence can be observed (see Figure 1) with
subsystems being both interdependent and interrelated with each other in which
disorganisation or immaturity in one subsystem will influence other subsystems (Als,
Butler, Kosta, & McAnulty, 2005; Holditch-Davis, et al., 2003).
10
In this framework infants strive to maintain homeostasis. Preterm infants are more
likely to exhibit difficultly in switching between the subsystems than fullterm
infants, who are able to do this smoothly and consistently. Immaturity of the
autonomic subsystem is illustrated through changes in heart and respiration rates,
skin colour and visceral signs such as twitching or stooling (Holditch-Davis, et al.,
2003). Motor control allows infants to maintain adequate tone and posture to engage
in feeding behaviour and is closely linked to the state subsystem (Als, 1986).
Organisation of the state subsystem refers to the ability of an infant to transition
between and maintain a level of awareness on the continuum from asleep to fully
aroused and crying (McGrath, 2004). Ability to maintain and move between sleep
and awake states increases over the preterm period (White-Traut, Berbaum, Lessen,
McFarlin, & Cardenas, 2005). The maturation of the attentional or interactive and
self-regularity subsystems is vital and allows the infant to demonstrate to the
caregiver how well they are coping. The attentional or interactive subsystem refers
to an infant’s ability to focus on the task at hand and their ability to engage in social
interaction, which is also dependent on the infant’s state subsystem (Holditch-Davis,
et al., 2003). Maturation of the self-regularity subsystem allows the infant to
maintain homeostasis despite a challenge (Vergara & Bigsby, 2004). Simply, the
ability to self-regulate refers to an infant being able to react to a challenge by
increasing their arousal state and respond with appropriate changes in heart rate (HR)
and breathing and return to homeostasis without external assistance (Vergara &
Bigsby, 2004). Preterm infants tend to be initially unstable when faced with any
challenge to their homeostasis resulting in sudden changes to their autonomic, motor
and state subsystems (Holditch-Davis, et al., 2003). As the infant matures, it is less
likely for the infant to totally decompensate; rather, there will be less dramatic signs
(Holditch-Davis, et al., 2003). Maturation of all the subsystems is required for
successful feeding (McGrath, 2004).
11
Figure 1: Model of the synactive organisation of behavioural development in Als
(1986).
A synactive model of neonatal behavioural organisation: Framework for the assessment of
neurobehavioural development in the premature infant and for support of infants and parents in the
neonatal intensive-care environment. Physical and Occupational Therapy in Paediatrics 6 (3–4), p.
15.
2.2.1.1 The Importance of Synactive Theory of Development in Providing Care to
Preterm Infants
As a consequence of this theory, neonatal care providers have re-examined how they
provide care to preterm infants. Historically, the care giving model was strictly
medical; however; now it is also incorporating early developmental interventions,
which are referred to as developmental care (Shaker, 1999). Developmental care is
based on Al’s synactive theory of development and involves interventions taken to
support the behavioural organisation of each individual infant, enhancing
physiological stability, protecting sleep rhythms and promoting growth and
12
maturation (Als, 1998). These interventions include handling and positioning
measures, reduction of noxious environmental stimuli and cue-based care. Providing
care in this way has been reported to have positive impacts on both current
neurobehavioural development as well as long-term developmental outcomes
(Pickler, 2005).
Therefore, this theory provides an overall appropriate framework of understanding
that assumes that a preterm infant can actively communicate its readiness to engage
in a particular activity, such as breast or bottle feeding, in a dynamic process of
continuous feedback between the infant and caregiver. Further, this theory has
highlighted a framework, which neonatal care providers may use to assess feeding
readiness when making decisions with regards to if an infant is to commence
nutritive suck feeding. The subsystems comprised in the synactive theory of
development have guided the areas that needed to be included into the instrument.
2.2.2 Neurobehavioural Development and Feeding Ability
Assessment of a preterm infant’s capacity to feed has also been seen as an important
‘window’ into the nature of the developing infant through allowing in-depth
examination of the integration of the neonate’s central nervous system (McGrath,
2004). The ability to transition from tube feedings to breast or bottle feeding is
dependent on the neurobehavioural status of the infant (Delaney & Arvedson, 2008).
Neurobehavioural maturity is indicated when an infant is able to coordinate the
autonomic, motor and behavioural state-organisation subsystems and requires the
development of the nervous system to support the integration and smooth
coordination of sucking, swallowing and breathing (McGrath, 2004). Maturation has
been shown to alter the modulation, presentation and availability for nutritive
sucking and interaction with the environment and therefore, feeding can be seen to
reflect the level of maturation of the developing brain (McGrath & Medoff-Cooper,
2002). This premise is based on breast or bottle feeding being a highly complex and
organised process (McGrath, 2004; Medoff-Cooper, McGrath, & Bilker, 2000)
requiring the infant to integrate the five subsystems described in the synactive theory
of development, thus, demonstrating central nervous system maturation (Als, 1986).
13
Neurobehavioural organisation appears to mature with increasing PMA with
differences noted between preterm and fullterm infants. In a study by Mouradian, Als
and Coster (2000), significant differences were found between preterm and fullterm
infants in relation to neurobehavioural maturity. The study consisted of 42 healthy
newborn infants aged at birth at either 40 weeks (n = 16); 37 weeks (n = 13) or 34
weeks (n = 13), in which, infants were compared. Neurobehavioural functioning was
measured by the Assessment of Preterm Infants’ Behaviour (APIB) system which
incorporates 12 variables including measures of autonomic, motor, state,
attention/interaction and self-regulatory sub-systems as well as a measure of overall
behavioural organisation. Significant differences between the three groups were
found, including that the preterm group (34 weeks GA) differed for all twelve
variables tested against the fullterm group. Difference between 37 and 40 weeks
were also seen, but not as dramatic (four out of the twelve variables). Results suggest
that with decreasing GA, infants are more likely to have motor system
disorganisation, potential difficulties with autonomic control, self-regulation and
attentional ability, which are all skills needed for feeding (Mouradian, et al., 2000),
Although there does appear to be a relationship between neurobehavioural maturity
and the development of feeding behaviour it is unclear what impact that early
experiences in feeding may have on the neurobehavioural development of the infant.
Specifically, it is unknown if neurobehavioural maturation develops feeding
behaviour or vice versa.
2.3 Development of Feeding Behaviour in Preterm Infants
A number of studies were identified that investigated the development of breast and
bottle-feeding behaviours. In Appendix A, the relevant details for each study have
been listed. Many of these studies were small and had included participants of
varying degrees of prematurity and morbidity making it difficult to apply findings to
all preterm infants. However, the findings do show trends and provide directions for
further research. A summary of these findings have been described below.
14
2.3.1 Development of Oral Structures and Function
Successful coordination of feeding is dependent on the adequate development of
structures of the upper airway including the lips, palate, jaw tongue, pharynx, larynx
and oesophagus (Hill, 2002). In other words, an infant’s ability to establish
successful feeding by breast or bottle will primarily depend on the degree of
structural and functional development of their GI tract and oral cavity (McGrath,
2004).
The functional development of sucking is dependent on the presence of adaptive and
protective reflexes. Adaptive reflexes, such as rooting and sucking, assist the infant
to direct the feed into the gut. Whereas, protective reflexes, such as gag and cough,
keep the feed away from the infant’s airway. Burns et al. (1987) investigated
protective and adaptive reflexes in infants ranging from 25 to 37.5 weeks PMA. The
investigators found that the gag reflex could occasionally be stimulated in 90 per cent
of infants at 29 weeks, several weeks before the age of 33 weeks in which the reflex
could be consistently evoked. All other reflexes such as rooting, coughing and NNS
were not consistently mature until 37 weeks. The development of the oral responses
was not significantly altered by age at birth, birth weight, mechanical ventilation or
gender. However, the presence of a grade III or IV intraventricular haemorrhage
significantly affected rooting, NNS and suck-swallow. It appears from the results of
this study that oral feeding should not be initiated before 33 weeks PMA and that
care should be taken until the cough reflex is consistently present at 37 weeks.
Nevertheless, individual infant differences were found within the study with some
infants consistently displaying more mature oral responses than infants of older ages.
2.3.2 Development of Sucking Behaviour
Ultrasound studies have shown that oral structures as well as sucking function
develop in utero in a predictable and stepwise order. From ten to sixteen weeks
gestation, jaw and lip movements can be observed starting with simple mouth
opening to repetitive open–close movements (Miller, Sonies, & Macedonia, 2003).
Swallowing can also be seen as early as ten to fifteen weeks; however, at this time, a
bolus of amniotic fluid is not brought into the mouth by a sucking movement, but
15
rather by an inhalatory-like movement (Devries, Visser, & Prechtl, 1985; Miller, et
al., 2003). It is not until approximately 28 weeks gestation that lingual movements
become more complex with anterior posterior suckling-like movements needed for
postnatal feeding are observed. Pre-swallowing stimulatory behaviours are seen
consistently in infants 26–36 weeks GA (Miller, et al., 2003). These behaviours may
involve infants sucking on their own fingers, hands, umbilical cord or touching their
cheeks with their own hands before swallowing. Therefore, it appears from
intrauterine studies that foetuses develop and perform complex sucking functions
well before the age of 34 weeks, the age in which many clinicians believe that infants
can be initiated at either the breast or bottle. However, these types of studies do not
necessarily indicate that an individual infant has sufficiently matured to be able to
breast or bottle feed.
Several studies have also explored the development of sucking function after the
infant is born. Infants have two types of sucking behaviour, nutritive and NNS
(Pinelli, Symington, & Ciliska, 2002). Nutritive sucking involves the intake of fluid
due to alterations in expression and suction with a mature sucking pattern being one
suck per second. In contrast, NNS is faster and does not involve the intake of fluid,
but can satisfy an infant’s urge to suck (Harding, 2009). While each type of sucking
behaviour has its own purpose, NNS is considered as a precursor to nutritive sucking
(Pinelli, et al., 2002).
2.3.3 Development of NNS
In a small study (n = 6) by Hack, Estabrook and Robertson (1985), it was found that
rhythmical NNS was demonstrated by one infant at 28 weeks PCA; however, this
was not demonstrated in the other infants until 31–33 weeks GA. Overall NNS
abilities seemed to increase with GA with the number of sucking bursts per minute
increasing as well as an increasing sucking pace within each sucking burst. The
findings of Hack (1985) were confirmed by a larger study conducted by Hafstrom
and Kjellmer (2000), who also found that infant’s sucking activity as well as sucking
frequency and sucking amplitude all improved with increasing age. The difference
between the two studies is that in the first study by Hack et al. (1985), it was found
that infants from about 30 weeks PCA did not alter the duration of sucking bursts,
16
while in the later study, they found sucking burst duration increased with age.
Interestingly, in the Hafstrom and Kjellmer study, it was also found that while
increasing age was the dominant predictor, gender, state of activity and weight also
influenced NNS abilities (Hafström & Kjellmer, 2000).
When NNS abilities were compared between infants who were born at term and
preterm infants with an equivalent age, the preterm group demonstrated a distinctly
different NNS pattern with a higher frequency and lower amplitude than the infants
born at term (Lundqvist & Hafstrom, 1999). In the preterm group, gender differences
were observed with female infants having higher frequency and larger amplitude.
The fullterm infants’ NNS patterns were also related to pacifier use during the first
few days of life. Experience, gender, maturity and level of tension are suggested as
explanatory factors for differences in NNS patterns.
NNS has also been seen as an intervention to improve physiological stability and
nutrition in preterm infants. A systematic review was conducted by Pinelli and
Symington (2005) in which 21 studies were found. NNS was found to significantly
decrease the length of hospital stay of preterm infants. The review did not
demonstrate a consistent benefit of NNS with respect to the other major clinical
variables (weight gain, energy intake, HR, oxygen saturation, intestinal transit time,
age at full oral feeds and behavioural state). However, the review did identify other
positive clinical outcomes of NNS including quicker transition from tube to bottle
feeds and better bottle-feeding performance. No negative outcomes were reported in
any of the studies.
2.3.4 Development of Nutritive Sucking
There have been a number of studies examining nutritive sucking patterns with the
studies near exclusively investigating the development of nutritive sucking in the
bottle-feeding population.
Several studies have investigated the role of GA at birth in the preterm infant
population. Medoff-Cooper (1991) showed that after 34 weeks PCA, a modest trend
for mean maximum sucking pressure existed; however, there was no trend across
17
GAs for the variables of duration of sucking burst or the number of sucks per sucking
burst. In addition, the pauses between sucking bursts also demonstrated a decreasing
trend from 32 to 36 weeks PCA. Gewolb, Vice, Schwietzer-Kenney, Tacial and
Brown (2001) also discovered a correlation in nutritive sucking skills and age.
However, in their study, they found the stability of the swallow did not change
significantly from 32 to 40 weeks PMA. Interestingly, sucking skills correlated more
strongly with PMA than with postnatal age (PNA). This finding suggests that oral
feeding is a congenital behavioural pattern rather than an acquired behaviour.
However, there is some evidence to suggest that experience plays a complementary
role in the development of nutritive sucking (Hedberg Nygvist, Sjoden, & Ewald,
1999; McGrath, 2004; Medoff-Cooper, McGrath, & Shults, 2002; Pickler, Best,
Reyna, Wetzel, & Gutcher, 2005).
Other studies have examined the relationship between feeding efficiency and GA.
Casaer, Daniels, Devlieger, De Cock and Eggermont (1982) studied 100 preterm
infants and reported that feeding efficiency as measured by volume consumed per
minute was greater in infants greater than 34 weeks GA than in those who were less
than or equal to 34 weeks. Additionally, they reported that there was a significant
correlation between feeding efficiency and duration of feeding experience at most
ages between 32 and 37 weeks. Martell, Martinez, Gonzalez and Rossello (1993)
also reported that feeding efficiency varied with GA. In this study of 62 infants, it
was found that the velocity of milk extraction during sucking was uniform at lower
GAs and then gradually became faster in the first minutes of the feed and, at the 36th
week, was somewhat similar to that of mature infants. This indicates that for preterm
infants, feeding efficiency improves with maturity.
It is noteworthy that studies examining the relationship between the development of
nutritive sucking behaviour and preterm infants’ oral feeding skills have found that
not all measures of sucking development consistently correlate with feeding
efficiency. In a study of eighteen preterm infants conducted by Daniels, Casaer,
Devlieger and Eggermont (1986), differences in feeding efficiency could not be
explained by differences in sucking rate alone. Quick efficient feeding was
characterised by long sucking bursts and by a large amount of milk intake during
each sucking movement. Conversely, slow inefficient feeding was characterised by
18
short sucking burst and by a small amount of milk intake during each sucking
movement (Daniels, Casaer, et al., 1986). Weaver and Anderson (1988)
demonstrated in a study of 30 preterm infants that there was no significant
correlation between sucking pressures and feeding scores, but there was a significant
correlation between the two feeding scales indicating that suck pressures alone
should not be used as an indicator of feeding efficiency. However, Lau,
Alagugurusamy, Schanler, Smith and Schulman (2000) developed a five-point
sucking scale, which incorporated both measures of development, nutritive sucking
as well as feeding efficiency, and demonstrated that as oral feeding performance
improves sucking skills also improve.
Differences in the development of sucking behaviour were investigated between
infants born at term and preterm infants who have reached a similar age. Medoff-
Cooper, Weininger and Zukowsky (1989) reported in their study of bottle-feeding
infants that fullterm infants had sucking patterns that were significantly different than
those of preterm infants. Fullterm infants had higher maximum suck pressures and
consumed more formula per suck, had more sucks per burst and maintained the
pressure longer for a wider suck width. A later study found similar results showing
that feeding efficiency significantly improved with age in fullterm infants
demonstrating lower number of sucks, greater intensity of sucking pressures, less
average time between sucking bursts and greater average time between sucks than
preterm infants (Medoff-Cooper, et al., 2000). Although there were improved
sucking parameters at full term, no relationship was found at either 34 or 40 weeks
with regards to sucking parameters and the variables of neurobehavioural
development score and infant’s behavioural state. This lack of relationship indicates
that maturation of central nervous system may not be the only contributor to preterm
infant’s ability to commence oral feeding.
2.3.4.1 Development of Coordination of Suck-Swallow-Breath
Another important aspect of an infant’s ability to safely feed is the coordination of
sucking, swallowing and breathing (SSB). A problem in any one of these functions
or a problem in coordinating all three functions can have a significant effect on
preterm infant’s sucking abilities. As swallowing and breathing utilise a common
19
space, inspiration, swallow and then expiration is the safest way for the infant to
coordinate sucking.
There is growing evidence to suggest that coordination of SSB in the preterm infant
population is correlated with increasing PMA. Mizuno and Ueda (2003) studied 24
preterm infants (28 to 31 weeks of gestation at birth), in which they found that
nutritive sucking behaviour significantly increased between 33 and 36 weeks PCA.
Between the ages of 32 and 33 weeks, swallowing occurred most commonly during
pauses in respiration; however, after 35 weeks, swallowing usually occurred at the
end of inspiration, with swallowing infrequently interrupting respiration.
Lau et al. (2003) also studied the coordination of SSB. In their study of both preterm
(n = 12) and fullterm (n = 8) infants, they demonstrated that as a preterm infant’s
amount of milk transfer improved, sucking and swallowing frequency, bolus size and
suction amplitude also increased. The investigators also found that preterm infants
swallowed preferentially at different phases of respiration than their fullterm
counterparts indicating that feeding difficulties in the preterm infant population may
be the result of uncoordinated SSB.
A study by Daniels, Devlieger, Casaer, Callens and Eggermont (1986) suggests that
nutritive sucking and NNS have different effects on SSB in the preterm infant
population. Nutritive and NNS was studied in nine preterm infants with PMAs
ranging from 28 to 33 weeks and PNAs ranging from zero to eight weeks (Daniels,
Devlieger, et al., 1986). During nutritive sucking, the respiratory rate was higher
during the pauses than during the bursts. During NNS, the respiratory rate was higher
during sucking. Further, there were also differences noted in transcutaneous oxygen
levels with higher transcutaneous oxygen levels being observed during NNS
(Daniels, Devlieger, et al., 1986). Therefore, NNS may not serve as a model for
studying feeding mechanisms in the preterm infant; however, it may be used to
indicate if the infant is not ready for nutritive feeding. If infants cannot maintain their
transcutaneous oxygen level during NNS, it is unlikely that they would be able to do
it during nutritive sucking.
20
2.3.5 Development of Breastfeeding Behaviour
As previously described, there are very few studies that specifically measure the
development of breastfeeding behaviour in preterm infants. Hedberg Nyqvist,
Farnstrand, Edebol Eeg-Olofsson and Ewald (2001) used surface electromyography
(EMG) and direct observation to investigate the development of breastfeeding
behaviour in 26 preterm infants. During this study, electrode pads were placed over
anatomical sites in the face and neck. Although EMG has been used before in
studying bottle-feeding infants, this was the first time that EMG had been used in
breastfeeding preterm infant population. The results of the EMG were highly
correlated with the direct observations of sucking. Interestingly, there was no
relationship with PMA for most of the variables including number of sucks per burst,
duration of sucking, duration of mouthing or pauses.
In another study by the same lead investigator, it was demonstrated that preterm
infants (n = 71) were initiated at the breast between 27.9 and 35.9 weeks PMA
(Hedberg Nygvist & Ewald, 1999). Although the investigators were only able to
observe the first contact at the breast for 36 infants, it is interesting to note that
irrespective of PMA, 34 (95.5 per cent) of these infants displayed rooting behaviour
on the first contact at the breast. Efficient rooting, areolar grasp and latching on were
observed at 28 weeks, and repeated bursts of less than or equal to ten sucks and
maximum burst of greater or equal to 30 sucks was observed at 32 weeks PMA.
Nutritive sucking was defined as confirmation by test weighing that the volume of
milk that the infant consumed was greater than or equal to five grams. This first
nutritive suck was recorded at 32.0 PMA. Although this study informs researchers
and clinicians about the minimum age at which specific breastfeeding behaviours
emerge, it does not test for the effect of age with only descriptive data analysis being
presented. Secondary analysis of this study describes factors that may affect the
development of breastfeeding behaviour. Efficient feeding performance was
predicted by higher birth weight, less need of ventilator and oxygen treatment,
absence of bottle feeding, no need for apnoeas treatment and no suspicion of
infection (Hedberg Nygvist & Ewald, 1999).
21
Furman and Minich (2004) studied 105 very low weight (≤1.5 kg) preterm infants to
compare the feeding efficiency between breastfed (n = 35) and bottle-fed infants (n =
70). In this study, breastfed infants took in smaller volumes, fed less efficiently and
spent less time with suckling bursts. The bottle-feeding and breastfeeding groups did
not differ in age at the time of observation, in birth data, in measures of neonatal
morbidity or in their neurobehavioural development. As these results are from only
one study and involve a particular subgroup of preterm infants, it is difficult to
determine the feeding efficiency of other preterm infants. In terms of feeding
readiness, this slower pace of intake may actually be beneficial over bottle feeding
when infants are first attempting to breastfeed with regards to better integration of
breathing and swallowing as well as more time in close contact with their mothers.
2.4 Infant Factors Related to Readiness to Commence Breast- or
Bottle Feeding
A number of infant-related factors have been highlighted in the literature as being
associated with the feeding readiness and may explain variation in different infant’s
ability to engage in sucking behaviour and determine whether neonatal care
providers decide to initiate nutritive sucking behaviour. As discussed below, these
factors include behavioural state, feeding-readiness behaviour cues, severity of
illness and physiological stability.
2.4.1 Behavioural State
Medoff-Cooper et al. (Medoff-Cooper, et al., 2000) defined behavioural state as a
group of behaviours that indicate a level of arousal along the continuum from sleep
to full arousal. Behavioural state is thought to be an important factor with regard to
feeding commencement as infants need sufficient behavioural state organisation to
remain awake. This occurs usually around 32 weeks when infants express a full
range of behaviours. However, expression is ‘less robust and more prone to
disorganisation than that of a fullterm infant’ (McCain, 2003, p. 46). Consequently,
behavioural states in preterm infants are generally less well defined and lack clarity,
22
leading to difficulties in determining when an infant is ready to engage in feeding
behaviour (McGrath & Medoff-Cooper, 2002).
Behavioural state has been positively correlated with readiness and availability to
engage in oral-nipple feeding opportunities (Medoff-Cooper, et al., 2000). A positive
relationship between level of alertness during a feed and successful outcome has also
been established (McCain, 1997; McGrath & Medoff-Cooper, 2002). However,
inconclusive evidence exists with McGrath & Medoff-Cooper (McGrath & Medoff-
Cooper, 2002) demonstrating that there is no significant effect found on robustness
of alertness prior to feeding on indicators of nutritive sucking competence.
2.4.2 Feeding-Readiness Behaviour Cues
Feeding-readiness behaviour cues can be defined as clusters of infant behaviours that
an infant exhibits to show its availability to engage in feeding behaviour (McGrath &
Braescu, 2004). Fullterm infants clearly demonstrate their desire to feed with clear
hunger signs such as loud crying (McCain, 2003). The preterm infant’s signals are
more subtle, making decisions with regard to signs of hunger and feeding readiness
more difficult. A number of pre-feeding behaviours have been highlighted in the
literature including: empty sucking, sucking on fingers, sucking on hands, sucking on
tongue, swipes at mouth, hand to mouth, tonguing and rooting (Cagan, 1995;
McGrath & Braescu, 2004; White-Traut, et al., 2002). There is limited evidence to
support the use of feeding-readiness behaviours to predict feeding efficiency (White-
Traut, et al., 2005).
2.4.3 Severity of Illness
Studies have shown that there is a negative relationship between age and severity of
illness with the youngest infants more likely to experience greater morbidity than
those born close to term (McGrath, 2004; Thoyre, 2003) resulting in some
researchers using PCA as a measure of maturity (Thoyre & Brown, 2004). Chronic
conditions often seen in the extremely preterm infants may influence the introduction
of breast or bottle feeds. A study of infants with bronchopulmonary dysplasia (BPD)
found that GA, days on mechanical ventilation, continuous positive airway pressure
23
or supplemental oxygen influence age at which breast or bottle feeds are commenced
(Pridham, et al., 1993). The severity of illness experience has also been related to
efficiency of feeding (Pickler, et al., 2005) and length of time taken to reach full
enteral feeds (Hook Morris et al., 1999).
2.4.4 Physiological Stability
Considering the variation in estimating PCA and the natural variation in maturation,
the transition from gavage to oral feeding is a period of risk for the preterm infant
because oral feeding requires an integration of SSB reflexes all of which are vagally
mediated (McCain, 2003, p. 47). Infants aged less than 32 weeks PCA are at a
greater risk for both apnoea (cessation of respiration for >20 seconds) and
bradycardia (HR of less than 100 beats per minute) (McCain, 2003); after this period
the autonomic system undergoes significant maturation (McCain, Fuller, & Gartside,
2005). Preterm infants’ immaturity of the autonomic system places them at high risk
for developing feeding bradycardia as they transition from gavage to oral feeding;
however, infants may have periods of bradycardia and apnoea not related to a
feeding event but autonomic immaturity alone, which is particularly true for infants
less than 34 weeks PCA (McCain, 2003).
Research findings have suggested that preterm infants remain more physiologically
stable during early breastfeeding than during early bottle feeds (Chen, Wang, Chang,
& Chi, 2000). Differences have been found between the two groups with breastfed
infants, who were found to breathe more during sucking bursts when compared to
bottle-feeding sessions and had fewer episodes of oxygen desaturation during
breastfeeding (Dowling, 1999). This is confirmed by other studies that have found
the infants who are bottle fed are at greater risk of a lower oxygen status (Thoyre,
2003).
24
2.5 Maternal Factors Related to Breast- and Bottle-Feeding
Initiation
The World Health Organization considers exclusive breastfeeding until six months
old to be the optimal feeding method for all infants (Lee, Lee, & Kuo, 2009). In
particular, preterm infants have shown to benefit from receiving breast milk. Benefits
include a lower incidence of infections including necrotising enterocolitis, decreased
rates of retinopathy of prematurity, shorter hospitalisation and fewer readmissions, as
well as higher scores on cognitive and developmental tests (Buckley & Charles,
2006; Sisk, Quandt, Parson, & Tucker, 2010; Sweet, 2008b). For these reasons
mothers of preterm infants are encouraged by health-care providers to provide breast
milk for their infant.
A mother’s decision to provide breast milk or infant formula is seen as a personal
choice and a product of rational decision making (Sheehan, Schmied, & Barclay,
2010). However, some argue that it is a more complex process, as the decision to
provide breast milk or breastfeed may be influenced by other factors such as
mothers’ beliefs and attitudes surrounding breastfeeding and external pressures from
family or health professionals (Sweet, 2008a). Consequently, there are a growing
number of studies into the socio-demographic, biomedical and psychosocial factors
that affect parental infant feeding decisions (Sheehan, et al., 2010). The majority of
research examining breastfeeding initiation and maintenance have focused on
mothers of fullterm infants with few studies specifically investigating women whose
infants were born preterm (Sisk, et al., 2010; Wheeler, Chapman, Johnson, &
Langdon, 2000). Although there are some common barriers for mothers of both
fullterm and preterm infants, mothers of preterm infants have their own challenges as
a result of immaturity and admission to a neonatal critical care unit (Buckley &
Charles, 2006).
25
2.5.1 Factors Affecting Mothers’ Decision to Initiate and Maintain
Breastfeeding
There are a number of quantitative studies that investigate the predictive factors of
breastfeeding initiation and maintenance within the preterm infant population
(Pineda, 2011). Results of these studies are at times conflicting with some studies
finding positive (Powers, Bloom, Peabody, & Clark, 2003), inverse (Colaizy &
Morriss, 2008) and no relationship (Andrews Espy & Senn, 2003; Lessen & Crivelli-
Kovach, 2007; Pineda, 2011) between breast milk feeding and the variables of GA
and birth weight. Studies that have demonstrated that the higher incidence of breast
milk feeding in infants of higher GAs and higher birth weights can be explained by
the fact that mothers of infants who are born closer to term may have less difficulties
to establish and maintain breast milk supply as they are less likely to have to express
breast milk and be separated from their infants for an extended period of time.
Whereas, higher breast milk feeding rate in the lower GA and birth weight infants
may be explained by the fact that these mothers having a greater desire to provide
breast milk to their infants due to the perceived special benefit to their smaller and/or
more premature infant. On the other hand, no relationship between breast milk
feeding with either GA or birth weight may be explained by the decision being made
to provide breast milk prior to their infant’s birth, with the occurrence of the preterm
birth not changing mothers’ minds (Sweet, 2008a). Maternal factors have also been
found to be related to breast milk feeding initiation and therefore, may influence
mother’s decision making to provide breast milk for her infant. These factors include
age, education, parity, race, smoking and marital status, pre-breastfeeding experience
and influence of father and prenatal education (Andrews Espy & Senn, 2003; Lessen
& Crivelli-Kovach, 2007; Pineda, 2011; Wheeler, et al., 2000).
Other qualitative studies have focused on the barriers of expressing breast milk,
which may also impact the mother’s decision to commence or cease breastfeeding. In
particular, there were a number of barriers in relation to the initiation of expressing
milk for their preterm infants that act as inhibitors. These include pregnancy
complications, anxiety surrounding their infant’s health and lack of privacy (Sisk, et
al., 2010). Once breast milk expression was established and mothers had been
discharged home, they encountered a new set of challenges, including using manual
26
or smaller electric breast pumps, travelling to the hospital, return to work and
difficulties with time management interfering with maintenance of breast milk
production (Sisk, et al., 2010). Identified factors of mother’s maintenance of
expressing breast milk were more related to emotional and social support, for
example, positive attitude to pumping and anticipation of breastfeeding assistance
(Sisk, et al., 2010).
2.5.2 Role of Mothers’ Experience of Breastfeeding on Decision to Breastfeed
Understanding the mother’s experience of breastfeeding preterm infants may also
highlight reasons why some women choose to initiate or cease breastfeeding. Few
qualitative studies have examined the mothers’ experience of breastfeeding their very
low-birth-weight infant. Lee et al. (Lee, et al., 2009) interviewed 31 mothers
following discharge home who had breastfed their very low-birth-weight infants. The
researchers found five themes: wanting to compensate, maintaining motivation and
connectedness, needing ‘extra helping hands’, controlling emotions and matching
infant’s individual pace. Mothers in the study felt to blame themselves for premature
birth of their infant and providing breast milk was one way to compensate for the
harm they caused (Lee, et al., 2009). The importance of maintaining a connection
with their preterm infant through breast milk was also found by Sweet (Sweet,
2008b). Sweet (Sweet, 2008b) found that providing expressed breast milk (EBM)
connected the mother and infant, although it also placed unexpected pressure on the
mother to produce breast milk with her ability to provide breast milk potentially
affecting her sense of motherhood. Another paper based on the same
phenomenological study by Sweet (2008a) found that women expected breastfeeding
to be natural despite their infant’s preterm birth and therefore, preterm birth was not
an impediment to initiating breastfeeding. Further, breastfeeding was seen as the best
option not just because of its nutritional value, but to aid in building a close bond
between the mother and her infant as well. It was an interesting finding of this study
that despite the appearance of strong beliefs surrounding breastfeeding, all
participants were aware that breastfeeding was an option and spoke of the situation
where they would formula feed, if or when they cease breastfeeding.
27
2.5.3 Decisions Surrounding Methods to Provide Breast Milk
Of the women who decided that they will provide breast milk for their infants, some
decided that at-breast feeding is most beneficial and convenient for them and their
infant, while others chose bottle feeding their EBM. The literature supports that
breastfeeding is the preferred method due to the many advantages for both the
mother and infant (Buckley & Charles, 2006). However, some mothers decide to
feed their breast milk via a bottle as it allows their infant to receive breast milk and
has other perceived advantages for some mothers such as ability to know the volume
of milk consumed, ability of other family members to participate in feeding and
avoidance of embarrassment of breastfeeding in public. Other contributing factors to
bottle feeding may relate to the mothers themselves including inadequate milk supply
and experiencing feelings of lack of confidence or vulnerability, lack of
informational and emotional support as well as lack of commitment or desire to
breastfeeding prior to the birth of the infant (Buckley & Charles, 2006). Another
factor that could affect mothers’ decisions to at-breast or bottle feed EBM may be
their experience of at-breast feeding in the neonatal care unit as well as the education
and support received from health-care professionals. However, there were no studies
found that specifically investigated women’s decision making with regard to bottle
feeding EBM or feeding at the breast during the hospitalisation of their preterm
infant.
In conclusion, while the views of mothers on when their infant may be ready to
commence breast or bottle feeding are essential, the decision to start feeding does
need to be made with the neonatal care provider. This is done to ensure the safety
and wellbeing of the preterm infant with the clinician needing to assess and make a
judgement that an infant is mature in both physiological and neurobehavioural terms
to engage in nutritive sucking behaviour. However, this does not mean that mothers
cannot be included into discussions surrounding their preterm infant’s readiness to
commence breast or bottle feeding.
28
2.6 Clinical Decision Making Surrounding Breast- or Bottle-Feeding
Initiation
Clinical decision making is a process in which a health-care professional chooses
between alternatives by making judgements about the care that they are going to
provide (Banning, 2008). In particular, it involves patient assessment, determining
accepting or rejecting the possible diagnoses and selecting best care strategies
(Ramezani Badr, Nasrabadi, Yekta, & Taleghani, 2009, p. 351). Although there is
much research into clinical decision making, there is little research into clinical
decision making in relation to preterm infant breast or bottle-feeding initiation, with
previous studies predominately in the neonatal field focusing on ethical or end of life
decision making (Monterosso et al., 2005; Spence, 1998, 2000).
Like many clinical decisions, the decision to commence nutritive suck feeding in
preterm infants is complex and requires assessment. However, decisions on when
preterm infants can commence nutritive feeding are often made by the bedside nurse,
frequently with no assessment tool or specific policy to guide them. A recent study of
feeding practices of special-care nurseries (SCNs) (n = 29) in Queensland, Australia
found no participating units had any formalised criteria to judge readiness to
commence breast or bottle feeds; instead used general guiding principles or ‘rule of
thumb’ with the most common criteria as indicators of feeding readiness being GA
and NNS ability (Pamela Dodrill, et al., 2008). These findings are consistent with a
study conducted in United States, which found that just over 50 per cent of nurseries
had no specific policy or guideline on when to commence suck feeds with nurses
predominantly using behavioural cues, GA and weight to determine readiness
(Burklow, et al., 2002; Kinneer & Beachy, 1994). These studies illustrate that
neonatal care providers rely heavily on clinical experience to underpin decision
making with regard to the initiation of nutritive sucking behaviour. It is hypothesised
that this reliance of clinical experience to make decisions has led to great variation in
the management of preterm infants’ commencement and transition to exclusive
breast or bottle feeding. It has been suggested in the literature that an instrument
would aid neonatal care providers’ decision making through giving a consistent and
29
systematic framework for assessment of an individual preterm infant’s feeding
readiness (McGrath & Braescu, 2004)
At the time of this literature review, a number of methods to assist staff in
determining feeding readiness in the preterm infant population were found including
a theoretical model (Pickler, 2005), clinical guidelines (McCain, 2003; Premji, Paes,
Jacobson, & Chessell, 2002; Premji, 2000), protocols (McCain, et al., 2001; Premji,
McNeil, & Scotland, 2004), and scales or instruments (McGrath, 2003; Thoyre,
Shaker, & Pridham, 2005 ). Chapter 3 will investigate the clinical utility of these
methods and why they were not used as a basis for this study.
2.7 Limitations of Current Research
Many of the studies that investigated the development of preterm infant feeding
behaviour were either based on small numbers or infants who had varying degrees of
prematurity and morbidity. Further, many studies did not provide important
demographic information relating to participants such as GA at birth, GA at time of
investigation or whether any specific morbidity was present. This may limit their
generalisability to particular subgroups of preterm infants. However, collectively,
these studies have brought to light a number of important issues that need to be
considered in future research and clinical practice. The limited knowledge of the
development of breastfeeding behaviour as well as breastfeeding efficiency in the
preterm infant population is of concern.
In particular, there is a paucity of research into the issues surrounding the
commencement of breastfeeding in the preterm infant population, including how
infants are currently introduced to breastfeeding. Of the few studies that have
investigated how neonatal care providers make decisions to initiate breastfeeding,
findings have focused on listing criteria. There are no studies that have aimed to gain
consensus of which possible criteria used by neonatal care providers would be the
most important to their decision to allow an infant to commence breastfeeding.
Further, these studies have focused on neonatal nurses and have not included other
health-care professionals’ views. Research into feeding has overwhelmingly focused
30
on the bottle-feeding population. There are currently no instruments available that
assesses readiness of preterm infants to engage in breastfeeding behaviour.
2.8 Research Directions and Significance
Neonatal care providers make decisions every day surrounding the introduction of
preterm infants to breast or bottle feeding. Despite the possible negative outcomes
surrounding mistiming of the initiation of feeding, there is little evidence to support
their decision making. Studies have found that factors such as behavioural state,
maturity and severity of illness have been associated with feeding readiness;
however, there is little research on what infant factors or characteristics can actually
predict feeding success with most studies being conducted primarily in the bottle-fed
population. Moreover, GA of the infant is often cited as a criterion to initiate
nutritive suck feeding, with conflicting evidence for its use in determining feeding
readiness. Research is urgently needed to identify and document the assessment
criteria used by neonatal care providers to make their clinical decisions with regard
to a preterm infants’ readiness to commence breastfeeding. Further, for these criteria
to be useful in assessment for clinical decision making, studies also need to provide
evidence that they can predict which infants will be safe and successful in attempting
breastfeed. A validated instrument that is specifically designed to assess an infant’s
readiness to engage in breastfeeding would fill this gap in knowledge and practice.
2.9 Conclusion
The literature shows that many preterm infants find it difficult to engage in breast or
bottle-feeding behaviour immediately after birth. There are many reasons identified
in the literature as to why preterm infants are hindered in engaging in feeding
behaviour. These problems include immature or dysfunctional sucking skills,
dysfunctional behavioural state control as well as immature SSB coordination.
Further, it has been shown that mature sucking patterns develop with GA, not PNA.
However, individual differences have been noted with some infants successfully
orally feeding before 32 weeks gestation and others not for several weeks later.
Therefore, GA of the preterm infant at birth is important, but cannot be used alone as
31
a criteria to decide when to initiate breast or bottle feeding. Other variables such as
gender and time spent on mechanical ventilation may also impact the development of
nutritive sucking. Further, the results of this literature review also suggest that a
method to decide feeding readiness is required in clinical practice as there are no
accepted criteria to determine that feeding readiness exists to guide neonatal care
providers’ clinical decisions. Many researchers and clinicians support the use of an
instrument to assess preterm infants’ feeding readiness. It is hypothesised that such
an instrument would aid neonatal care providers’ decision making as well as reduce
the physical and physiological risks to both infant and mother. In order for such an
instrument to be incorporated in clinical practice, evidence needs to be collected and
appraised to determine if these presumptions can be supported.
32
33
Chapter 3: Systematic Review
3.1 Introduction
As described in the previous chapter, there are existing methods available to assess
feeding readiness in the preterm infant population; however, the clinical utility of
these instruments are unknown. The purpose of this systematic review was to answer
a well-defined question using systematic and explicit methods in order to identify
and critically appraise all of the relevant research. This systematic review follows the
methods of the Cochrane Collaboration’s neonatal group and has been submitted for
publication in The Cochrane Library. An abstract, which contains a summary of the
systematic review, is followed by the full review.
3.2 Abstract
3.2.1 Background
One of the most challenging milestones for preterm infants is the acquisition of safe
and efficient feeding skills. The majority of healthy fullterm infants are born with
skills to coordinate SSB. However, this is not the case for preterm infants, who
develop these skills gradually as they transition from tube feeding to suck feeds. For
preterm infants, the ability to engage in oral feeding behaviour is dependent on many
factors. The complexity of factors influencing feeding readiness has led some
researchers to investigate the use of an individualised assessment of an infant’s
abilities. A limited number of instruments that aim to indicate an individual infant’s
readiness to commence either breast or bottle feeding have been developed.
3.2.2 Objectives
The objective was to determine and compare the effects of using a feeding readiness
instrument, using no instrument or using another instrument, on the outcomes of time
to establish full oral feeding and duration of hospitalisation.
34
3.2.3 Search Methods
We used the standard methods of the Cochrane Neonatal Review Group and did not
apply any language restriction.
3.2.4 Selection Criteria
Randomised and quasi-randomised trials comparing a formal instrument to assess
preterm infant’s readiness to commence suck feeds with either no instrument (usual
practice) or another feeding readiness instrument were used.
3.2.5 Data Collection and Analysis
The standard methods of the Cochrane Neonatal Review Group were used.
3.2.6 Results
No studies met the inclusion criteria.
3.2.7 Author’s Conclusions
There is currently no evidence to inform clinical practice with no studies meeting the
inclusion criteria for this review. Research is needed in this area to establish an
evidence base for the clinical utility of implementing the use of an instrument to
assess feeding readiness in the preterm infant population.
3.3 Plain Language Summary
Unlike infants born at term who are able to breast or bottle feed soon after birth,
preterm infants need time to learn to feed, which may take days or weeks after they
are born. Preterm infants are commenced breast or bottle feeding at a time when the
infant is deemed to be ready by health-care professionals looking after the infant.
The optimal timing of the introduction of suck feeds is unclear in both the literature
35
and practice. An individualised assessment specifically designed to assess an
individual infant’s readiness to commence breast or bottle feedings has been
suggested as the best way to promote consistency in identifying when it is safe for an
infant to commence suck feeding. A limited number of assessment tools to determine
readiness have been developed. However, no studies were found that evaluated the
benefit or risk of their use.
3.3.1 Instruments for Assessing Readiness to Commence Suck Feeds in Preterm
Infants: Effects on Time to Establish Full Oral Feeding
3.3.1.1 Background
Description of the Condition. One of the most challenging milestones for preterm
infants is the acquisition of safe and efficient feeding skills (Hill, 2002). The majority
of healthy fullterm infants are born with skills to coordinate SSB that allow safe oral
feeding (Lau, et al., 2003). However, this is not the case for preterm infants who
develop these skills gradually as they transition from tube feeding to suck feeds (P.
Dodrill, et al., 2008; Thoyre, 2003). This transition to full oral feeding is an
important competency to attain prior to discharge home (Pickler & Reyna, 2003).
Delays in discharge are often secondary to feeding difficulties leading to increased
financial costs (Simpson, et al., 2002). Strategies to avoid delays must be the focus of
care without compromising the safety of the infant (McGrath & Braescu, 2004).
Introducing suck feeds as soon as the neurologic development and physical condition
of the infant permits has been reported to have several advantages, including shorter
transition time to all suck feeds, greater maternal satisfaction and shorter hospital
stay (Pridham, et al., 1993; Simpson, et al., 2002). However, feeding infants who are
unable to safely commence feeding may lead to problems with respiration, growth
and nutritional status with infants being at higher risk of 1) aspiration pneumonia, 2)
readmission to NICU, 3) fatigue, 4) increased energy expenditure, 5) hypoxia, 6)
bradycardia and 7) deglutition apnoea (Breton & Steinwender, 2008; Hill, 2002).
Therefore, careful timing is vital to ensure that the commencement of feeding is
beneficial rather than detrimental to the health of the infant (McGrath & Braescu,
2004).
36
Factors influencing the preterm infant’s ability to feed efficiently include
neurobehavioural maturation, physiologic stability, control of tone, behavioural state
organisation and coordinated SSB (McGrath & Braescu, 2004). Successful
coordination of feeding is also dependent on the adequate development of structures
of the upper airway including the lips, palate, jaw, tongue, pharynx, larynx and
oesophagus (Hill, 2002).
Differences in the ability of infants to engage in feeding behaviour at a particular GA
have been shown through studies of preterm infant sucking (Hedberg Nygvist, et al.,
1999; Lemons, 2001), Although GA is a guide to expected maturity, disparities are
evident in the rates at which infants mature (Hedberg Nygvist, et al., 1999; Simpson,
et al., 2002). Further, a preterm infant’s feeding ability may not always be consistent
at each feed, while infants are transitioning from gavage feeds (McGrath & Braescu,
2004). Differences in the sucking patterns between breast and bottle feeding have
also been found and may significantly impact the infant’s ability to commence suck
feeds (Thoyre, 2003).
Studies examining current practices in neonatal nurseries have found that over 50 per
cent of nurseries have no specific policy or guideline on when to commence suck
feeds, with nurses predominantly using behavioural cues, gestation age and weight to
determine readiness (Kinneer & Beachy, 1994; Siddell & Froman, 1994).
Description of the Intervention. A preliminary search revealed three instruments
designed to aide neonatal care providers in determining preterm infants’ readiness to
commence feeding. The Preterm Infant Nipple Feeding Readiness Scale (PINFRS)
was a ten-item scale that scores variables such as GA, PCA, colour or activity, state
regulation, hunger cues and tone (McGrath, 2003). However, this instrument has
been renamed as the Feeding Readiness and Progression in Preterms Scale
(FRAPPS). The second instrument found, the Early Feeding Skill (EFS) assessment
tool not only aims to assess feeding readiness, but also feeding skill and how well
they recover after the feed (Thoyre, et al., 2005). The feeding readiness section of the
EFS consists of five items that assess an infant’s readiness to commence oral feeds
by observing an infant’s tone, energy level, state of arousal, and oxygen saturation
37
(Thoyre, et al., 2005). Lastly, Fujinaga, deOliveira Rodarte, Goncalves and Scochi
(2007) developed and tested an eighteen-item-preterm-infant-oral feeding readiness
instrument consisting of items in relation to corrected GA, behavioural state, global
posture and tone, gag reflex, tongue movement and cupping, jaw movements and
maintenance of alert state. Each item was scored from zero to two with a possible
maximum score of 36. All instruments were designed such that the infant being
assessed for feeding readiness could ‘pass’ or ‘fail’. These assessments aim to
determine whether breast or bottle feeding is attempted and may easily be repeated
prior to each feed, while feeding is being established.
How the Intervention Might Work. The use of a formal screening instrument that
encompasses individual infant’s behaviour and development has been suggested as a
way to improve the accuracy of determining when an infant is ready to commence
feeding (McGrath, 2003). It is thought that many premature infants may be ready to
breast or bottle feed; however, as this readiness is often not identified, they continue
to be fed via a tube for longer than necessary. Alternatively, some infants who are
slower at developing these skills may be introduced to breast or bottle feeding too
soon. It is hypothesised that by identifying their readiness, neonatal care providers
could ensure that infants have more successful feeding attempts, reduce the time
taken to achieve all suck feeds and reduce the possibility of adverse events. The use
of a formalised instrument could also standardise measurement of feeding readiness
and facilitate documentation of feeding attempts.
Why it is important to do this review. The possible benefits of a screening
instrument to assess feeding readiness must be weighed against the additional staff
time required and other costs and possible detrimental effects, such as introducing
oral feeds when infants are not ready and withholding of oral feeding. This review
addresses the balance of benefits and risks of screening instruments for
commencement of suck feeds in preterm infants to assist in establishing an evidence
base for clinical decision making.
38
3.3.1.2 Objectives
Primary Objectives. 1. To assess and compare the effects of using a formal feeding
readiness assessment instrument and no formal instrument in preterm infants deemed
ready to commence feeds based on general clinical grounds on the outcome of time
to establish full oral feedings and duration on hospitalisation.
2. To assess the effects of different formal feeding readiness assessment instruments
in preterm infants deemed ready to commence feeds based on general clinical
grounds on the outcome of time to establish full oral feedings and duration of
hospitalisation.
The secondary objective of this review is to explore possible differential effects of
applying a formal feeding readiness assessment instrument according to the
following subgroups:
1. GA at birth:
• extremely preterm (<28 weeks)
• moderately preterm (28 to 31 weeks)
• mildly preterm (32 to 37 weeks)
2. Chosen method of feeding
• breast or bottle feeding
3.3.1.3 Methods
Criteria for Considering Studies for This Review. Types of studies: Randomised,
quasi-randomised controlled trials (including cluster trials) in which an instrument is
compared with either no assessment instrument or alternate instrument. Crossover
trials were excluded.
Types of Participants. Studies which enrolled preterm infants (<37 weeks gestation)
after being deemed ready to commence either breast or bottle feeds by general
clinical grounds. Exclusion criteria included congenital malformations, syndromes or
severe neurological problems.
39
Types of Interventions.
1. Experimental group involved infants who have been deemed ready on general
clinical grounds, who are then assessed for readiness to commence oral feeding
through the use of an instrument prior to the initiation of the first breast or bottle
feed. The instrument must have included assessment of one or more of the following:
a) motor development and abilities including posture, movement, tone and
reflexes
b) behaviour state and cues, including state of arousal and presence of feeding
behaviour cues
c) physiological parameters
d) integrity of oral structures
2. Control group involved infants who were not assessed by any formal instrument as
feeding was commenced once readiness was determined on general clinical grounds.
3. Comparison group involved infants who were deemed ready on general clinical
grounds, who are then assessed for readiness to feed by an alternate feeding
assessment instrument.
General clinical grounds were defined as a clinical impression that may include GA,
medical stability or infant cues, but excludes the use of a formal assessment
instrument. Instruments must have undergone psychometric evaluation including
tests for criterion-related or construct validity. In groups where an instrument is used,
infants must pass prior to commencement of feeding. Physiological parameters may
include HR, respiration rate and oxygen saturation levels. Other physiological
parameters used by individual trials were acceptable provided these were defined in
the trial protocol.
Types of Outcome Measures.
Primary outcome measure:
1. time from randomisation to full oral feeding (days)
2. duration of hospitalisation (days from randomisation until the end of the trial)
40
Secondary outcome measures include:
1. time from randomisation to introduction of first feed (days)
2. age (PCA and days from birth) at establishment of full oral feeding
3. daily weight gain (g/day or g/kg/day) from time of randomisation until the
end of the trial
4. breastfeeding (partial or full) on hospital discharge (number of infants)
5. time from randomisation to regain birth weight (days)
6. parental satisfaction (validated assessment tool)
7. number of apnoea or bradycardia episodes that required intervention from the
caregiver (stimulation, oronasal suction, increase in delivery of oxygen,
assisted ventilation)
3.3.1.4 Search Methods for Identification of Studies
The standard search methods of the Cochrane Neonatal Review Group were used.
Electronic searches. This included electronic searches of the Cochrane Central
Register of Controlled Trials (The Cochrane Library, Issue 2, 2010), MEDLINE via
EBSCO (1966 to July 2010), EMBASE (1980 to July 2010), CINAHL via EBSCO
(1982 to July 2010), Web of Science via EBSCO (1980 to July 2010) and Health
Source (1980 to July 2010). The search strategy for each database is described in
Appendix B. Language restriction was not applied.
Searching other resources. The meta-Register of clinical trials
(http://www.controlled-trials.com/mrct/search.html) and the US National Institutes of
Health registry of clinical trials (http://clinicaltrials.gov/) web sites were searched for
completed or ongoing trials. The authors also searched cited references from
retrieved articles. A number of researchers, who had either previously published an
article on the topic of feeding readiness or were known to have completed
preliminary psychometric testing of an instrument measuring feeding readiness to
identify any other potential studies meeting the inclusion criteria were contacted.
41
3.3.1.5 Data Collection and Analysis
The standard methods of the Cochrane Neonatal Review Group were used.
Selection of studies. Two authors (LC, KW or AC) independently screened the title
and abstract of all studies identified by the above search strategy. Articles identified
as potentially relevant based on title and abstract were retrieved in a full-text format
and were then reassessed for selection. Those studies that did not fulfil the inclusion
criteria were excluded. The authors resolved any disagreements by discussion. For
studies which were written in a language other than English, the English title and
abstract provided by the journal was used to decide their relevance to this review.
Only one article was found that contained no English title or abstract (Frischknecht,
2005). This article was translated from German to English by a doctor, training to be
neonatal specialist, who was fluent in both English and German languages.
Data extraction and management. If eligible studies were found, two authors
would have independently extracted and entered the data into tables using RevMan 5
software.
Assessment of risk of bias in included studies. If eligible studies were found, it was
planned that these studies would be evaluated independently by two review authors
(LC, KW and AC) for methodological quality in accordance with the methods of the
Cochrane Neonatal Review Group.
Studies were to be assessed with regard to blinding of randomisation, intervention,
and outcome as well as completion of follow-up. The results of this assessment
would be added to the table ‘Characteristics of Included Studies’. It was also planned
that consideration would be given to the following methodological issues:
• Sequence generation: Was the allocation sequence adequately generated?
• Allocation concealment: Was the allocation adequately concealed?
• Blinding of participants, personnel and outcome assessors: Was knowledge of
the allocated intervention adequately prevented during the study? At study
entry? At the time of outcome assessment?
42
• Incomplete outcome data: Were incomplete outcome data adequately
addressed?
• Selective outcome reporting: Are reports of the study free of suggestion of
selective outcome reporting?
• Other sources of bias: Was the study apparently free from other problems that
could put it at a high risk of bias?
This data was to be reported in the risk of bias table.
For consistency one review author (LC) was to act as primary reviewer for all studies
to be assessed and the secondary reviewer was to be one of the members of the
review panel (AC or KW). It was planned that any differences between the review
authors would be resolved either by discussion or by consensus after negotiation with
the third review author (KW or AC).
Measures of treatment effect. Weighted mean difference (WMD) would have been
calculated for continuous data and relative risk (RR) or risk difference (RD) for
dichotomous data. For each treatment effect a 95 per cent confidence interval (CI)
was calculated.
Assessment of Heterogeneity. If there were studies to synthesise in meta-analysis,
heterogeneity would have been assessed through the visual inspection of forest plots
as well as calculating the degree of heterogeneity statistically using the I2 statistic. If
statistical heterogeneity was found, the review authors would have searched for an
explanation (inter-study variations, intra study variations, methodological error,
publication bias and control effect) and either remove the heterogeneous study or not
conduct the meta-analysis depending on the explanation for and the degree of
heterogeneity. An I2 statistic above 40 per cent would have been considered
moderate heterogeneity and a value over 75 per cent considered high.
43
3.3.1.6 Data Synthesis
We planned to use the standard methods of the Neonatal Review Group to synthesise
the data. If there were eligible studies to conduct meta-analysis, WMD with a 95 per
cent CI would be used for the continuous variables and the RR and RD with 95 per
cent confidence intervals for categorical variables. Number needed to treat and
number needed to harm also would have been calculated if appropriate. To conduct
the meta-analysis, it was planned to use a fixed effect model. If any cluster trials
were included into the review, the studies would have been analysed separately from
non-cluster trials using the inverse variance (IV) method in consultation with the
Cochrane Neonatal Review Group statistician. Data analysis was to be undertaken by
using RevMan 5 software. If there were studies not suitable for meta-analysis then
results of these studies would have been summarised either in narrative form or in
tables. Instruments were to be analysed using separate comparisons according to the
type of instrument.
Subgroup analysis and investigation of heterogeneity. Subgroup analysis was
planned for the following subgroups if data was available: GA at birth (extremely
preterm, <28 weeks; moderately preterm, 28–31 weeks and mildly preterm, 32–37
weeks) and chosen method of feeding (breast or bottle feeding). Post hoc subgroup
analysis would be performed to detect heterogeneous trials.
3.3.1.7 Results
No studies were included.
Description of studies. See: Characteristics of excluded studies table (Appendix C).
Although the initial search found 955 publications, the number to be reviewed was
reduced to 716 once duplicates were removed. Two review authors reviewed the title
and abstract of all 716 publications. Only 44 articles were retrieved in the full-text
format for further consideration. However, no studies were found that met the
inclusion criteria for this review.
44
Within the 44 excluded articles, nine articles were found not to be research, but a
review of the literature. These nine articles were retrieved in the full-text format to
search the reference lists to ensure no studies were missed during the electronic
searching of databases. Topics of the literature review articles included initiation and
transition to suck feeds (Breton & Steinwender, 2008; Fernandez Díaz &
Valdebenito, 2007; Frischknecht, 2005; Lau, 2007; Lemons & Lemons, 1996;
McGrath & Braescu, 2004; Ross & Browne, 2002; Thoyre, 2003) as well as
diagnostic tools used to determine feeding readiness (da Costa & Bos, 2008).
A number of methods to assist staff in determining feeding readiness in the preterm
infant population were found including a theoretical model (Pickler, 2005), clinical
guidelines (McCain, 2003; Premji, et al., 2002; Premji, 2000), protocols
(Drenckpohl, Dudas, Justice, McConnell, & Macwan, 2009; McCain, et al., 2001;
Premji, et al., 2004; Shaker & Woida, 2007), clinical pathway (Kirk, Alder, & King,
2007) and scales/instruments (Fujinaga, de Oliveira Rodarte, et al., 2007; Ludwig &
Waitzman, 2007; McGrath, 2003; Thoyre, et al., 2005).
Although there were two randomised trials (McCain, Gartside, & Greenberg, 2002;
McCain, et al., 2001) found in the search that evaluated the clinical utility of the
implementation of a feeding protocol, these studies did not compare the assessment
of feeding readiness with any other assessment, but rather compared no NNS with
the use of ten minutes of NNS prior to assessing behavioural state as an indicator of
feeding readiness. There were two studies (Drenckpohl, et al., 2009; Kirk, et al.,
2007) that used historical controls to evaluate their implementation into practice as
well. Other articles related to methods to determine feeding initiation or transition
were either a description of the method (Ludwig & Waitzman, 2007; McCain, 2003;
Pickler, 2005; Premji, et al., 2004; Premji, et al., 2002; Shaker & Woida, 2007;
Thoyre, et al., 2005) or psychometric testing of an instrument (Fujinaga, de Oliveira
Rodarte, et al., 2007; Fujinaga, Zamberlan, Rodarte, & Scochi, 2007; McGrath,
2003; Neiva, Leone, & Leone, 2008; Rossarolla, Menon, Scochi, & Fujinaga, 2009).
Psychometric testing of the instruments did not involve an experimental design, but
other non-experimental designs such as observational studies and expert panels.
45
A further three observational studies were found that described the psychometric
testing of instruments that either did not measure or measured indirectly the construct
of feeding readiness. The Dynamic-Early Feeding Scale (D-EFS) instrument is an
observational coding scheme to continuously code videotaped oral feeding (Thoyre,
2009). This instrument should not be confused with another instrument developed by
the authors, the EFS (Thoyre, et al., 2005) described in the background of this review
which contains a checklist of five questions to determine feeding readiness. The
other two observational studies used an existing instrument, the Neonatal Oral Motor
Assessment Scale (NOMAS) that measures infants’ nutritive sucking behaviours.
These studies investigated the NOMAS psychometric characteristics within a healthy
preterm population (Howe, Sheu, Hsieh, & Hsieh, 2007) and as an indicator of
feeding readiness (Church et al., 2006).
Other studies were found that contributed to the knowledge of feeding readiness and
progression but did not involve assessment of feeding readiness. Staff surveys were
used to document how staff decide to commence breast or bottle feeding (Kinneer &
Beachy, 1994; Siddell & Froman, 1994) as well as manage the transition period from
tube feeding to all breast or bottle feeds (Pamela Dodrill, et al., 2008). Current
management of feeding initiation and progression has also been investigated using
chart audits (P. Dodrill, et al., 2008; Flint, et al., 2007). Observational studies were
utilised to explore factors that may relate to feeding readiness (Bauer, Prade, Keske-
Soares, Haëffner, & Weinmann, 2008; Buhler & Limongi, 2004; Cagan, 1995;
McGrath, 2005; McGrath & Medoff-Cooper, 2002; Pickler, et al., 2005) as well as
interventions that may enhance preterm infants’ ability to engage in feeding
behaviour (White-Traut, et al., 2005; White-Traut, et al., 2002). The effects of
feeding experience, maturity and morbidity on feeding milestones (Pickler, Best, &
Crosson, 2009) as well as sucking patterns (Cunha, Barreiros, Gonçalves, &
Figueiredo, 2009) were also studied.
Risk of bias in included studies. No studies met the inclusion criteria.
Effects of interventions. No studies met the inclusion criteria.
46
3.3.1.8 Discussion
The absence of randomised or quasi-randomised studies evaluating the use of a
formalised instrument to assess a preterm infants’ readiness has resulted in this
systematic review being unable to determine the effects of using such an instrument
on time to establish full oral feeding or duration of hospitalisation.
The excluded studies of this review show there is an interest among researchers in
how to best approach the dilemma of when to commence breast or bottle feeds. This
review focused on validated instruments, but there were a number of other methods
found (e.g., care pathways, protocols and clinical guidelines) that could also
potentially aide clinicians in managing suck feeding initiation and progression. There
were a few studies that demonstrated that the application of a feeding protocol may
improve outcomes including time taken to all suck feeds (Drenckpohl, et al., 2009;
Kirk, et al., 2007; McCain, et al., 2001) and length of hospital stay (McCain, et al.,
2001). The benefit of using a formalised instrument over other methods, such as
clinical judgement, or using a criterion, such as GA, is that an instrument ensures that
a systematic and consistent method of assessing feeding readiness is utilised.
There were a number of instruments that specifically assessed feeding readiness;
however, the clinical utility of these instruments was not investigated in an
experimental study. The studies were observational with the focus of establishing the
validity and reliability of the tool (Fujinaga, de Oliveira Rodarte, et al., 2007;
Fujinaga, Zamberlan, et al., 2007; McGrath, 2003; Neiva, et al., 2008; Rossarolla, et
al., 2009). The lack of any experimental studies to establish the clinical utility of the
instruments may simply be that they are too newly developed to have undergone
such testing. The absence of randomised or quasi-randomised trials may also be a
reflection of the practical difficulties to ensure the comparison group is not exposed
to the intervention, particularly, in the situation when the use of an instrument is
compared to normal clinical practice with direct care givers collecting data.
47
3.3.1.9 Implications for Practice
There is no evidence to inform clinical practice with no studies meeting the inclusion
criteria for this review.
3.3.1.10 Implications for Research
Randomised or quasi-randomised trials are needed to evaluate the clinical utility of
using an instrument to assess feeding readiness in the preterm infant population.
Researchers need to also consider the use of a feeding readiness instrument in the
preterm infant breastfeeding population as the majority of observational studies
investigating feeding readiness and progression are predominately focused on bottle
feeding.
3.3.1.10 Implications of Results on Thesis
The results of this review demonstrates that there has been a surge in interest over the
past few years in how best to determine when preterm infants are ready to engage in
feeding behaviour. At the time of confirmation in early 2007, only two instruments
designed to aide neonatal care providers in determining preterm infants’ readiness to
commence feeding existed in the literature. The FRAPPS was formerly known as the
PINFRS (McGrath, 2003) and the EFS assessment tool (Thoyre, et al., 2005). The
instruments themselves are not published in the literature with only the instrument
characteristics being described. The authors of the two instruments were approached
with regard to the availability of the tools for use. Both authors report that their
instruments continue to undergo validity and reliability testing and are not ready for
publication.
The other guidelines and protocols available in early 2007 were generally focused on
feeding in the preterm population and not specifically related to the concept of
feeding readiness. There was one feeding readiness model by Picker (Pickler, 2004);
however, this model was specifically developed for bottle feeding. Therefore, the
overall aim of the study was to develop and test an instrument to assess readiness to
commence breastfeeding. More importantly, the study sought to address the
48
identified limitations within the literature chapter and the systematic review with
respect to the assessment of readiness to initiate breastfeeding in the preterm infant
population.
49
Chapter 4: Development of the Preterm Infant Readiness
Scale
4.1 Introduction
A mixed method methodology and design was used to explore preterm infant feeding
readiness, an area of practice in which little is known. The chapter starts with an
overview of the debate between quantitative and qualitative inquires that underpin
the movement towards combining the two research paradigms. An overview of how
the mixed-methods design was incorporated in the study is then presented. Finally,
each of the first three phases of the study, which are focused on the development of
the instrument are presented including specific description of methods used, ethical
clearance and findings. The methods and results of the final phase involving the
testing of the instrument are described in Chapter 5.
4.2 Overview
Both quantitative and qualitative research methods have been used in health research
with researchers traditionally choosing one or the other as the basis on their study.
The separation between the two methods came to a head in the 1980’s and 1990’s
with the debate about which is the optimal method for research being referred to in
the literature as the ‘paradigm war’ (Tashakkori & Teddlie, 2008). During this time,
purists from both quantitative and qualitative camps focused on the differences
between the two methods rather than on the similarities and therefore, viewed their
paradigm as the appropriate model for research and implicitly or explicitly advocated
the ‘incompatibility thesis’ (Tashakkori & Teddlie, 2008; Teddlie & Tashakkori,
2003). The incompatibility thesis stated that paradigms that underpin the methods of
quantitative and qualitative research such as positivism and constructivism are
incompatible, thus, associated methods cannot and should not be mixed. Scholars
have criticised the incompatibility thesis, with many stating that the differences
between the two are overdrawn and that there are strengths and weakness of both the
positivist and the constructivist traditions (Bergman, 2008; Johnson, Onwuegbuzie,
& Turner, 2007; Tashakkori & Teddlie, 2008). The incompatibility thesis was
countered by Howe (1988), who posited a different paradigm: pragmatism.
50
Pragmatism underpins the mixed method design and presents a very practical applied
research philosophy with the focus being not the methods used but on the research
problem and the use of all approaches available to understand it (Creswell, 2009;
Teddlie & Tashakkori, 2003).
Although mixed-methods research is an increasingly popular paradigm it is still an
emerging research approach with no singular definition accepted (Johnson, et al.,
2007). For the purpose of this study, mixed method research was defined as a type of
research in which a researcher or team of researchers combines elements of
qualitative and quantitative research approaches (e.g., use of qualitative and
quantitative viewpoints, data collection, analysis, inference techniques) for the broad
purpose of breadth and depth of understanding and corroboration (Johnson, et al.,
2007).
The aims and benefits of mixed-methods research are simple with the method taking
the best of qualitative and quantitative approaches and combining them into a single
study so that a research problem can be solved better (Bergman, 2008). The mixed-
methods design can be particularly helpful in researching complex phenomena that
require data from different perspectives in which either a quantitative or qualitative
study would not provide on their own (Creswell, 2009). However, the mixed-
methods approach is not appropriate to be used in all research and should not be used
in all studies. To guide researchers on its appropriate use, a model of four rationales
was developed by Collins, Onwuegbuzie and Sutton (2006). One of these four
rationales is instrument fidelity in which mixed methods may be used to create new
instruments or assess the appropriateness and/or utility of existing instruments
(Collins, Onwuegbuzie, & Sutton, 2006). Although mixing of methods has been
informally used for many years to test the various types of validation and reliability
required in the development of an instrument, it is only recently that a number of
papers (Collins, et al., 2006; Onwuegbuzie, Bustamante, & Nelson, 2010) are
providing structured guidance on how a mixed-methods approach may be utilised for
the purpose of instrument fidelity. This has resulted in an increasing number of
researchers formally using mixed method as an approach to validate their instrument
(Gélinas, Fillion, & Puntillo, 2009; Owen-Smith, Sterk, McCarty, Hankerson-Dyson,
& DiClemente, 2010).
51
A four-phase mixed-method design was used to develop and test an instrument to
assess feeding readiness in the preterm infant population. This inclusive method was
chosen as the appropriate research approach as it allowed the study to meet the
project aims of developing and testing an instrument to assess feeding readiness
through providing greater content coverage and alternate levels of analysis that could
not be undertaken if a single method was used. There are a number of different
designs of mixed-methods research that could have been chosen; however, a
sequential, exploratory mixed-methods design was used. The distinguishing
attributes of the sequential, exploratory design is that the researcher first explores the
topic and then tests the findings of the earlier phases (Creswell, Piano Clark, &
Garrett, 2008; Tashakkori & Teddlie, 2003). In this study both elements of
quantitative and qualitative research were used with the first three phases exploring
decisions surrounding feeding readiness to develop the PTSR scale and the final
phase testing the validity and reliability of the instrument (refer to Table 1). In the
first phase of the study, quantitative methods were used to collect and analyse data to
explore the appropriateness of age as a criterion for deciding when preterm infants
are ready to commence nutritive sucking. The second phase of the study used
predominately qualitative methods to explore how neonatal care providers decide
feeding readiness; however, as the purpose of the phase was also to develop items for
an instrument, quantitative data analysis also occurred to establish how commonly
criteria were mentioned between the groups. Phase 3 was conducted to build upon
the findings of phase 2 and to explore which criteria were thought to be the most
important. This was achieved primarily by quantitative methods; however,
qualitative data collection and data analysis occurred to support the findings. The
final phase used quantitative methods to test and validate the instrument developed in
the first three phases.
52
Table 1: Priorities of quantitative and qualitative methods in each phase
Sequence order Strategy of
inquiry
Data collection Data analysis
Phase 1 Chart audit QUAN QUAN
Phase 2 Focus groups QUAL QUAL/quan
Phase 3 Delphi survey QUAN/qual QUAN/qual
Phase 4 Observational study QUAN QUAN
QUAN = priority is quantitative; QUAL = priority is qualitative; quan = quantitative is not the emphasis but included; qual = qualitative is not the emphasis but included
4.3 Phase 1: Chart Audit
4.3.1 Background
One of the important aspects of instrument development is to identify and describe
the behaviours that underlie the construct (Onwuegbuzie, et al., 2010). As described
in Chapter 2 of this document, there has been great variation in the management of
preterm infants’ commencement of breast and bottle feeding. Anecdotally, this great
variation is also occurring at the study hospital. However, data on when suck feeding
is being initiated or when infants reach feeding milestones are not routinely being
collected and placed into the study hospital’s database. The only data available was
the date they first received breast milk or formula and what the method of feeding
was at discharge. To discuss the development of the instrument, data would need to
be collected from infants’ medical charts.
4.3.2 Aim
To assist in the development of the PTSR scale, the overall aim of this phase was to
gain a greater understanding of infant factors that may influence the decision that a
preterm infant is ready to engage in suck feeding behaviour. The specific objectives
were to explore: 1) age as a criterion to determine feeding readiness and 2) factors
associated with prolonged time taken to reach first attempt at nutritive sucking.
53
4.3.3 Research Questions
On an average, at what age are preterm infants meeting their feeding milestones of
first attempt at suck feed, first successful suck feed and first exclusive suck feed?
What is the relationship between time taken from birth to first attempt at nutritive
sucking and specific demographic and disease characteristics?
4.3.4 Design
Retrospective, descriptive design was used in which a chart audit was conducted to
obtain data with regard to feeding practices in the study hospital.
4.3.5 Sample
The Mater Mothers Hospital (MMH) is a large, tertiary referral hospital in Brisbane,
Australia and at the time of data collection incorporated a 60-bed nursery: 20-bed
intensive-care nursery (ICN) and a 40-bed special-care nursery (SCN). The records
of all preterm infants born ≤34 weeks and discharged home from the MMH during
the period of January 2005 to May 2006 were included in the study. As the study
hospital cares for infants from all over Queensland and northern New South Wales,
only those discharged from the MMH were included. This was performed to reduce
the number lost to follow-up as many infants are transferred back to the referring
hospital with no further information provided about their ongoing feeding
management. Infants in the sample were fed as per hospital policy at the time of the
study, where mothers were provided with breastfeeding support and encouraged to
express breast milk and breastfeed their infants. However, in some circumstances,
infants were fed via a bottle when their mothers were not present at feed times or
when their mothers were not able to breastfeed or did not wish to breastfeed.
54
4.3.6 Data Collection Tools
4.3.6.1 Neonatal Clinical Reporting System Data Form (Appendix D)
The Neonatal Clinical Reporting System (N-CRS) database is a comprehensive
database that provides detailed discharge summaries for doctors, parents and
community nurses and a data warehouse for researchers. Data on each individual
infant is entered into the database either by the Neonatal Audit Officer or the clinical
staff caring for the infant. Standardised variable fields were exported from the N-
CRS database into an Excel file for all infants meeting the inclusion criteria.
Exported data included:
• infant characteristics such as GA at birth, birth weight and gender
• maternal characteristics (e.g., Type of delivery, reason for delivery)
• treatment or interventions required during admission (e.g., mechanical
ventilation and antibiotic use)
• feeding milestones and interventions (e.g., Method of feeding at discharge
and days requiring gavage feeding)
4.3.6.2 Chart Audit Data Form (Appendix E)
A standardised data collection form was developed by the researcher to collect
information from the infants’ charts.
4.3.7 Procedure
Both the Mater Health Services (MHS) and Queensland University of Technology
(QUT) Human Research Ethics Committee granted an exemption from full ethical
review and provided permission to conduct the study. Once permission was granted
to access the infants’ charts, the Neonatal Audit Officer was contacted. An Excel file
was provided containing demographic and feeding data of all infants meeting the
inclusion criteria. The data in this Excel file was exported from the MMH’s N-CRS
database. However, the previously collected data did not contain all variables of
interest or was not in a form useful for the purposes of this research. Therefore, to
collect data not found in the N-CRS database and to check the data found within, the
55
Medical Records Department of the hospital was also contacted to obtain the infants’
medical records. The medical record of each infant identified by the Neonatal Audit
Officer was reviewed and had a separate chart audit data form completed. The time
to reviewing each chart varied, with the average time spent being 45 minutes with
more complex medical records taking up to two hours. Data collected via the
standardised form was entered into an Excel file before being exported into an SPSS
data file. Initially 25 per cent of records were audited to check for errors; however, as
a number of inaccuracies were found between the data obtained from the N-CRS
database and the chart audit, all data was re-checked against the medical chart. A
further check of the data was completed by conducting frequency distributions to
check for invalid values and outliers.
4.3.8 Data Analysis
The main variables of interest were infants’ age at first attempt at suck feeding, age
at first successful suck feed and age of attainment of first exclusive suck feed. For
the purposes of this research the feeding milestones were defined as follows:
First attempt at suck feeding was defined as any documented attempt made by the
infant at breast- or bottle feeding. For those who were breastfed, infants did not have
to attach to the breast as licking at the breast was considered an attempt. An attempt
was also recorded if clinical staff at the time of the feed recorded ‘feeding attempt’ in
the infant’s chart.
First successful feed was defined as the first breast or bottle feed in which any
volume of milk was subtracted from their feed via a gavage tube. For breastfed
infants, the amount of additional milk via the tube depends on how well the clinical
staff feel the infant has breastfed. Subtracting the volume given via the tube helps to
assume that the infant has obtained some milk from the breast.
First exclusive suck feed was defined as the first breast or bottle feed in which no
additional milk was given to the infant via a gavage tube.
56
Time taken to reach the feeding milestones, the first successful feed, was calculated
in days by subtracting the date of first attempt at suck feeding from the date of first
successful feed. Similarly, the time in days to the first exclusive feed was calculated
by subtracting the date of the first feeding attempt from the date of the first exclusive
suck feed.
Descriptive statistics of frequencies and percentages were used to summarise
categorical infant and maternal characteristics. Prior to undertaking statistical
analysis, continuous variables were checked for normality. As all continuous
variables were not found to be normally distributed, medians and ranges were used to
summarise the age at which infants reached feeding milestones and other continuous
data. To test differences between the three groups based on the degree of prematurity
(extremely, moderately or mildly) with regard to dichotomous sample characteristics,
Pearson’s chi-square test was used unless the assumption of a cell count less than
five was violated than the exact test was reported. Further, the Kruskal-Wallis test
was used to examine any difference between the three groups and continuous
variables. Relationships between the time taken to reach feeding milestones and
infant characteristics and other continuous variables (e.g., days on mechanical
ventilation or days receiving caffeine) were tested by Spearman’s correlation
coefficient. A Cox regression model was used to explore the differences between
gestational groups in the time taken to reach first feeding attempt. Further, an
additional Cox regression model was used to determine the influence of infant and
maternal factors on the time taken to attain the feeding milestone of first attempt at
suck feeding. The level of significance was set at 0.05 and data analysed using SPSS
version 18.0.
4.3.9 Ethical Considerations
4.3.9.1 Consent and Information to Participants
The chart audit involved the use of secondary data; therefore, consent was not
obtained and information was not provided to participants. However, permission to
use the data was granted from the study hospital’s executive as well as its HREC.
Permission was granted based on the researcher being a staff member already with
access to the information, no identifying information would be reported or published,
57
and the use of the data was to improve care. The study hospital deemed data
collection for the purposes of audits to be reasonably anticipated by the participants
and their families.
4.3.9.2 Risks to Participants
Phase 1 was considered to be ‘negligible risk’. The National Health and Medical
Research Council (NHMRC) defines negligible risk in human research as ‘no
foreseeable risk of harm or discomfort; and any foreseeable risk is no more than
inconvenience…’ (National Health and Medical Research Council, Australian
Research Council, & Austraian Vice-Chancellors' Committee, 2007).
The risks of the chart audit to individual infants were minimal as there was no
contact between the researcher and the participants. However, there was a small risk
that identifiable information may be viewed by those not authorised to do so. To
minimise this risk, only those who were directly involved in the research had access
to identifiable information. Data collectors were all employees of the study hospital
involved and would normally have access to this information. Data forms were kept
in a locked cabinet at MHS. Access to the Excel database set up to manage the data
was restricted.
4.3.9.3 Confidentiality
Any identifying information was only seen by the researcher and the research
assistants. When analysed and reported, all identifying information was removed
with each participant given a numerical code. Confidentiality of contact details or
documentation collected for the studies was maintained, and is yet stored in a locked
filing cabinet and on a restricted access computer drive. No individuals were
identified with only the hospital identified in this thesis and any publications.
4.3.9.4 Approvals
This phase of the study was reviewed by both the Mater and QUT HREC and was
deemed a negligible risk. As a result, both ethics chairs provided a letter of
exemption from ethical review.
58
4.3.10 Results
4.3.10.1 Sample Characteristics
A total of 338 infants were found to meet the inclusion criteria of being born at less
than or equal to 34 weeks PCA and discharged from the Mater Mothers’ Hospital
between January 2005 and May 2006.
The median PCA of infants at birth was 33 weeks, with ages ranging from 24 to 34
weeks PCA. The sample was predominately male (53.6 per cent). At birth, the
weight of the infants ranged from 463 to 3968 g with the median weight of 1800.5 g.
The condition of the infant at birth was measured by Apgar scores at one and five
minutes, with the median score of eight and nine, respectively. No infants aspirated
meconium liquor at birth.
During their admission to the neonatal unit, only 87 infants (25.7 per cent) required
mechanical ventilation and for those who required ventilation, the median number of
hours was 53.25 hours (7 to 1269 hours). A greater number of infants required
continuous positive airway pressure (CPAP) to support ventilation with 205 infants
(60.7 per cent) requiring from one to 3058 hours. Oxygen therapy was provided via
mask, cot, head box or nasal prongs to 148 infants (43.8 per cent) for an average of
four days (one to 163 days). Only sixteen (4.7 per cent) infants were discharged
home receiving oxygen therapy; however, 24 infants (7.1 per cent) were diagnosed
with chronic neonatal lung disease (CNLD). Despite a relatively small number of
infants having an infection confirmed (8.3 per cent), intravenous (IV) antibiotics
were also administered to 209 infants (61.8 per cent) when sepsis was suspected. The
use of caffeine to treat apnoea of prematurity was also a relatively common
intervention with 135 (39.9 per cent) infants receiving caffeine for an average of 31
days (one to 104 days). Total parenteral nutrition (TPN) was administered to 204
infants (60.4 per cent) with a median of eight days (three to 62 days). A total of 337
infants (one missing case) were fed via a nasogastric or orogastric tube from three to
165 days with the median length of tube feeding 24 days.
59
For 275 infants (81.4 per cent) breastfeeding was the feeding method at first attempt
with only 61 infants (18.0 per cent) first attempting bottle feeding. However, there
were two cases, where it was unclear as to whether the infant had initially attempted
breast or bottle feeding from the review of medical records. At hospital discharge,
just over half of infants (52.1 per cent) were exclusively breastfed with the next most
common method of feeding being infant formula via a bottle (24 per cent). There
were some infants who were partially breastfed in which in addition to the breast
milk they consumed from breastfeeding they were given either EBM or infant
formula via the bottle as well, with percentages of these infants being thirteen per
cent and 7.4 per cent, respectively. Only a small percentage of infants were given
EBM exclusively via the bottle (three per cent). When the different combinations of
methods utilised at discharge are considered, the total number of infants receiving
any breast milk was 255 infants (75.5 per cent). Only one infant (0.3 per cent)
received infant formula via a nasogastric tube at discharge (Table 2).
Table 2: Frequency of feeding methods at discharge
Method of feeding
Total sample (n = 338)
Frequency Percentage (%)
Exclusive BF 176 52.1
Infant formula via bottle 82 24.3
BF and EBM via bottle 44 13 BF and infant formula via bottle 25 7.4 EBM via bottle 10 3 Infant formula via bottle and n/g tube 1 0.3
BF = breastfeeding; EBM = expressed breast milk; n/g = nasogastric
It was anticipated that there would be significant differences in the sample
characteristics as a result of the degree of prematurity and therefore, data was also
analysed using three categories based on GA at birth: extremely preterm (24 to 28
weeks); moderately preterm (29 to 32 weeks) and mildly preterm (33 to 34 weeks).
Only fourteen per cent of infants were categorised as extremely preterm, whereas a
further 34 per cent were categorised as born moderately preterm with just over half
of the total sample categorised as being born mildly preterm (52 per cent).
Statistically significant differences were found across age category groups for all
infant characteristics except for gender (Pearson χ22 = 1.74, p = .433). The extremely
60
preterm group had the highest rate of confirmed infection with 42.6 per cent of
infants in this group having an infection confirmed by blood cultures in comparison
with both the moderately preterm and mildly preterm groups, who had 4.1 per cent
and 1.7 per cent of infants, respectively, with an infection confirmed. Further, the
extremely preterm group also the highest frequency of antibiotics given (Pearson χ22
= 19.72, p ≤ .001).
Across the three groups, the extremely preterm group had the lowest median birth
weight (H(2) = 200.86, p ≤ .001) and Apgar score at one minute (H(2) = 44.48, p ≤
.001). There was no difference in the median Apgar score at five minutes between
the three groups, although the minimum and maximum values were different.
However, there was a statistically significant difference in the distribution of scores
across the three groups (H(2) = 19.93, p ≤ .001) (Table 3).
Table 3: Continuous infant characteristics by GA at birth
Extremely Preterm
Moderately Preterm
Mildly Preterm
24–28 weeks PCA
29–32 weeks PCA
33–34 weeks PCA
(n = 47) (n = 115) (n = 176)
Characteristics Median Median Median Significance
(Min, Max) (Min, Max) (Min, Max)
Age at delivery In completed weeks
27 31 34 ** (24, 28) (29, 32) (33, 34)
Birth weight Grams 915 1543 2053 ** (4, 631, 434) (610, 2876) (1059, 3988)
Apgar Score at 1 minute 6 8 8 ** (0, 9) (2, 9) (1, 10)
Apgar Score at 5 minutes 9 9 9 0.655
(3, 10) (3, 10) (5, 10) PCA = post-conceptual age; min = minimum; max = maximum. Significance level p < .001**
As shown in Table 4, the percentage of infants requiring support was the greatest in
the extremely preterm groups; the only age category in which 100 per cent of infants
required treatment. There were both statistical and clinical significant differences
across the three age categories in the length of time that infants required for each
intervention (Table 4). Jonckheere’s test also revealed significant trends in data: as
infants’ PCA at birth increased, the length of time that the infant required each of the
intervention to support their physiological stability decreased. Significant trends
were found for all interventions, including mechanical ventilation (J = 827.00, z =
61
−2.61, p < .009), CPAP (J = 22,254.00, z = −9.80, p < .001), oxygen therapy (J =
1,749.50, z = −6.29, p < .001), caffeine (J = 283.50, z = −10.03, p < .001) and TPN
administration (J = 1,140.50, z = −6.58, p < .001).
Table 4: Interventions required during admission by GA at birth
Age categories Test statistic
Intervention
Extremely Moderately Mildly Kruskal-Wallis
Significance level Preterm Preterm Preterm
24–28 weeks PCA
29–32 weeks PCA
33–34 weeks PCA
N = 47 N = 115 N = 176
Mechanical Ventilation
Frequency (%)
40 (85.1%) 33 (28.7%) 14 (8.0 %)
Median hours (min, max)
84a 42.5 45 H(2) = 9.36, p ≤ .009**
(9, 1269) (7, 196) (8, 131)
CPAP Frequency (%)
47 (100%) 94 (81.7%) 64 (36.4%)
Median hours (min, max)
689.35a 75.25 24.5 H(2) = 92.55, p ≤ .001**
(11, 3058) (4, 1734) (11, 110)
Oxygen therapy
Frequency (%)
41 (87.2%) 61 (53%) 46 (26.1%)
Median days (min, max)
50 4.0b 2a H(2) = 39.86, p ≤ .001**
(1, 165) (1, 117) (1, 68)
Caffeine Frequency (%)
47 (100%) 72 (62.6%) 16 (9.1%)
Median days (min, max)
58 25.5 11 H(2) = 89.63, p ≤ .001**
(27, 104) (1, 65) (5, 20)
TPN Frequency (%)
47 (100%) 67 (58.3%) 20 (11.4%)
Median days (min, max)
13a 7.0 5 H(2) = 43.46, p ≤ .001**
(4, 60) (3, 32) (3, 62)
a = one missing case; b = two missing cases; CPAP =continuous positive airway pressure; TPN = Total Parenteral Nutrition; min = minimum; max = maximum; Significance level, p ≤ .001**
Two hundred and thirteen mothers had singleton pregnancies (77.2 per cent) with 57
mothers having twins (20.7 per cent) and a further six mothers having triplets (2.2
per cent), for a total of 276 mothers and 338 infants.
Singleton infants (n = 213) were most predominately delivered by caesarean section
(62 per cent) followed by spontaneous vaginal birth (31.0 per cent) with the least
number of infants (seven per cent) being delivered via breech or instrumental
delivery. Although instrumental or assisted vaginal deliveries were not common,
forceps deliveries (4.2 per cent) were the most common instrumental or assisted
62
delivery. When a multiple birth occurred, the method of delivery was sometimes
different for each infant; however, this was only the case for twin births (Table 5)
with all infants in each triplet set delivered by caesarean section. Differences
between the numbers of each type of delivery for twin pregnancies are also the result
of the death of one of the infants, as only infants discharged home were included in
this study.
Table 5: Type of delivery for twin births
Type of delivery
Twin 1 (n = 57) Twin 2 (n = 50)
Frequency (%) Frequency (%)
Caesarean Section 47 82.5 39 78.0
SVD 9 15.8 7 14.0
Breech/Instrumental 1 1.8 4 8.0
SVD = spontaneous vaginal delivery
The average age of mothers (n = 276) was 32 with ages ranging from 16 to 47 years
old (Table 6). The most common age group was between 28 and 33 years with 112
women (40.6 per cent). Most women were living with their partner at the time of the
birth (92.4 per cent) and were identified as Caucasian (84.1 per cent). For one third
(33.3 per cent) of women, this was their first pregnancy with a median number of
two pregnancies (one to twelve pregnancies). Further, for most women (81.5 per
cent) this pregnancy was their only premature delivery.
The most common reason for delivery (n = 276) was preterm labour with 89 mothers
(32.2 per cent) experiencing this complication (Table 7). This was followed by
preterm pre-labour rupture of membranes (PPROM) (n = 68, 24.6 per cent) and then
pregnancy-induced hypertension (PIH) (n = 44, 15.9 per cent). There were other
presenting obstetric problems; however, these were less common.
63
Table 6: Maternal characteristics n = 276
Characteristic Frequency %
Age Group
16–21 14 5.1
22–27 55 19.9
28–33 112 40.6
34–39 81 29.3
40–45 13 4.7
Over 46 1 0.4
Ethnic group
Caucasian 232 84.1
Aboriginal/Torres Strait Islander 6 2.2
Asian 22 8.0
Other 16 5.8
Marital status
Lives with the father of the infant 255 92.4
Lives apart from the father of the infant 20 7.2
Mother died 1 0.4
Number of previous pregnancies (including this pregnancy)
1 92 33.3
2 86 31.2
3 35 12.7
≥4 63 22.8
Number of previous premature deliveries (including this pregnancy)
1 225 81.5
2 40 14.5
3 9 3.3
≥4 2 0.7
Table 7: Presenting obstetric complications
Presenting obstetric problem Frequency Percentage (per cent)
Not stated 1 0.4
PPROM 68 24.6
Preterm labour 89 32.2
PIH (BP greater than 130/90) 44 15.9
APH/Abruption 24 8.7
Foetal distress leading to
intervention
9 3.3
IUGR 19 6.9
Decreased Doppler flow 1 4.0
Other 21 7.6
PIH = Pregnancy induced hypertension; APH = Antepartum haemorrhage; PPROM = Preterm pre-labour rupture of membranes; IUGR = Intrauterine growth restriction; BP = blood pressure
64
Age at reaching feeding milestones. There appears to be little variation in the
overall median age in which infants reached the feeding milestones when compared
to the significant clinical variation in the minimum and maximum ages at which
infants reached these milestones (Table 8). When tested by Friedman’s analysis of
variance (ANOVA), there was a statistically significant difference found across the
three time points (χ22 = 507.967, p < .001). As shown in Table 8, the earliest GA at
which any infant attempted a suck feed was 29.86 weeks PCA; however, this attempt
was noted as not meeting the successful feed milestone with the infant still requiring
a top-up tube feed. It was not until 31.86 weeks PCA that the first infant had its’ first
successful and exclusive feed. The maximum age for the first feeding attempt was
41.73 weeks.
Table 8: Median age (PCA) in reaching feeding milestones
Total Sample (n = 338)
Age at first attempt
at suck feed
Age at first successful suck
feed Age at first
exclusive suck feed
(weeks) (weeks) (weeks) Median
34.00a 34.71 35.00 Minimum
29.86 31.86 31.86 Maximum
41.73 41.43 41.57 a = three missing cases
The median age to reach feeding milestones by GA group is presented below in
Table 9. Interestingly, the moderately preterm group has the lowest median age at
first attempt and first successful feed with both the extremely preterm and mildly
preterm having higher median post-conceptual age at these milestones.
65
Table 9: Median age at feeding milestones by GA group
Extremely Preterm Moderately Preterm
Mildly Preterm
24–28 weeks PCA 29–32 weeks PCA
33–34 weeks PCA
(n = 47) (n = 115) (n = 176)
Characteristics Median Median Median
(Min, Max) (Min, Max) (Min, Max)
Age at first attempt Weeks 33.71 33.28 34.14 a
(29.86, 40.57) (30.00, 41.40) (33.00, 40.14)
Age at first successful feed
Weeks 35.57 34.57 34.71
(32.00, 40.57) (31.86, 41.43) (33.00, 40.14)
Age at first exclusive Weeks 34.71 34.86 35
(32.00, 40.71) (31.86, 41.57) (33.00, 40.14)
a = three missing cases
Kruskal-Wallis tests were performed which revealed that there were statistically
significant different ages for attainment of each of the feeding milestones across the
three groups. For the milestone of the first attempt, the moderately preterm had the
lowest median value when the compared to the two other groups (H(2) = 43.78, p ≤
.001). Further, at the feeding milestone of the first successful feed, the moderately
preterm group also had the lowest median age at the first successful feed (H(2) =
12.68, p = .002). However, for the milestone of the first exclusive suck feed, the
lowest median value was in the extremely preterm group with the median age
increasing as the age of the groups increased (H(2) = 16.09, p ≤ .001).
Mann-Whitney tests were used to follow-up this finding. A Bonferroni correction
was applied and therefore, 0.01 was considered the level of significance. It appeared
that at age of the first feeding attempt there was no statistical difference between the
extremely preterm group when compared to the moderately preterm (U = 2138.00, z
= −2.86, p = .04) or when the extremely preterm group was compared to the mildly
preterm group (U = 3816.50, z = −.65, p = .519). However, when the moderately
preterm group was compared with the mildly preterm group the age of the first
attempt was significantly lower in the moderately preterm group (U = 5081.50, z =
−7.04, p ≤ .001).
66
When examining the age at the first successful feeding attempt, there was no
statistical difference between the moderately preterm group when compared with the
mildly preterm group (U = 8546.00, z = −2.246, p = .025). However, the extremely
preterm group had a significantly higher age at first successful feed than the mildly
preterm group (U = 3098.00, z = −2.65, p = .008). Further, when the extremely
preterm group was compared to the moderately preterm group (U = 1916.50, z =
−2.90, p = .004) the age of first successful feed was also significantly higher in the
extremely preterm group. Therefore, both the moderately and mildly preterm groups
had a significantly lower age than the extremely preterm group.
Interestingly, for the age of the feeding milestone of the first exclusive suck feed, all
group comparisons were significantly different. The extremely preterm group had a
significantly lower age at first exclusive suck feed when compared to the moderately
preterm group (U = 205.00, z = −9.22, p ≤ .0001). Further, the extremely preterm
group also had a significantly lower age at reaching the milestone of first exclusive
suck feed than the mildly preterm group (U = 10.50, z = −10.51, p ≤ .0001). For the
comparison between the moderately preterm group and the mildly preterm group, the
moderately preterm had a significantly lower age at the first exclusive suck feed (U =
2061.00, z = −11.49, p ≤ .0001). Thus, it can be concluded that for the feeding
milestone of the first exclusive suck feed, the extremely preterm group had the
lowest age at first exclusive suck feed, followed by the moderately preterm group
with the mildly preterm group having the highest age at first exclusive suck feed.
Relationship between time to attain feeding milestones and demographic disease
characteristics. Table 10 shows the median time taken from birth to reach the
feeding milestones of the first feeding attempt, first successful feed and first
exclusive suck feed. For each of the milestone there is great variation in the time
taken with infants achieving the milestone on the day of birth up to 107 days for time
taken to first exclusive suck feed.
67
Table 10: Median time taken to reach feeding milestones
Total Sample (n = 338)
Time to first
attempt at suck
feed
Time to first
successful suck
feed
Time first
exclusive suck
feed
Transition time
(days) (days) (days) (days)
Median 7a 14 16 7
Minimum 0 0 0 0
Maximum 96 103 107 45
a = three missing cases; Transition time = time from first attempt to first exclusive suck feed
The differences in time taken from birth to reach the feeding milestones across the
GA birth groups were also examined. Initially from Table 11, it appears that the
extremely preterm group had the longest time taken to reach all three feeding
milestones as well as the longest transition time from the first attempt to exclusive
suck feed. A univariate, Cox regression model was used to further explore the
independent effect of age at birth on time taken to reach the first breastfeeding
attempt.
Table 11: Median time taken to reach feeding milestones by GA group
Extremely Preterm
24–28 weeks
PCA
Moderately Preterm
29–32 weeks
PCA
Mildly Preterm
33–34 weeks PCA
(n = 47) (n = 115) (n = 176)
Characteristics Median Median Median
(Min, Max) (Min, Max) (Min, Max)
Time to first attempt at suck feed
Days 54 13 4.97
(20, 98) (2, 87) (0, 50)
Time to first successful suck feed
Days 64 26 7.5
(35, 103) (2, 87) (0, 50)
Time first exclusive suck feed
Days 69 27 10
(35, 107) (2, 88) (0, 50)
Transition time from 1st attempt to exclusive suck feed
Days 9 9 5
(0, 45) (0, 32) (0, 21)
Min = Minimum; max = maximum; PCA = post-conceptual age
68
Table 12 and Figure 2 show that both the extremely preterm group and the
moderately preterm group were more likely to have prolonged time to first feeding
attempt when compared to the mildly preterm group.
Table 12: Effect of GA at birth on time taken to reach first breastfeeding
attempt
GA at birth Number of subjectsa
Hazard ratiob 95% CI Significance
level
Extremely preterm 24–28 weeks
47 0.54 .035–.082 <.0001
Moderately preterm 29–32 weeks
115 0.26 .201–.337 <.0001
Mildly preterm 33–34 weeks
173 1 referent
Figure 2: Time taken to first attempt at breastfeeding by GA at birth
69
To test the effect of PCA of the infant at birth as well as other demographic and
disease characteristics on the time taken to achieve the feeding milestone of time to
first attempt, a Cox regression was performed. For the model, the continuous
variables of GA at birth and birth weight were used as well as the categorical
variables of mechanical ventilation, CPAP, oxygen therapy and caffeine. All
variables have been mutually adjusted for all other variables in the model. Other
variables such as gender, infection, Apgar scores and whether the mother had a
previous preterm infant were tested at the bivariate level and were not found to be
significant so were not included in the model.
This model (n = 335) examines the effect of variables on the time taken to the first
feeding attempt. The overall model was significant (χ26 = 388.02, p ≤ .0001).
Table 13: Model of variables associated with time taken to first feeding attempt
(days)
Number of subjectsa
Hazard ratiob 95% CI Significance level
GA at birth Weeks 335 1.269 1.166–1.381 <.001
Birth weight Kg 335 2.638 2.032–3.426 <.001
Mechanical No 248 2.227 1.594–3.111 <.001
Ventilation Yes 87 1 referent
CPAP No 131 2.172 1.629–2.897 <.001
Yes 204 1 referent
Oxygen Therapy
No 188 1.504 1.127–2.009 0.006
Yes 147 1 referent
Caffeine No 200 1.543 1.128–2.111 0.007
Yes 135 1 referent
GA = gestational age (PCA); CI = confidence interval; a = three missing cases; b = hazard ratios mutually adjusted for all other variables in the table.
Table 13 shows that the variable GA at birth (PCA), birth weight, mechanical
ventilation, CPAP, oxygen therapy and caffeine all have a statistically significant
effect on the dependent variable of time taken in days to the first attempt at suck
feeding.
70
Infants born at younger PCA were at the most risk of prolonged time to first attempt
with each week older at birth, the odds of attempting to breastfeed was also increased
(95 per cent CI = 1.166–1.381, p ≤ .0001). Similarly, lower birth weight infants were
also found to have a prolonged time to first attempt with each additional kilogram;
infants had a 2.6-fold increase in the odds that breastfeeding would occur (Hazard
ratio = 2.638, 95 per cent CI = 2.032–1.381, p ≤ .0001).
The strongest intervention predictor for prolonged time to first feeding attempt was
mechanical ventilation, in which infants who were not mechanically ventilated had a
2.2-fold increased odds of the first attempt at breastfeeding occurring more quickly
when compared to infants who were ventilated (Hazard Ratio = 2.23, 95 per cent CI
= 1.594–3.111, p = .006). This was also true for the other interventions in which
those infants who received the intervention had a prolonged time to first attempt
(Table 13).
4.3.10.2 Summary
This phase of the study has retrospectively examined when infants are meeting breast
and bottle feeding milestones. The overall median age at attaining the feeding
milestone of the first attempt was 34 weeks (29.86–41.73 weeks), which is consistent
with the several reports of the acceptable age to initiate breast or bottle feeding.
There were statistically and clinically significant differences shown between the ages
at which infants reached the feeding milestones of the first attempt, first successful
and first exclusive suck feeds. With regard to time taken to reach feeding milestones,
it appears that the lower the GA at birth and birth weight, the more prolonged the
time taken to first feeding attempt. This is also true for the interventions in the model
in which if an infant had the intervention, they were at greater risk of prolonged time
to feeding. Although this study was beneficial to discuss the development of the
PTSR scale by documenting that there are differences in the age of infants to reach
the feeding milestones as well the existence of variables that impact on time taken to
breast or bottle feed, it does not explore how the introduction and transition to breast
or bottle feeding is managed by clinical staff. This next study will explore how
71
neonatal care providers decide when an individual infant is ready to attain the
feeding milestones.
4.4 Phase 2: Focus Groups
4.4.1 Background
Phase 1 of this study identified infant and maternal factors that may influence the
achievement of feeding milestones. However, there may be other factors or criteria
used by clinical staff to determine feeding readiness that were not revealed in the
chart audit. In addition, although the literature has described clinicians using the age
of the infant as the main criteria for feeding readiness, recent research suggests that
there should be less emphasis on age as a criteria (Pickler, et al., 2009). It is
unknown whether this research finding is reflected in clinical practice.
4.4.2 Aims
The main aim of this study was to explore how clinical staff determine if a preterm
infant is ready to engage in breastfeeding so that items for possible inclusion in a
breastfeeding readiness instrument could be generated. Specific objectives of the
study were to: 1) explore the use of age as a criterion to initiate breastfeeding by
neonatal care providers and 2) explore what criteria neonatal care providers use to
determine when preterm infants are ready to initiate breastfeeding.
4.4.3 Research Questions
1. Does age or age range remain an important criteria to initiate breastfeeding in
the preterm infant population?
2. What criteria do neonatal care providers use to decide when to initiate
breastfeeding in the preterm infant population?
3. What criteria do neonatal care providers use to decide when not to initiate
breastfeeding in the preterm infant population?
72
4.4.4 Design
Focus groups were used to determine items for inclusion in the instrument. This
method has been used in nursing research for several purposes including instrument
development (McLafferty, 2004; Willgerodt, 2003). The purpose of this design is to
obtain the participants’ perceptions and personal views on a specific topic (Willis,
Green, Daly, Williamson, & Bandyopadhyay, 2009). A benefit of conducting a focus
group over other alternative data collection methods such as one-to-one interviews is
that it allows participants to react and build upon the responses of other participants
(Plummer-D'Amato, 2008).
4.4.5 Sample
Three focus groups involving nursing staff and other interested health-care
professionals from the MMH’s SCN (two groups) and ICN (one group) were formed.
This neonatal unit is one of two tertiary referral units in south-east Queensland and
the second largest in Australia.
Optimal sample size ranges from four to twenty in the literature, as any fewer may
lead to inadequate discussion (McLafferty, 2004). In this study, ten participants were
recruited to each group to ensure a sufficient group size if a couple of participants did
not attend the focus group discussion. Purposeful sampling was utilised so that staff
with expertise in the initiation of breastfeeding were targeted and invited to
participate. Inclusion criteria included: 1) volunteer to be involved in the focus
group; 2) employed as a midwife or nurse, speech pathologist, physiotherapist or
neonatologist in the SCN or ICN at the MMH or 3) have had experience in the
initiation of breastfeeding in preterm infants.
4.4.6 Data Collection Tools
A semi-structured focus interview guide was developed to elicit participants’ ideas
including the following questions:
73
1. Is there a particular age or age range when you allow mothers to attempt
breastfeeding?
2. If so, why have you decided on this age?
3. How do you decide when an infant is ready to attempt a breastfeed?
4. How do you know when an infant is not ready to attempt a breastfeed?
4.4.7 Procedure
Both the MHS and QUT HREC gave approval to conduct this study. Participants
were recruited from the SCN and ICN at the MMH. An e-mail with the participant
information letter attached (Appendix F) was sent out to all staff working in the
study hospital’s neonatal unit. In addition, notices were placed on the electronic, staff
notice board to encourage staff to read their e-mail and consider being involved.
Focus groups were held during working hours but at different times to allow greater
flexibility for potential participants to attend. Further, it was organised with the
Nursing Director and Nursing Unit Managers for nursing staff to be relieved for the
period of attendance at the focus group. The focus groups’ discussions were held in a
small conference room which was a non-threatening, comfortable and easily
accessible setting with minimal distractions (Jamieson & Mosel-Williams, 2003;
Willis, et al., 2009). Refreshments were provided to participants prior to the
commencement of the focus groups to create a relaxed setting. The researcher acted
as a moderator who guided the discussion and clarified responses. Before
commencement of the discussion, participants were welcomed by the moderator and
told that: 1) there is no right or wrong answer with all answers being valid; 2) the
discussion will be audiotaped and 3) methods will be used to maintain their
confidentially (Morrison-Beedy, Cote-Arsenault, & Fischbeck Feinstein, 2001). To
provoke discussion, a short video was played to the participants of an infant prior to
a feed. Authors have found that the use of stimuli like a poster or other material can
encourage and enrich discussion (Jamieson & Mosel-Williams, 2003). The focus
groups were expected to last one hour; however, focus groups may last from one
hour to three hours depending on the scope of the problem discussed and the number
of participants (Boswell & Cannon, 2007). All discussions were audiotaped and
74
transcribed verbatim. All transcripts were audited to ensure quality of transcription
by listening to the audiotapes when reading the transcripts.
4.4.8 Data Analysis
There is some debate on how data collected from focus groups should be analysed
with various techniques being described (Bender & Ewbank, 2004; Jamieson &
Mosel-Williams, 2003). The depth of data analysis and whether data should be
analysed at a group or individual level depends on the purpose of the focus groups
(Jamieson & Mosel-Williams, 2003). As the purpose of this phase was to develop
items for an instrument, analysis occurred at group level and only content analysis
was performed (Creswell, 2009). Content analysis involved reading the transcribed
interviews to identify key quotes, form codes and grouping the codes into themes
(Creswell, et al., 2008; Morrison-Beedy, et al., 2001). Once analysis was completed
the key quotes, codes and themes were used in the construction of a preliminary
version of the instrument to be used in the next phase of this study (Creswell, et al.,
2008). To determine the agreement for the discussed criteria between the three
groups of participants, the number of groups mentioning the item was also reported
(Bender & Ewbank, 2004).
4.4.9 Ethical Considerations
4.4.9.1 Consent and Information to Participants
Participants in the focus groups received an information sheet prior to participation
with sufficient time to consider their participation (Appendix F). The information in
the sheet was also repeated verbally to participants prior to commencement. To allow
for open disclosure of opinions during the focus group sessions, written consent
forms were not signed, instead attendance at the focus groups were deemed as
consent to participate.
An HREC approved information letter was provided to the participants giving a
description of the study and its purpose. The letter described what their possible
involvement would entail and any risk and benefits to participating in the study.
Potential participants were also informed of what steps would be taken to protect
75
their privacy and maintain confidentiality. Each study’s information letter stated that
participation was voluntary and that their participation would in no way impact any
current or future relationship with either the study hospital or the university. Specific
information was given to the participants in the focus group including: 1) they were
being audiotaped; 2) they could not participate unless they agreed to being
audiotaped and 3) they could withdraw prior to and during the focus group; however,
information already collected would be used as no identifying information was
collected from participants.
4.4.9.2 Risks to Participants
Phase 2 was considered to be ‘negligible risk’. For the focus groups, the risk or
burden associated with this project was not expected to be greater than for their
normal day-to-day work. However, there was a possibility of a small risk of a breach
of their privacy and confidentiality. This was minimised for the focus groups
through: 1) audio recordings not being kept after accurate transcription has been
verified; 2) individuals not identified within the transcript; 3) transcription by a
person who does not have an association with the MMH staff involved; 4) the list of
staff who volunteered to be involved will be stored separately from the transcripts on
a restricted access computer drive. The burden for the staff was minimal as
individuals volunteered and could be conducted during work time. No identifying
information was used in any report or publication.
4.4.9.3 Confidentiality
Any identifying information was only seen by the researcher and the research
assistants. When analysed and reported all identifying information was removed with
each participant given a numerical code. Confidentiality of contact details or
documentation collected for the studies was maintained, and is still stored, in a
locked filing cabinet and a restricted access computer drive. No individuals were
identified with only the hospital identified in this thesis and any publications.
76
4.4.9.4 Other Approvals
This phase of the study was reviewed by both the Mater and QUT HREC and was
deemed a negligible risk. As a result, both ethics chairs provided a letter of
exemption from ethical review. Approvals to conduct the research projects were also
granted from the relevant Nursing Unit Managers and Nursing Director of
Neonatology to conduct data collection at the MMH.
4.4.10 Results
4.4.10.1 Response Rate and Sample Characteristics
It was predicted that between fifteen and thirty health-care professionals from the
MMH would attend the focus group sessions. However, this did not occur with a
total of nine staff attending one of the three focus group sessions. Out of the nine
participating health-care professionals, only two attended the first focus group with
three and then four participants attending the second and third groups, respectively.
Although the sample size was smaller than anticipated, the three groups represented a
cross section of health-care disciplines that are involved in assisting the mother to
breastfeed her preterm infant including: a physiotherapist, an occupational therapist,
a speech pathologist, two lactation consultants and four neonatal nurses. All
participants in each group were female and had at least five years of professional
experience in the initiating of breastfeeding within the preterm infant population.
Each group had some participants who worked across both SCNs and ICNs, but there
were two participants in group 2 who worked predominantly in ICN with one
participant in group 3 in SCN. The length of the focus groups ranged from
approximately 36 minutes to 46 minutes with the total amount of time being just over
two hours.
4.4.10.2 Age as a Criterion to Determine Breastfeeding Readiness
The PCA of the infant at the time of deciding when to initiate breastfeeding appears
to be an important criterion with all groups stating that the age of the infant was a
consideration in making their decision. However, no specific age or age range was
found to be mentioned across the three groups (Table 14).
77
Table 14: Age criteria for preterm infants to attempt breastfeeding mentioned
by focus group
Minimum Age criteria Number of focus groups mentioning item
(n = 3)
No Minimum 1
30–32 weeks PCA 1
32 weeks PCA 1
32–34 weeks PCA 2
34 weeks PCA or greater 2
PCA = post-conceptual age
A reflection of why no specific age or age range was accepted by all groups may be
the result of the criterion of age not being strictly adhered to for all infants, with age
being seen as a general rule when infants may be commenced. Each group
acknowledged that although they have a minimum age or age range in mind when
considering the commencement of breastfeeding, they would also take into account
the infant and may commence early or later depending on individual differences.
Participants during the focus groups often referred to the current gestation age of the
infant in terms of weeks, for example 34-weeker.
Participant from group 1: ‘I suppose there is in general terms an age range but that
doesn’t mean to say that there aren’t babies outside that range or at the lower end of
that range…there are other criteria why I would let them breastfeed. I even can recall
having let a baby breastfeed at 29 weeks because it met other criteria …’
Participant from group 2: ‘I think it depends on the babies cues to some degree,
rather than an age say at 34 weeks we should do this, but I think having worked as an
LC [lactation consultant] some babies at 33 are quite ready to do it, so I think it is
hard to put an exact age, sometimes on that…’
Participant from group 3: ‘I personally probably wouldn’t put a 28, 29 below that
sort of weeker unless they were showing real signs of wanting to ….’.
78
Another explanation to observed differences may have been how they determined
their stated age or age range criteria. To further understand the differences in why
groups stated a particular age or age range, the groups were asked how they decided
on using their stated minimum age or age range criteria. Common reasons across all
three groups were their own clinical experience and reading journal articles (Table
15). Groups spoke of reading both literature reviews on preterm infant feeding as
well as research studies in which sucking patterns and the presence of protective
reflexes such as cough and gag reflexes are examined at differing GAs. Some of the
groups also discussed that they were taught that 34 weeks PCA was the minimum
age to attempt breastfeeding, which they state is often still the unit culture
particularly in the ICNs.
Table 15: Stated reason for minimum age criteria mentioned by focus groups
Reason for Minimum Age criteria Number of groups mentioning item
(n = 3)
Own clinical experience 3
Reading literature/research articles 3
Unit culture 2
What I was taught
1
Talking to lactation consultants 1
Participant from group 1: ‘But I think the research tells you that the responses may or
may not be there earlier, ….And absolutely, my experience as well’
Participant from group 2: ‘a bit of reading and a bit of seeing what happens in
practice to some degree…’
Participant from group 3: ‘and most of the literature even with breastfeeding will say
that babies have developed, you know well developed SSB coordination around that
34 weeks mark, but it also say that babies from other, from 30–32 weeks even will
do quite well at the breast occasionally’.
79
However, clinical experience with assisting mothers of preterm infants to breastfeed
was cited by two of the groups as one of the reasons why they would decide to
initiate feeding outside the stated age or age range. Clinical experience was also
reported to have the potential to change previous criteria for stated age, particularly,
unit culture and what they have been taught.
Participant from group 2: ‘yes, that’s what I was taught. But having worked in the
field and seen it, I’m probably a bit more flexible now.….’.
Participant from group 3: ‘so experience is definitely something that I see as a
change in your thinking about when babies might be ready and I think that’s just
because you see more babies and you know what to look for and things like that’.
Participant from group 3: ‘Follow the culture that’s been set out. But now that you’re
experienced and would just take that on yourself, wouldn’t you? And, you wouldn’t
have to worry about what they’re thinking’.
Interestingly, all groups also discussed that there was a maximum age before which
infants should be attempted at the breast. However, of the three groups, only group 1
specified a maximum age at which infants should attempt breastfeeding; they
mentioned either 35 weeks or 36 weeks PCA as the maximum age. All groups said
that if feeding was delayed too long, infants may experience more difficultly in
transitioning to breastfeeding.
Participant from group 1: ‘I think there comes a point in time where you have to
weigh up, like the amount of oxygen they may be on, is very high but you’ve reached
that point where they need the practice at sucking, proper sucking practice and
feeding practice, because if beyond that point, we all seen the babies that they don’t
remember how to suck or they don’t have that sucking coordination …’.
Participant from group 2: ‘And that is where it falls down sometime, because the
window of opportunity is limited and I find that if you miss the window, it makes it
that much harder, so it draws that time out’.
80
Participant from group 3: ‘But I think you certainly have ranges in your head, like if
you got a baby who is 36 weeks who hadn’t tried it, you would go wow, haven’t tried
it…in that way there’s almost a maximum age that we should have tried by, but I
don’t think there’s a minimum age where you should try’.
4.4.10.3 Criteria to Determine if Infant Is Ready to Breastfeed
The focus groups mentioned a number of different criteria apart from age that they
would use to decide when an infant is ready to attempt breastfeeding, many of which
all groups mentioned (Table 16).
Table 16: List of criteria to decide feeding readiness mentioned by all groups
Feeding readiness criteria
Number of groups mentioning item
(n = 3)
Respiration within normal limits at baseline 3
Heart rate within normal limits at baseline 3
Oxygen saturation levels within normal limits at baseline 3
No dramatic change in oxygen requirement during handling/cares 3
No dramatic change in baseline heart rate during handling/cares 3
No dramatic change in baseline respiratory rate during handling/cares 3
No dramatic change in baseline transcutaneous oxygen during handling/cares
3
Sucking regularly on dummy, fingers, feeding tube, etc. 3
Mature non-nutritive sucking pattern 3
Presence of rooting reflex 3
Moves smoothly between behavioural (arousal) states 3
Infant alert just prior to feed time 3
Infant showing interest at the breast (licking, moving to nipple, etc.) during kangaroo care
3
Although the above criteria (Table 16) were mentioned by all groups, they did not
seem to have a single criterion but rather a group of criteria in which an infant had to
81
meet a number of criteria before they would consider that infant to be ready to
attempt breastfeed.
Participant from group 1: ‘I think you would be looking at, well behaviourally for
sure, baby is out for a cuddle and baby is mouthing and looking and wanting, baby is
very competent with the non-nutritive sucking …very able to non-nutritively suck
quite happily, keep that little dummy in and very interested in sucking, while the
baby is having a gavage feed …but also there is the respiratory side of it, the work of
breathing and the amount of oxygen the baby’s on all those’.
Participant from group 2: ‘that you’ll find they are sucking or they’ll start—you put a
dummy near them and they will start rooting around, …but again I think you would
be looking they are quite—that there are no issues with breathing or something like
that…. I take the cues from the baby’.
Participant from group 3: ‘any baby that is making any feeding cues at all, so they
need to be awake and alert, and if they look like they are wanting to suck, so they are
bring their hand to their mouths, looking, moving their head from side to side…’.
However, there were also other criteria not all of the groups had cited (Table 17).
Possible explanations of this could be that the other groups either did not mention the
criteria as they had not thought of it or not agreeing that it should be a criterion.
Table 17: List of criteria to decide feeding readiness not mentioned by all
groups
Feeding readiness criteria
Number of groups mentioning item
(n = 3)
Baseline body temperature within 36.4 to 37.2 degrees Celsius 1
Infant has not had a bradycardia or apnoea episode in the last 24 hours 1
Infant requiring no supplemental oxygen at baseline 1
Infant not requiring any additional supplemental oxygen during handling/cares
1
82
All groups mentioned that the medical stability of the infant was very important to
maintain. However, there were many different criteria mentioned across the three
groups of how to define whether an infant is medically stable to commence feeding
as seen in Table 16 and Table 17.
Similar to the minimum age criterion discussed earlier, medical stability criteria may
be relaxed depending on the individual infant. For example, when group 2 were
asked about their criteria for deciding breastfeeding readiness, the criterion not
receiving oxygen therapy was mentioned. However, later in the focus group, they
discussed that infants receiving oxygen therapy could still be allowed to attempt
breastfeeding.
Participant from group 2: ‘and if the baby is in oxygen, then I think it would have to
around—that the oxygen requirement goes no greater than say five per cent during
the activity or something like that’.
There were also a number of criteria described in the literature not mentioned by the
groups. When groups were asked about these criteria, each group mentioned that
they did not think that they were important criteria to decide an infant’s readiness to
commence breastfeeding. These criteria included caffeine usage, minimum weight
and if the infant was fussing/demanding at the time of feeding.
Criteria to determine infants are not ready to be breastfed. There were also a
number of criteria to decide when not to attempt breastfeeding mentioned during the
focus groups, with most of these criteria focused on the medical stability of the infant
(Table 18).
83
Table 18: List of criteria to decide an infant is not ready to feed mentioned by
all groups
Not ready to attempt breastfeeding criteria
Number of groups mentioning item
(n = 3)
Infant not awake when changing nappy/doing cares 3
Frequent periods of oxygen desaturation 3
Increased baseline heart rate 3
Increased baseline respiration rate 3
High oxygen requirement 2
Frequent apnoea and bradycardia episodes 2
Infant not staying awake once woken 2
Marked increase in work of breathing 2
Infant currently requiring CPAP 2
Infant currently requiring mechanical ventilation 1
Umbilical artery or venous catheter in situ 1
Absence of protective reflexes, for example gag, cough 1
Not interested in sucking 1
CPAP = continuous positive airway pressure
All groups gave examples of situations in which they thought that the infant was too
unstable to attempt or continue with a breastfeed.
Participant from group 3: ‘So say, you had a CPAP baby who was on like 22–23 per
cent [oxygen], and it’s requirements gone up to 40, I would probably just stop, would
probably put the baby back in by that sort of stage. Like if you see a great change and
that kind of thing, and its needs are up’.
A participant from group 1 went on to explain that there may not be a consistent
approach by which clinicians define the term “medically stable infant”. This may
reflect the number of different criteria when an infant is not ready to commence
breastfeeding.
84
Participant from group 1: ‘Some people might worry, like you know a couple of
brady’s [bradycardia] or apnoeas might be as an unstable criteria but you know if
you get the baby out for a kangaroo cuddle and that settles the baby down . What is
stability? And some people would look at a stable baby and say that a baby has had
an apnoea in the last 24 hours is unstable…Like some of the people say you
shouldn’t start them feeding when they are oxygen, but then you have all those
chronic lung babies, if you don’t start them when they’re…follow the piece of
string’.
However, all groups noted that there was a difference between infants who either
require additional support to safely attempt a breastfeed and those who were too
unstable to attempt.
Participant from group 1: ‘And even then they may not be completely stable, but you
would still offer them suck feeds so you don’t miss out on that window…and you
would boost them, absolutely, if they got to 35, 36 and they haven’t been offered a
suck, you would be very concerned and you may need to up their oxygen and
support’.
Participant from group 2: ‘If mum is coming up to look after the baby and he’s very
stable just on air and CPAP … we’ll often put the baby on intranasal oxygen for the
couple of hours they’re active, that makes nuzzling a lot easier for mum, you know,
and I think also for the baby than CPAP prongs or something. But again it depends
on the baby’.
Participant from group 3: ‘And that’s what I was saying about you know, you do
everything you can to optimise the stability and if that has, you know, if the baby is
on oxygen or caffeine…it doesn’t mean it’s a contraindication to them starting to
feed. It’s something in helping them to be stable enough to do that…’.
Interestingly, all focus groups mentioned that in a few cases they may attempt to
introduce breastfeeding to an unstable infant, who they might not consider to be
allowed to attempt if they were an earlier GA. As medical stability is very important
and may risk the safety of infant, the groups described that this would only be in
85
cases where infants had a long medical course and was either getting too close to or
have been delayed past the ‘window of opportunity’ in which infants tend to more
easily learn to breastfeed. This decision would need to be weighed carefully with the
benefits to the infant of commencing feeding against the risk to the infant’s health
and would only be done under close supervision by experienced staff. This is
illustrated well by group 1.
Participant from group 1: ‘We’ve been often asked to offer the first feed for the
longer term respiratory kids…They are at 35 weeks and they have never had any
sucks and someone says it’s now or never, well because there is that window, isn’t
there where if you let it go too long, you have to give them that opportunity. Two
suck with some fluid in the back of their mouths and bang they are down, their sats
are down in their boots, they have the big vagal response and that might happen for
the first feed, and then you try again the next day, a little bit better, and the next day
a little bit better, and then they are off, bang. So sometime it’s a stage, it’s a stage
thing. It’s a learnt thing, absolutely but it just requires such close supervision in the
very high-risk kids’.
Although there were a number of behavioural cues that were mentioned by all groups
that would indicate readiness, there was only one criterion that was mentioned by all
three groups, to indicate that the infant was not ready to feed. This criterion was with
regard to the infant’s behavioural state or how alert the infant was just before feeding
time.
Participant from group 1 : ‘I guess also at the bottom of the list [of not ready to
attempt] is if the baby is in a deep sleep, sleep wake cycles, there is not really much
point waking the baby up to try and breastfeed, especially if it’s the first attempt’
Participant from group 3: ‘…they need to be awake and um alert and not sort of very
drowsy or anything so because you’re probably not going to see any sort of cues
when they are really drowsy anyway so’.
86
4.4.11 Summary
The focus groups have demonstrated that age remains a criterion used by neonatal
care providers, with both minimum and maximum age criteria described. Although
all groups mentioned minimum age as a criterion, there was variation in how this
criterion was defined as well as descriptions that the minimum age criterion was not
being consistently applied to all infants. There was also a number of other criteria in
which all three focus groups mentioned that they used to decide when to initiate
breastfeeding in the preterm infant population. These items generally focused around
the broad criteria of behavioural cues, behavioural state and medical stability.
Although the sample size of the focus groups was smaller than anticipated it has
provided abundant information with regard to the criteria that neonatal care providers
utilise to determine an infant’s readiness to commence breastfeeding. However, the
number of groups mentioning criteria as a method of analysis can only give a
perspective of the relative importance of the criteria but cannot indicate or gain
consensus of what criteria should be included as an item in the PTSR scale. The next
phase used information obtained in the focus groups and the literature to attain which
criteria should be retained to form items for the PTSR scale.
4.5 Phase 3: Delphi Survey
4.5.1 Background
Studies have found that clinicians use a variety of criteria to decide when a preterm
infant is ready to commence breastfeeding. Phase 2 of this study utilised focus
groups to obtain a list of criteria that clinicians use to determine when to initiate
breastfeeding. For an instrument to be developed and used in clinical practice,
consensus of items to be included in the instrument has to be reached.
87
4.5.2 Aim
The aim of this phase was to reach a consensus on which items should be included in
the instrument. In particular, this study aimed to establish the content validity of the
items contained within the newly developed tool, the PTSR scale.
4.5.3 Research Question
What are the most important criteria to include in an instrument to determine when to
initiate breastfeeding in the preterm infant population?
4.5.4 Design
The Delphi survey utilises a ‘process of response-analysis-feedback-response’ which
is repeated approximately three times until general consensus is reached allowing a
list of decision-making criteria to be formed (Portney & Watkins, 2000). The Delphi
technique has become widely used by researchers to obtain agreement on an issue
(Keeney, Hasson, & McKenna, 2006) and is particularly useful in the validation and
development of the instrument (Waltz, Strickland, & Lenz, 2010). However, the
Delphi technique is not without limitations and careful consideration must be given
to the design of the study, particularly with regard to the establishment of the expert
panel, communication with panel members, structured versus unstructured first
round, the number of rounds and the process of gaining consensus (Chang, Gardner,
Duffield, & Ramis, 2010) .
4.5.5 Sample
Purposeful sampling was used to ensure: 1) the expertise of the panel members, 2)
that the participants were interested in the topic, 3) credibility within their field and
4) willingness to participate throughout the entire study, all of which is important in a
Delphi survey (Chang, et al., 2010). The inclusion criteria included: 1) willing to be
involved in a Delphi survey; 2) professional experience as a midwife or nurse,
speech pathologist, physiotherapist or neonatologist and/or knowledge of the
initiation of breastfeeding in preterm infants. Obtaining a sample involving variety of
88
different disciplines and units was conducted to reduce bias and improve the
reliability of the results (Hardy, O’Brien, & Gaskin, 2004; Waltz, et al., 2010).
There is no set sample size for a Delphi survey in the literature with some authors
using as few as fifteen, while others using over 60 participants (Hasson, Keeney, &
McKenna, 2000). Sample size is dependent on the scope of the problem and
resources available (Keeney, et al., 2006; Powell, 2003). In this study, 78 surveys
were sent out in an initial round.
4.5.6 Data Collection Methods
Each of the panel members was sent an e-mail with a link to an electronic survey.
Although other techniques could have been used, it has been found that when
compared to mail-based response systems, e-mail-based responses were more
legible, eased data entry and enhanced communication tracking (Snyder-Halpern,
Thompson, & Schaffer, 2000). Further, the use of e-mail also allowed participants to
remain anonymous to other panel members so that participants could respond more
openly and avoid the influence of dominant personalities, which may arise from
other response methods such as face to face meetings (Keeney, Hasson, & McKenna,
2001). Internet-based surveys have been found to be particularly useful when
participants are all interested in particular specific topic domains (Polit & Beck,
2008).
4.5.6.1 Feeding Readiness Indicators in Preterm Infants Survey (Appendix L)
This survey was developed by the researcher. At first, the items for the first round
survey were derived from the results of the focus groups with all criteria discussed
within the groups included. However, not all criteria to decide feeding readiness
described in the literature were discussed in these groups. Therefore, to limit any
possible bias of the focus groups, the researcher added these criteria as items into the
survey.
The survey asked participants to rate each of the criteria for possible inclusion into
the instrument on a four-point Likert-type scale (1 = not important, 4 = very
89
important). There are no standardised rating scales used in Delphi surveys; however,
traditionally, a four-point rating scale has been used (Polit & Beck, 2006).
The content of the items in the second round questionnaire remained unchanged from
the first round, although there were slight modifications in wording due to comments
received from panel members after round one. Only one panel member specifically
stated that she was unsure of the meaning of the wording of items B13 and B14.
However, changes to the wording of items B7, B8 and B12 were also made to be
more specific after a few participants’ comments indicated that these items needed to
be further clarified. The third round questionnaire only included items, where
agreement on inclusion of the item was not obtained in rounds one and two. Items in
the questionnaire for the third round were not modified.
4.5.7 Procedure
Both the MHS and QUT HREC gave approval to conduct this phase of the project.
The study used an expert panel to obtain consensus on which items were to be
included into the pilot instrument. All 26 level three neonatal nurseries in Australia
and New Zealand were contacted via the phone and asked if they could nominate
three health-care professionals with expertise and interest in determining the
readiness of preterm infants to breastfeed. Once the names were obtained,
participants were contacted initially via the phone to gauge their willingness to
receive information about the study and ensure correct work e-mail address. An e-
mail including a participant information letter (Appendix G) was then sent to
potential participants to inform them about the study and ask if they would be willing
to participate. It was expected that completing the survey would take about 10–15
minutes of the participant’s time for each round of the survey.
Information gathered from the literature review and focus groups was used to design
an initial questionnaire. In some Delphi surveys, the first round is unstructured,
giving participants a series of open-ended questions. This method of first round is
most appropriate in the situation where the aim is to generate solutions for an
existing problem (Hardy, et al., 2004). However, due to the work already undertaken
in the focus groups and existing literature, and that the purpose of the study was to
90
validate existing items, the first round was structured (Hardy, et al., 2004). At the
bottom of the questionnaire was a field in which participants could write any
comments.
Prior to the first round, the questionnaire was pilot tested by three neonatal care
providers to ensure the readability and functionality of the electronic survey form.
Minor alterations in content and format were made. Minor changes included
providing additional information to clarify the meaning of questions, such as in
Question A5 ‘i.e., putting babies to the breast’ was added to the end of the question
‘How many years of experience do you have in directly helping mothers breastfeed
preterm infants?’. Further, slight changes were also made in the terminology used in
a couple of the items; for example, ‘transcutaneous oxygen saturation levels’ was
changed to ‘pulse oximetry levels’ in items B5 and B6 and ‘gestational age’ was
changed to ‘corrected postnatal age’ in item B13.
The first round involved panellists being sent an e-mail with a link to an electronic
survey form, the Feeding Readiness Indicators in Preterm Infants Survey (Appendix
L). Participants were asked to rate the importance of each item separately on a 4-
point scale. Once completed, participants were asked to check their responses and
click submit so that their results were automatically sent to the researcher by e-mail.
Two weeks after the round one survey was distributed, e-mail reminders were sent to
the panel members who had not responded.
Round two involved panel members again scoring the list of the same items but this
time they were provided with their own score and the panel’s overall mean scores
(Schell, 2006). Rounds continued until an agreed list of decision-making criteria was
developed. Consensus was achieved in three rounds.
4.5.8 Data Analysis
The content validity index (CVI) was used to analyse data. The CVI agreement or the
percentage of agreement was calculated from the proportion of the number of
panellists scoring an item three or four divided by the total number of panellists in
the sample. There is no standardised cut-off point in Delphi surveys, with cut-off
91
points determined by scope and purpose of the survey. As the survey had more than
six participants, a CVI for the total instrument of .78 or 78 per cent, based on the
percentage of all items rated three or above, was used as the level of acceptable
content validity (Polit & Beck, 2006). Although CVI is the most commonly used
statistic by nurse researchers to measure consensus in Delphi surveys, there is some
debate of its use as it does not measure chance agreement (Polit, Beck, & Owen,
2007; Polit & Beck, 2006). Therefore, a modified kappa coefficient was also used to
test for chance agreement (Polit, et al., 2007).
4.5.9 Ethical Considerations
4.5.9.1 Consent and Information to Participants
The participants involved in the Delphi survey were neonatal care providers.
Participants received an information sheet prior to participation with time to consider
their participation (Appendix G). Participants were advised that completing and
returning the survey indicated that they understood the information provided and
consented to participate in the study.
An HREC approved information letter was provided to the participants giving a
description of the study and its purpose. The letter described what their possible
involvement would entail and any risk and benefits to participating in the study.
Potential participants were also informed of what steps would be taken to protect
their privacy and maintain confidentiality. Each study’s information letter stated that
participation was voluntary and that their participation would in no way impact on
any current or future relationship with either the study hospital or the university.
4.5.9.2 Risks to Participants
Phase 3 was considered to be ‘negligible risk’. For the Delphi survey, the risk or
burden associated with these projects was not expected to be greater than for their
normal day-to-day work. However, there was the possibility of a small risk of a
breach of their privacy and confidentiality. This was minimised for the Delphi survey
through the list of participants’ names and contact details which was not kept with
the information that they provided. However, this was not sensitive information and
92
relates to how they scored each item. Confidentiality of contact details or
documentation collected for the Delphi survey was stored at the Mater in a filing
cabinet or on a restricted access computer drive. The burden for the staff was
minimal for either project as individuals volunteered and could be conducted during
work time. No identifying information will be used in any report or publication.
4.5.9.3 Confidentiality
Any identifying information was only seen by the researcher and the research
assistants. When analysed and reported, all identifying information was removed
with each participant given a numerical code. Confidentiality of contact details or
documentation collected for the studies was maintained, and is still stored, in a
locked filing cabinet and a restricted access computer drive. No individuals were
identified, with only the hospital identified in this thesis and any publications.
In the Delphi survey, there was a small risk of a breach of their privacy and
confidentiality as a list of participants’ names and work e-mail addresses were kept
with the information that they provided. However, the information provided by
participants was not sensitive in nature and was needed so that participants have
access to their previous survey results.
4.5.9.4 Approvals
This phase of the study was reviewed by both the Mater and QUT HREC and was
deemed negligible risk. As a result, both ethics chairs provided a letter of exemption
from ethical review. Approvals to conduct the research projects were also granted
from the Director of Neonatology and the Nursing Director.
4.5.10 Results
4.5.10.1 Sample Characteristics
Characteristics of the first round sample (n = 43) are outlined in Table 19.
Participants were predominately female (88.4 per cent) and from a nursing or
midwifery background (65.1 per cent), although both medical (16.3 per cent) and
93
allied health (18.6 per cent) staff were represented. Just over half of the participants
(53.5 per cent) did not have one main area in which they worked as they worked
across both intensive and SCNs. However, there were twelve (21.9 per cent)
participants who worked exclusively in the ICN and eight (18.6 per cent) participants
that worked exclusively in SCN. Although neonatal care providers were targeted to
be experienced in working with preterm infants there was one participant (2.3 per
cent) who had less than one year and three participants (7.0 per cent) who had one to
five years of experience. However, the majority 28 participants (65.1 per cent) had
over eleven years of experience. There were eight participants (18.6 per cent) who
had no direct experience at assisting a mother to breastfeed her preterm infant. These
participants were either experienced medical staff (seven participants) or allied
health staff (one participant) who would not necessarily assist mothers to breastfeed;
however, they do participate greatly in the decision on when to commence
breastfeeding particularly in the high-risk preterm infant population. There were only
ten participants (23.3 per cent) who identified themselves as being an international
board-certified lactation consultant.
94
Table 19: Delphi survey sample characteristics (n = 43)
Characteristics Frequency %
Gender
Male 5 12
Female 38 88
Occupation
Medical 7 16
Nursing/Midwifery 28 65
Allied health 8 19
Main work area
ICN 12 28
SCN 8 19
No main area (Both ICN & SCN) 23 54
Years worked with preterm infants
less than one year 1 2.3
1 to 5 years 3 7
6 to 10 years 11 26
11 or more years 28 65
Years experienced in assisting mothers to breastfeed
No experience 8 19
less than one year 1 2.3
1 to 5 years 3 7
6 to 10 years 7 16
11 or more years 24 56
Qualified Lactation Consultant
No 33 77
Yes 10 23
4.5.10.2 Round One
Out of 78 health-care professionals, who were invited to participate in the Delphi
survey, 43 (55.1 per cent) decided to participate and complete the first round of the
survey (Appendix L). In round one, six items had an item CVI above the cut-off of
.78 (Table 20) indicating that 79 to 88 per cent of participants agreed that these were
important criteria in their decision to commence breastfeeding. Three of these items
related to the respiratory status of the infant with the remaining items related to the
infant’s behavioural state and showing interest at the breast.
Errors in chance agreement were checked by consulting the modified kappa statistic
table for 43 experts (Appendix P). From reviewing this table, it was noted that there
95
were two items that despite not reaching the cut-off of 78 per cent in this round, had
a modified kappa value above its cut-off for excellent agreement of 0.74. These items
related to changes in the baseline pulse oximetry during handing (77 per cent) and
sucking regularly on dummy (77 per cent).
Nineteen panellists completed the comments section of round one (Appendix Q).
Three of the comments made by the panellists were a clarification of why they
choose a particular response; for example, Participant 18 stated ‘B13 refers to a
preterm baby being at least 32–43 weeks’. Other comments made by the panellists in
round one suggest that deciding feeding readiness is complex with no single criterion
having to be achieved prior to breastfeeding being commenced. Eight panellists’
comments were in relation to the fact that they assessed the infant individually on a
variety of criteria, some of which being medical stability, level of activity, degree of
prematurity and showing interest at the breast. Five participants commented on the
importance of the use of skin-to-skin contact or kangaroo care in preparing the infant
for initiation of breastfeeding as well as its benefit both for the mother and baby. The
need to carefully observe infants during initial feeds was also mentioned by one
participant with a different participant stating the need to start breastfeeding during
the ‘window of opportunity’ even if they need support, for example increased oxygen
to do it. No additional feeding readiness criteria were suggested by the panel
members.
4.5.10.3 Round Two
Each of the 43 participants that completed the first round was sent the second
questionnaire containing the same 21 items with some minor modifications in
wording (Appendix L). Participants were also sent both their own score and the
group’s mean score for each item in round one as well as all comments received.
Due to some participants deciding not to participate in Round 2, the initial sample
was further reduced to 37 participants (47.4 per cent) completing the second Delphi
round. However, the level of agreement was slightly improved in round two with the
same six items above the cut-off 78 per cent agreement in round one, remaining
above the cut-off in round two with the percentages of agreement now ranging from
96
81 to 89 per cent (Table 20). An additional item B6, which described change in pulse
oximetry levels during care as a criterion was also now above the cut-off, which was
previously just below at 0.77. However, the other criteria B15, which was also just
under the CVI cut-off in round one decreased, with the percentage of agreement now
73 per cent. Further, when the modified kappa value was checked (Appendix P), the
category of agreement had also reduced from excellent to good. This item, B15,
related to the infant sucking regularly on a dummy or fingers.
Conversely, there were two other items which had very low percentage of agreement
in both rounds indicating agreement that the items B11 and B14 were not of
importance in their decision to commence breastfeeding. These items related to
whether the infant was receiving caffeine to treat apnoea of prematurity and the
current weight of the infant. The percentage of agreement for B11 only slightly
decreased from seven per cent agreement in the first round to five per cent in the
second round. However, there was a more substantial decrease in the percentage of
agreement for B14, which went from 23 per cent to five per cent. In both rounds,
each of these item’s modified kappa was found to be below the cut-off for fair
agreement also indicating poor relevance to the criteria of deciding feeding readiness
(Appendix P).
Only seven participants submitted comments in round two with some of the
comments reflecting views already expressed in the first round (Appendix Q).
However, there were additional comments made by the panellists. One comment
reflected that she believed that visual observation of an infant’s behaviour and
medical stability was more indicative of feeding readiness than purely relying on
equipment that measures an infant’s physiological stability. Another participant
commented that she did not believe that a ‘tick off’ criterion of whether an infant met
certain criteria would appropriately allow the initiation of breastfeeding. Further, she
stated the decision to commence breastfeeding was one that had to include critical
thinking and assessment of the individual infant. After changes were made to the
wordings of the item, there was only one panellist who commented on the wording of
items. This panellist stated that there was a difference between the behavioural states
of crying and fussing, which had been placed together in one item of B21.
97
Interestingly, this item was one of the items in which agreement was achieved in
both round one and round two.
4.5.10.4 Round Three
All panellists who responded to round two were sent the questionnaire again with
their own item ratings and the corresponding panel’s mean. The items in the
questionnaire had the same wording as round two, but the number of items were
reduced (Appendix L). Items (n = 6) in which agreement was already reached with a
percentage of agreement of above 78 per cent in the first and second rounds were
excluded. Further, the two items in which there was a clear agreement that they were
not important in deciding feeding readiness was also excluded; resulting in thirteen
items being included into the final round. These thirteen items had not reached the
specified CVI of 78 per cent and were either close to the cut-off level of agreement
or a difference between CVI values were noted between the first two rounds.
Twenty-seven panellists (34.6 per cent) completed the third and final rounds. In
round three, no item had a percentage of agreement above 79 per cent (Table 20) or a
modified kappa that could be categorised as demonstrating an excellent agreement
with its inclusion (Appendix P). There was little difference between this round and
the first round in relation to the level of agreement for a number of items with CVI
relatively constant for items B1, B7, B9, B10, B16 and B18. However, there are a
number of items such as B4, B5, B15 and B17 in which the CVI decreased with each
round. Only two items increased in the second round and then decreased in the third.
These items were B3, which related to the infant’s baseline respiratory rate and B12,
which related to whether an infant had a recent bradycardia or apnoea episode.
Panellists provided no further comments.
98
Table 20: Item CVI for three rounds
Item
Round 1
Round 2
Round 3
(n = 43)
(n = 37)
(n = 27)
CVI
%
CVI
%
CVI
%
B1 Baseline heart rate within 120 to 180 beats per minute 70 70 70
B2 Less than 5–10% difference in baseline heart rate during handling/cares
72 76 56
B3 Baseline respiratory rate within 30 to 60 breaths per minute prior to commencement of feed
88* 84* #
B4 Less than 5–10% difference in baseline respiratory rate during handling/cares
84* 73 0.59
B5 Baseline pulse oximetry level greater than 90% prior to commencement of feed
88* 89* #
B6 Less than 5–10% difference in baseline pulse oximetry levels during handling/cares
77 89* #
B7 Infant requiring no supplemental oxygen 21 27 22
B8 Infant not requiring any additional supplemental oxygen during handling/cares
44 35 19
B9 Infant's oxygen requirement not increased by 5–10% of baseline 58 68 63
B10 Baseline body temperature within normal limits prior to commencement of feed
63 59 67
B11 Infant not on caffeine/theophylline medication 7 5 #
B12 Infant has not had a recent bradycardia or apnoea episode 40 57 44
B13 Corrected Postnatal age 49 30 11
B14 Infant’s current weight 23 5 #
B15 Sucking regularly on dummy, fingers, feeding tube, etc. 77 73 63
B16 Mature non-nutritive sucking pattern 65 76 59
B17 Presence of rooting reflex 70 68 56
B18 Moves smoothly between behavioural (arousal) states 7 62 74
B19 Infant quietly alert just prior to feed time 86* 84* #
B20 Infant showing interest at the breast (licking, moving to nipple, etc.) during kangaroo care
84* 86* #
B21 Infant fussy/crying 79* 81* #
CVI = Content Validity Index * = Agreement reached with CVI > 0.78; # = item not included in questionnaire as agreement already established
99
4.5.11 Summary
Consensus on the importance of seven criteria to deciding feeding readiness in the
preterm population was achieved in the first two rounds of this Delphi survey. The
third round did not provide any more agreement and therefore, it was decided that
this would be the final round. Aiding in the decision to stop after round three was the
loss to follow-up with the number of participants continuing with the survey
decreasing from 43 panellists in round one to only 27 in round three.
In conclusion, the development of the PTSR scale was based on the results of phases
1 and 2, which informed the development of items as well as phase 3, which allowed
the formation of an agreed list of important criteria to determine breastfeeding
readiness in the preterm infant population. Nine criteria highlighted by the focus
groups and Delphi survey as important indicators of breastfeeding readiness were
incorporated into the final scale.
100
101
Chapter 5: Testing of the Instrument
5.1 Introduction
Chapter 5 follows on from the previous chapter which described the development of
the PTSR scale. This chapter details the methods and techniques used to
preliminarily test the PTSR scale’s ability to determine breastfeeding readiness in the
preterm infant population. A discussion on the major ethical implications is then
described. Finally, an overview of the findings of this phase is presented including a
description of how construct and criterion-related validity as well as inter-rater
reliability were established in this phase of the study.
5.2 Phase 4: Observational Study
5.2.1 Background
For an instrument to be beneficial in clinical practice, the validity and reliability of
the instrument needs to be established. There are a number of types of validity and
reliability, which may be tested including: content, construct and criterion-related
validities as well as inter-rater reliability and internal consistency (Onwuegbuzie, et
al., 2010).
5.2.2 Aim
The overall aim of this study was to initially establish the validity and reliability of
the PTSR scale. The specific objectives of this study were to initially establish the
following types of validity and reliability:
• Construct validity
• Criterion-related validity
• Inter-rater reliability
102
5.2.3 Research Questions
1. Is there a difference between the preterm and fullterm group in relation to items
in the newly developed PTSR scale?
2. Within the preterm infant group, is there a difference between the ages at birth in
relation to items in the newly developed PTSR scale?
3. Is there a relationship between the feeding outcome of milk intake and each item
in the newly developed PTSR scale?
4. Is there a relationship between the feeding outcome of total Preterm Infant
Breastfeeding Behaviour score and each item in the newly developed PTSR
scale?
5. Has the PTSR scale demonstrated inter-rater reliability?
5.2.4 Design
A prospective, observational design was used. Observational studies involve
collecting data by directly observing the behaviour or event of interest without
intervention. In structured observational studies, researchers specify in advance the
behaviour or events that they want to observe, and have outlined pre-determined
methods of data collection in which numerical data are collected (Polit & Beck,
2004). Unlike participant observation, it does not aim to capture a broad section of
everyday life but to document specific behaviours, actions or events (Polit & Beck,
2008).
5.2.5 Sample
A convenience sample was recruited from the MMH and Royal Brisbane and
Women’s Hospital (RBWH). Preterm infants included into the study were those who
would normally commence breastfeeding according to current clinical practice. The
timing of the initiation of feeding was not decided by the researcher but by the
clinical staff based on when they deem the infant ready to commence breastfeeding.
Infants included in the preterm group comprised of those aged between 30–36 weeks
PCA.
103
Other criteria included:
• inpatients of the MMH’s SCN
• tolerating enteral feeds
• medical stability
• their mothers’ wish to breastfeed
Another group of infants were recruited as a comparison group. This group
comprised of healthy, fullterm infants from the postnatal ward of the MMH. The
majority of infants in this group were able to breastfeed competently and therefore,
used for the purposes of this research as a known group.
Infants who were excluded from the study were: a) at risk of neonatal abstinence
syndrome; b) who have congenital malformations, syndromes; or c) severe
neurological problems. Infants from these groups have additional difficulties in
transitioning from enteral feeding to breastfeeding.
There is no consensus with regard to the sample size required for testing a tool, with
different authors suggesting a variety of ‘rules of thumbs’ to be used (Polit & Beck,
2008). Some authors suggest a minimum total of participants such as 300, while
others support a minimum number of participants per item, ranging from five to 20
(Polit & Beck, 2008). The instrument includes six items, which would mean that 120
preterm infants would be required if 20 infants per item were used. To check the
fullterm infant sample size required for testing construct related validity, analysis of
the first 20 observations was used to calculate the numbers needed. Therefore, for
this study a total sample size was calculated to be 140, including 120 preterm infants
and 20 fullterm infants.
5.2.6 Data Collection Tools
5.2.6.1 Demographic Data Sheet (Appendix M)
Infant and maternal demographic data such as birth weight, PMA, parity and reason
for delivery was collected from the chart prior to the commencement of the feeding
104
event. Other data was also collected including those who helped the mother to
breastfeed and the mother’s previous experience breastfeeding.
Feeding success was determined by both amount of milk intake and the sum of
breastfeeding behaviour exhibited.
Milk Intake. Weighing of infants has been used commonly by researchers to
determine the amount of milk taken by an infant during a breastfeed and has been
found to be a reliable indicator of milk intake (Chertok, Schneider, & Blackburn,
2006; Furman & Minich, 2004; Furman & Minich, 2006; Hedberg Nygvist, 2005;
Riordan, Gill-Hopple, & Angeron, 2005). The Wedderburn TI BLB12 Digital Baby
Scale was used. These scales are accurate to two grams for infants weighing less than
six kilograms (kg). To calculate the number of millilitres taken, one millilitre was
equal to one gram.
5.2.6.2 Preterm Infant Breastfeeding Behaviour (PIBBS) Scale (Appendix N).
The PIBBS can be used to describe the development of breastfeeding behaviour in
preterm infants (Hedberg Nygvist, et al., 1996; Hedberg Nygvist, et al., 1999).
PIBBS has been specially developed for use with preterm infants from 26 weeks
gestation (Hedberg Nygvist, et al., 1996). Six items (rooting, areolar grasp, latch,
sucking, longest sucking burst and swallowing) are assessed with a score being
attributed to each item (Hedberg Nygvist & Ewald, 1999). This instrument is used to
collect data during a breastfeeding event to assess how successful the infant was to
engage in feeding behaviour. Inter-rater reliability has been tested using two
observers with the cut-off for kappa values ranging from 0.68 to 0.84 (Hedberg
Nygvist & Ewald, 1999).
5.2.6.3 PTSR Scale (Appendix O)
The PTSR scale was developed by the researcher stemming from the work
completed in phases 1 to 3 of this research project.
105
The first part of the scale included three items relating to the physiological stability
of the infant with baseline respiration rate and oxygen saturation levels measured as
well as changes to oxygen saturation levels during handling. Data for these items
were both dichotomous and continuous.
The second part of the scale also included three items; however, these items relate to
the infant’s displayed behaviour. These items involve the infant’s behavioural state
and ability to display feeding-readiness behaviours. These items were initially
measured on a four-point adjective label scale. However, after pilot testing, the PTSR
scale items were measured on a five-point scale.
The initial version of the PTSR scale was piloted by three clinical nurses or
midwives to assess the tool’s usability and ease of use. Each clinical nurse or
midwife used the newly developed instrument when assessing feeding readiness in
their clinical practice for five feeds. The scale was also given to a neonatologist,
lactation consultant and a speech pathologist for comment and feedback.
As a result of this piloting there were minor changes in wording of the items, for
example the word ‘few’ was replaced with ‘occasional’. An additional adjective label
was also added to each of the items in the second part of the scale. This was done as
a result of its use in clinical practice by nurses or midwives, who felt that the four
original adjectives did not fully describe the item. Further, the order of the adjective
labels for the item ‘behavioural state just prior to disruption at feeding time’ was
changed. In this item, the adjective label of crying was moved from the final
adjective label to be placed prior to the alert states. Although behavioural states are
usually measured on a scale from asleep to crying, the neonatal care providers felt
that this item should instead reflect on optimal behavioural states for feeding with
‘quietly alert’ being the last adjective label. This is supported by studies in which the
quiet alert state is the optimum behavioural state to attempt breastfeeding (McCain,
2003; McGrath & Medoff-Cooper, 2002).
106
5.2.7 Procedure
Neonatal nurses/midwives were employed to work at either the MMH or RBWH by
a small grant obtained by the researcher to collect data. Each data collector worked at
the hospital that they normally worked and had previous experience at assisting
mothers to breastfeed preterm infants; however, not all had formal experience in
assessing breastfeeding behaviour and intake. All data collectors attended a one-hour
session explaining the study and the data collection tools including instructions on
how to assess breastfeeding and infant behaviour using the PIBBS and PTSR tools.
This education was provided by the researcher and an expert lactation consultant
knowledgeable in assessing preterm infant breastfeeding. Further, all research
nurses/midwives observed at least two breastfeeds in the study units with the
researcher to ensure that research nurses/midwives were assessing infants correctly
and allowing any further questions or issues to be raised. Moreover, before data
collection began, the research nurses/midwives in pairs also assessed infants using
the data collection tools, this continued until data collectors consistently obtained the
same results for three infants. The research nurses/midwives collected data on days
in which they were not rostered to work in the study unit. An e-mail with an attached
information letter was sent in the study units to inform them about the study
(Appendix H) and when possible nursing staff and lactation consultants received
one-to-one in-service on the study and their possible involvement.
Once a preterm infant was deemed ready to commence breastfeeding by the clinical
staff and informed consent was obtained from the infant’s parent, a time which was
convenient for both mother and clinical staff to be present was organised. Data
collection involved a single observation occurring during the first five attempts at
commencement of breast/bottle feeding. Collection of the data began with infants
being assessed by the research nurse/midwife using the PTSR scale just prior to a
scheduled feeding time. To complete the PTSR scale, infants were required to be
connected to a monitor that measured HR and oxygen saturation levels, although
most infants were already monitored, there were a few infants that were not. In this
situation, infants were connected to monitoring equipment to measure their HR and
oxygen saturation level with the respiratory rate calculated by observing the infant
breathing. Infants were not connected to the monitoring equipment until PTSR items,
107
which required the infant not to be disturbed were collected. The research
nurse/midwife then weighed the infant before the feed ensuring the warmth and
condition of the infant. A lactation consultant or clinical nurse helped the mother and
infant during the feed as it common practice in the study hospital for initial feeds to
be monitored by nurse/midwife to ensure the physiological stability of the infant and
assist the mother if she is unsure. The person collecting the data did not actively
participate in advising/assisting with the feed. Questions or issues arising during the
observational period were referred to the nurse assigned to the care of the particular
infant. The infant was observed at the breast for a period of at least 20 minutes. After
the breastfeed was completed and before any additional milk was given via the
gavage tube, infants were re-weighed. For both the pre and post-feed weights, the
infants were weighed in the same nappy, clothes and blanket that they were fed in.
A comparison group of 20 fullterm healthy infants was also observed during a
breastfeed as per the preterm infant protocol. A total of 20 per cent of the feeding
readiness assessments using the PTSR scale was conducted by two people for the
inter-rater reliability. To check the data, frequency distributions were noted for
invalid values and outliers (Polit & Beck, 2004).
5.2.8 Data Analysis
Descriptive statistics were used to summarise categorical data. Prior to undertaking
statistical analyses, continuous variables were checked for normality. Depending on
the normality of the data, means and standard deviations (SDs) or medians and
ranges were to be used to summarise continuous data. The level of significance was
set at 0.05 and data analysed using SPSS version 18.0.
To establish construct validity of the items in the PTSR scale, the known groups
method was used (Polit & Beck, 2004). This method was used to compare the scores
for each item of the PTSR scale between the preterm infant group and those of the
fullterm infant group. For dichotomous data and categorical data, a chi-square test
was used. In the situation when the assumption of a cell count greater than five was
violated, the analysis would be performed using a Fisher’s exact test. As the
assumption of normalcy of data was violated, the non-parametric statistic of Mann-
108
Whitney test was used to compare continuous data between the group of preterm and
fullterm infants. Further, the relationship between preterm age categories and the
items in the PTSR scale was also examined. This analysis was also conducted by chi-
square test.
On a bivariate level, relationships among the time taken from birth to reach feeding
milestones and infant characteristic and other continuous variables was tested by
Pearson’s Correlation Coefficient if normally distributed (Portney & Watkins, 2000).
If not, Spearman’s correlation coefficient was used to determine correlations among
continuous variables and time taken to reach feeding milestones (Portney & Watkins,
2000). Point-Biserial Correlation Coefficient was used to determine relationships
between dichotomous variables such as proven infection and time taken to attain
feeding milestones (Portney & Watkins, 2000). A multivariable model was then used
to examine the influence of items in the PTSR scale and other infant and maternal
factors on the time taken to attain a feeding milestone (e.g., first attempt at
breastfeeding).
The use of factor analysis in this study was not statistically or theoretically
appropriate. Factor analysis is used by researchers to test whether items in an
instrument truly cluster around one or more dimensions as well as indicating whether
the items reflect a single construct or multiple constructs (Lo-Biondo-Wood &
Haber, 2006). It was initially thought from visual inspection of the items that perhaps
there might be two constructs; with the first three items reflecting safety to
commence breastfeeding and the next three items reflecting infant behaviour related
to readiness to commence breastfeeding. However, on closer examination, these
items were found to be single items and not a reflection of a construct with infants
able to have convergent score for items in the same proposed construct. For example,
infants may have a low, normal or high respiratory rate per minute and still have an
oxygen saturation level over 90 per cent. The results of the Delphi survey did not
provide the researcher with constructs, but rather provided specific criteria to be
included into the instrument and therefore, factor analysis was not performed.
Testing the predictive validity of the instrument included examining the relationship
between each of the items in the newly developed PTSR scale and the short-term
109
outcomes of milk intake and level of breastfeeding behaviour. As both of these short-
term outcomes were not normally distributed, the Mann-Whitney was used for
dichotomous items, Kruskal-Wallis ANOVA for categorical items with three or more
groups and Spearman’s correlation coefficient for continuous variables (Lo-Biondo-
Wood & Haber, 2006).
Inter-rater reliability was determined firstly by calculating the proportion of
agreement between the two raters and then a Cohen’s kappa statistic was used to
adjust for chance agreements (Polit & Beck, 2008). However, no further reliability
was undertaken in this study. The stability of the instruments could not be examined
by test-retest. Individual infants in the early transition to breastfeeding period have
been shown to vary significantly in their abilities to engage in breastfeeding
behaviour from feed to feed and even across a given feed event (Thoyre, et al.,
2005). Further, as items in the PTSR scale are single items and cannot be added, it
was inappropriate to measure the internal consistency of the instrument, which is
designed to measure the items of the same construct (Polit & Beck, 2008).
5.2.9 Ethical Considerations
Both the MHS and QUT HREC gave approval to conduct this project. In addition,
the RBWH HREC gave permission to conduct stage II of the research project, the
only stage involving the RBWH.
5.2.9.1 Consent and Information for Participants
In the observational study, mothers of breastfeeding infants meeting the inclusion
criteria, were approached, and their possible involvement in the study was verbally
explained to them. They were also given a written information sheet (Appendix I and
J). When possible, mothers were given a few days to decide whether to participate.
Once they agreed to participate, they were asked to sign a consent form (Appendix
K). As the study involved observing a breastfeed, the mother had to consent for both
herself and her infant to be involved. An information letter was also provided to all
staff working in the areas in which data collection was to occur. Further, several in-
service education sessions were provided to lactation consultants and nursing staff.
110
Although the study wards’ lactation consultants and nursing/midwifery staff would
not collect data, they would assist the mother to breastfed as per the usual duties as
well as identify infants ready to commence breastfeeding (Appendix H).
An HREC approved information letter was provided to the participants giving a
description of the study and its purpose. The letter described what their possible
involvement would entail and any risk and benefits to participating in the study.
Potential participants were also informed of what steps would be taken to protect
their privacy and maintain confidentiality. Each information letter stated that
participation was voluntary and that their participation would in no way impact on
any current or future relationship with either the study hospital or the university.
5.2.9.2 Risks to Participants
Stage IV was considered to be low risk. The NHMRC define low risk research
‘where the only foreseeable risk is one of discomfort. Where the risk, even if
unlikely, is more serious than discomfort, the research is not low risk’ (National
Health and Medical Research Council, et al., 2007). It was expected that by
participating in this research the infants were not placed at any greater risk than their
usual care involves. Exclusive breast milk is recommended for infants under the age
of six months, particularly in the preterm infant population. Mothers of infants in this
study would have chosen to breastfeed their infant prior to commencement of the
study. Participation included observation of a feed and weighing pre and post feed.
Observation of a feed by health-care professionals is a common practice, especially
in the early stages of breastfeeding to ensure infants are safely and competently
feeding. The weighing of the infants to see weight patterns is also a common
practice, although pre and post-feed weights are not. However, this was a one-time
event with an experienced nurse/midwife who was able to quickly weigh the infant.
Infants were weighed pre and post feed in the same clothes and nappy that they were
wearing to minimise disruption and heat loss. The inclusion criteria also minimised
risk. To minimise any risk of mistiming of the feed, breastfeeding was commenced
via the normal routine of the unit by the clinician caring for the infant. Requiring
infants to tolerate enteral feeds and medical stability also ensures the safety of the
infant.
111
5.2.9.3 Confidentiality
Any identifying information was only seen by the researcher and the research
assistants. When analysed and reported all identifying information was removed with
each participant given a numerical code. Confidentiality of contact
details/documentation collected for the studies was maintained, and is still stored, in
a locked filing cabinet and a restricted access computer drive. No individuals were
identified with only the hospital identified in this thesis and any publications.
5.2.9.4 Other Approvals
Approvals to conduct the research projects were also granted from the relevant
Nursing Unit Managers, Directors of Neonatology, Directors of Nursing and
Executive Directors to conduct data collection at both the MMH and RBWH.
5.2.10 Results
5.2.10.1 Response Rate
During the ten-month study period 202 preterm infants were identified as eligible to
participate in the study. However, only 145 mothers were approached to participate
because additional information about 57 infants indicated they were no longer
eligible for inclusion. Of the mothers who were approached, a further 18 mother and
infant dyads did not participate in the study, six declined and 12 did not participate
for a number of reasons. Four of the six mothers who declined to participate in the
study did so as they were not interested and two mothers stated they felt too
overwhelmed dealing with the circumstances arising from their infants’ preterm
birth. Reasons for the remaining 12 mothers not participating were: 1) their infant
already established breastfeeding; 2) that they were not available during the time that
the data collector was available to collect data (e.g., visited mainly in evening) or 3)
they agreed to participate but the data collectors were not contacted to observe the
breastfeed or the mothers did not visit at the arranged time.
112
Therefore, a total of 127 mother–preterm infant dyads (87.6 per cent) were included
in the study with 67 recruited from the MMH and 60 from RBWH. However, only
107 preterm infants’ data were used in analysis to test the validity of the PTSR scale
with the remaining 20 preterm infants’ data used to test inter-rater reliability. Figure
3 shows the summary of participants available for the study. There were also 20
fullterm mother and infant dyads not specified in Figure 3 that were recruited from
the MMH to act as a comparison group.
Figure 3: Summary of study participants
Inter-Rater Reliability
n = 8
Single Observation
n = 59
Inter-Rater Reliability
n = 12
Single Observation
n = 48
MMH
n = 67
RWBH
n = 60
Participated
n = 127
Did not participate
n = 18
Declined (n=6)
Not Available (n=4)
Data Collector Not Contacted (n=3)
Baby had 5 BF Attempts (n=5)
Mothers Not Approached
n = 57
Mothers Approached
n = 145
Eligible Participants
n = 202
• Mother decided to bottle-feed
• Had Already reached 5 Breastfeeding Attempts
• Medical Reasons
113
5.2.10.2 Sample Characteristics
As no statistically significant differences were found between the pre-term infant
sample from MMH and RBWH and therefore, analysis was performed as one sample
(Table 21).
The median age at birth was 32.93 weeks PCA, ranging from 24.28–36.9 weeks. Just
over half of the infants (54.3 per cent) were male. At birth, the median weight was
1754 g (range = 652–3278g) with a median Apgar score at 1 minute of 8 (range = 2–
9). The majority of preterm infants (81.3 per cent) experienced at least one
complication following birth with the most common complication of physiological
jaundice of the newborn (59.8 per cent). This was closely followed by respiratory
distress syndrome (RDS), which was diagnosed in 62 infants (57.9 per cent). Other
respiratory complications were not as common, with nine infants (8.4 per cent)
having transient tachypnoea of the newborn, sixteen infants (fifteen per cent) with
recurrent apnoea and finally only three infants (2.8 per cent) diagnosed with CNLD.
Mechanical ventilation was provided to 27 infants (25.2 per cent) to treat respiratory
issues, with only nine infants (8.4 per cent) requiring mechanical ventilation for 49
hours or longer. CPAP was more commonly provided to support respiration with 69
infants (64.5 per cent) requiring CPAP and 36 infants (33.6 per cent) requiring CPAP
for 49 hours or longer. Infection was only confirmed in 3.7 per cent of infants;
however, it was suspected in 35 cases (32.7 per cent).
Sixty-preterm infant mothers (72.3 per cent) had singleton pregnancies, nineteen
(22.9 per cent) having twins and four (4.8 per cent) with triplets resulting in a total of
83 mothers. In singleton pregnancies (n = 60), infants were predominately delivered
by caesarean section (61.7 per cent) followed by spontaneous vaginal birth (31.7 per
cent) with the least number of infants being delivered by breech or instrumental
deliveries (6.6 per cent). When a multiple birth occurred, the method of delivery was
sometimes different for each infant; however, this was only the case for twin births
(Table 22) with all infants in each triplet set (n = 11) delivered by caesarean section.
Differences between the numbers for twin and triplet pregnancies may also be the
result of the death of one of the infants or that both infants were not ready to feed
during the study period.
114
Table 21: Sample characteristics of the preterm (n = 107) and fullterm (n = 20)
infants
Characteristic Preterm Infants Fullterm infants
Frequency % Frequency %
Hospital Mater Mothers’ Hospital 59 55.1 20 100
Royal Brisbane and Women’s Hospital 48 44.9 0 0
Gender Male 58 54.3 6 30
Female 49 45.7 14 70
Multiple Birth a No 59 55.7 20 100
Yes 47 44.3 0 0
Any neonatal complication No 20 18.7 17 85 Yes 87 81.3 3 15
Hours of mechanical ventilation a None, did not receive mechanical ventilation
79 74.5 20 100
Less than or equal to 48 hours 18 17 0 0
49 hours to 7 days 5 4.7 0 0 8 to 14 days 2 1.9 0 0
15 to 28 days 1 0.9 0 0
Over 29 days 0 0
Hours of CPAP a None, did not receive mechanical ventilation
38 35.8 20 100
Less than or equal to 48 hours 32 30.2 0 0
49 hours to 7 days 13 12.3 0 0
8 to 14 days 5 4.7 0 0
15 to 28 days 13 12.3 0 0 Over 29 days 5 4.7 0 0
a = one missing case; CPAP = continuous positive airway pressure # fullterm sample used as a known group to establish construct validity
Table 22: Type of delivery for twin births (n = 36)
Type of delivery
Twin 1 (n = 19) Twin 2 (n = 17)
Frequency (%) Frequency (%)
Caesarean Section 14 73.7 15 88.2
SVD 5 26.3 2 11.8
Breech/Instrumental 0 0 0 0 SVD = spontaneous vaginal delivery
115
The presenting obstetric complication leading to preterm birth is reported for each
first born infant (n = 83) in Table 23. Preterm labour (28.9 per cent) was the common
obstetric problem leading to intervention. This was followed by PPROM (22.9 per
cent) and PIH (14.5 per cent).
Table 23: Obstetric complications leading to delivery
Presenting obstetric problem Frequency
Percentage (%)
PPROM 19 22.9
Preterm labour 24 28.9
PIH (BP greater than 130/90) 12 14.5
APH/Abruption 8 9.6
Foetal distress leading to intervention 3 3.6
IUGR 6 7.2
Decreased Doppler flow 2 2.4
Other 9 10.8 PIH= Pregnancy induced hypertension; APH = Antepartum haemorrhage; PPROM = Preterm pre-labour rupture of membranes; IUGR = Intrauterine growth restriction; BP= blood pressure
The mean age at first feeding attempt was 34.24 weeks PCA (SD = 1.81) with the
mean age at data collection slightly higher at 34.66 weeks PCA (SD = 1.69). Thirteen
infants (12.1 per cent), had not previously attempted breastfeeding prior to data
collection. The remaining infants had attempted breastfeeding between one and five
times. Of the infants that had previously attempted breastfeeding 72.9 per cent
attached and sucked at the breast (Table 24).
Table 24: Frequency and quality of previous feeding attempts
Characteristic Frequency Percentage (%)
Number of feeding attempts
No previous attempt 13 12.1 One previous attempt 23 21.5 Two previous attempts 21 19.6 Three previous attempts 19 17.8 Four previous attempts 29 27.1 Five previous attempts 2 1.9 Was any attempt successful
Has not had any previous attempts 13 12.1
No interest 2 1.9 Licked at the breast 13 12.1 Attached, occasional suck 36 33.6 Attached sucked well 42 39.3
116
Over half of the 83 mothers (60.2 per cent) had no previous breastfeeding
experience. At data collection, only a small proportion of mothers used a nipple
shield (10.3 per cent). The stage of the mothers’ breast milk was predominately
mature (65.1 per cent) with the remaining mothers classified as either having
colostrum or transitional milk. Although all mothers provided their infants with
breast milk, less than half (41 per cent) did not produce enough to meet their infant’s
feeding demand with infants requiring supplementation with either formula or IV
fluids. Twenty mothers’ milk supply (24.1 per cent) was adequate for their infant’s
demand. Finally, 34.9 per cent of mothers produced more than what their babies
required.
At data collection (n = 107) preterm infants and mothers were assisted to breastfeed
by either the clinical nurse looking after the infant (54.2per cent), lactation consultant
(10.3per cent) or another clinical nurse working in the unit (0.9per cent). Thirty-
seven (34.6 per cent) mother–infant dyads did not receive any help with
breastfeeding. Only two (1.9 per cent) out of 107 infants received oxygen therapy.
A total of 20 fullterm breastfeeding infants were recruited from the MMH’s postnatal
floors for comparison with the preterm infant group, in testing the construct validity
of the PTSR scale (Table 21). The mean age at birth was 39.42 (SD = 0.1) weeks
PCA. The majority of infants were female (70 per cent). At birth, the mean weight of
infants was 3453.4 g (SD = 364.93) with the average Apgar score at one and five
minutes being 8.25 (SD = 1.33) and 9.05 (SD = 0.51), respectively. All infants were
singletons. Just over half of infants were delivered via a caesarean section (55 per
cent) with a further six infants (30 per cent) being born by spontaneous vaginal
delivery. Only three infants (15 per cent) were delivered by instrumental delivery.
The only presenting obstetric complication was PIH (15 per cent) with the remaining
seventeen infants (85 per cent) having no presenting obstetric complications. Infants
predominantly had no neonatal complication (85 per cent) and did not require
medical intervention. Only one infant was diagnosed with tachypnoea of the
newborn, who required only a few hours of cot oxygen. There was no confirmed
infection in the sample although it was suspected in two cases (ten per cent).
117
All infants had at least four attempts at breastfeeding with majority of infants having
five previous attempts (95.0 per cent). Further, all of these were considered
successful feeding attempts in which infants required no supplementation. The
average age of infant at data collection was 39.65 (SD = 0.97) weeks PCA. No
infants received oxygen therapy at the time of the feed. Just over half of mothers had
breastfed previously (55.0 per cent). During the breastfeeding observation period for
each infant, no mother received assistance with breastfeeding from staff.
Testing of the Preterm Sucking Readiness Scale
5.2.10.3.1 Milk Intake
Although not statistically significant, the median milk intake in the fullterm infant
group was higher than in the preterm infant group. In the fullterm infant group (n =
19) the median milk intake was five millilitres (0–76), whereas in the preterm infant
group (n = 103), the median milk intake was two millilitres (0–56). The data was not
evenly distributed with 43.8 per cent of preterm infants (n = 103) and 30 per cent of
fullterm infants (n = 20) documented as consuming no breast milk.
5.2.10.3.2 Breastfeeding Behaviour
Total scores for the PIBBS ranged from 0 to 20 with a score of zero indicating no
breastfeeding behaviour exhibited. For the preterm sample (n = 107) the mean total
PIBBS score was 12.37 (SD = 5.49), which was significantly lower than the fullterm
infant group’s mean score of 16.53 (SD = 3.6) indicating that the fullterm group
exhibited greater breastfeeding behaviour than the preterm infant group (t35.32 =
−4.223, p ≤ .0001).
5.2.10.3.3 Comparisons between fullterm and preterm groups and within preterm infants group
Item 1: Baseline Respiration Rate is within Normal Limits of 30–60 Breaths Per
Minute. There was no significant difference between the fullterm and preterm
118
groups at the dichotomous level of measurement (Fisher’s exact test = .499). In the
preterm group, (n = 107) nearly all infants (96.3 per cent) met the criteria of having a
respiratory rate between 30 and 60 breaths per minute. While, in the fullterm group
(n = 20), all infants met the criteria. However, at the continuous variable level there
was a statistically significant difference between the two groups, with the preterm
group (n = 107) having a lower mean respiratory rate of 46.08 (SD = 9.45) breaths
per minute than the fullterm infant group (n = 20) with a mean respiratory rate of
51.65 (SD = 34–59) (t33.5 = −3.07, p = .004). Although this is a statistically
significant result, it is unlikely such differences would be clinically relevant. Further,
when the preterm infant group was categorised by age at birth and compared, no
differences were found across the three groups with regards to baseline respiratory
rate (H2 = 4.309, p = .12).
Item 2: Baseline Pulse Oximetry Level above 90 Per Cent. All infants in both the
preterm group (n = 105) and the fullterm group (n = 20) met the criteria of a baseline
pulse oximetry level above 90 per cent at baseline. No difference was found between
the preterm and fullterm groups in the lowest baseline pulse oximetry level observed
(U = 88.50, z = −1.192, p = .233). Further analysis of lowest baseline pulse oximetry
level in the preterm sample (n = 104), by age group at birth revealed no statistically
significant differences (H2 = 1.318, p = .517).
Item 3: Change in Baseline Pulse Oximetry Level during Handling or Care. In
the preterm group (n = 101), 64.7 per cent of infants had less than five per cent
change in the baseline pulse oximetry level during handling, which was similar to the
fullterm group (n = 19), where 63.2 per cent also had less than five per cent change
(Pearson χ21 = 0.317, p = .574). When examining if infants had less than ten per cent
change in their baseline pulse oximetry level during handling, there was again no
difference between the groups with the 94 per cent of preterm infants and 100 per
cent of fullterm infants having less than ten per cent change (Fisher’s exact test =
.588). There was also no significant difference (U = 793.5, z = −1.140, p = .254) in
the median lowest pulse oximetry level during handling which was 94 per cent in
both the preterm and fullterm groups. However, there was a wider range of lowest
pulse oximetry levels recorded in the preterm group (56–100 per cent) than the
fullterm group (86–98 per cent). When the lowest pulse oximetry level during
119
handling was compared across the preterm infants’ age at birth categories, there was
a significant difference (H2 = 7.22, p = .027). There was no statistically significant
difference between the extremely preterm and moderately preterm (U = 83.5, z =
−1.60, p = .109), or the moderately preterm group and the mildly preterm group (U =
747.5, z = −1.62, p = .105). Although close to the cut-off, the difference between the
extremely preterm group and the mildly preterm did not reach statistical significance
once the Bonferroni correction was applied (U = 110.0, z = −2.37, p = .018).
Item 4: Behavioural State Prior to Feed. The preterm group had a statistically
lower mean rank score of 60.83 when compared to fullterm group who had a mean
rank score of 80.97 (U = 730.5, p = .020). This is illustrated in Table 25 when the
percentages of infants who were asleep or drowsy were greater in the preterm infant
group when compared to the fullterm infant as well as fewer preterm infants were
crying. Interestingly, the percentages of those infants who were actively alert and
quietly alert were similar between the two groups. In the preterm infant group, no
statistically significant difference was found between the age at birth groups with
preterm infants in each group having similar mean rank scores (H2 = .688, p = .716).
Table 25: Preterm and fullterm infants’ behavioural state just prior to feed time
Preterm Group Fullterm Group
(n = 107) (n = 20)
Frequency % Frequency %
Asleep (0) 46 43.0 1 5.0
Drowsy (1) 23 21.5 6 30.0
Crying (2) 2 1.9 6 30.0
Active alert (3) 15 14.0 3 15.0
Quietly alert (4) 21 19.7 4 20.0
Item 5: Transition between Behavioural States during Handling. There was a
statistically significant difference between the preterm and fullterm groups for
transition between behavioural states during handling, with the preterm group having
lower mean rank (61.14) than the fullterm group (79.28) (U = 764.5, z = −2.329, p =
.020). Table 26, shows the preterm group had a higher percentage of infants who
were either asleep or drowsy during handling. However, there was no significant
120
difference in the preterm sample when infants were compared by age at birth with
similar mean ranks found across the age at birth categories (H2 = .775, p = .659).
Table 26: Preterm and fullterm infants’ transition between behavioural state
during handling
Preterm Group
(n = 107)
Fullterm Group
(n = 20)
Frequency % Frequency % Remain asleep (0) 1 0.9 0 0 Briefly alert but goes quickly asleep during handling/cares(1)
13 12.1 1 5
Briefly alert but gets drowsy during cares/handling(2)
17 15.9 0 0
Alert during cares but then becomes drowsy in immediate post-care/handling period (3)
14 13.1 2 10
Alert during cares and the immediate post-care/handling period (4)
62 57.9 17 85
Item 6: Feeding-Readiness Behaviours. When comparing the preterm and fullterm
infant groups, the preterm group had a significantly lower mean rank (58.28) than the
fullterm infant group (91.15) with the preterm infant displaying less feeding-
readiness behaviours than the fullterm group (U = 507.0, z = −3.832, p ≤ .0001).
However, when comparing preterm infants by age at birth, there was no significant
difference between the three age categories (H2 = 1.291, p = .524).
Table 27: Preterm and fullterm infants’ display of feeding-readiness behaviours
Preterm Group Fullterm Group (n = 106)a (n = 20)
Frequency % Frequency % Displays no feeding-readiness behaviours(0) 16 15.1 0 0 Displays occasional feeding-readiness behaviours (1)
44 41.5 3 15
Displays intermittent feeding-readiness behaviours(2)
24 22.6 6 30
Displays frequent feeding-readiness behaviours(3)
12 11.3 4 20
Displays continuous feeding-readiness behaviours (4)
10 9.3 7 35
a = one missing case
121
5.2.10.3.4 Relationship between Milk Intake and Items in the PTSR Scale
There was no relationship between whether an infant had milk intake and the
baseline respiration rate (Fisher’s exact test = 1, p = .630). Further, there was also no
relationship between whether the infant recorded a milk intake and had a change in
baseline pulse oximetry levels at either the five per cent (Pearson χ21 = .794, p =
.373) or ten per cent (Fisher’s exact test = 1, p = .386) level. The relationship
between milk intake and whether infants’ had a pulse oximetry level above 90 per
cent was not tested as all infants had a pulse oximetry level above 90 per cent.
There was also no relationship between milk intake and the final three items in the
PTSR scale. A relationship was tested for behavioural state (U = 1207.0, z = −1.021,
p = .307), transition between behavioural states (U = 1338.0, z = −0.137, p = .891)
and displaying feeding-readiness behaviour (U = 1184.0, z = −1.030, p = .303).
As there was no significant relationship between milk intake and the items in the
PTSR scale, no multivariate analysis was performed.
5.2.10.3.5 Relationship between Breastfeeding Behaviour and Items in the PTSR
Scale
At the bivariate level there was no relationship between total PIBBS score and
whether an infant’s respiratory rate was within normal limits (t103 = .232, p = .817);
however, when the association between actual baseline respiratory rate and total
PIBBS score was investigated there was a weak, significant association found
between the two variables (n = 105, Kendall’s r = .147, p = .033). This was similar
for the variable change in baseline pulse oximetry levels during handling, where
there was no significant relationship at the dichotomous level of measurement, but
there was a weak, significant relationship between the continuous variable lowest
pulse oximetry level during handling and total PIBBS score (n = 98, Kendall’s r =
.178, p = .014). The relationship between total PIBBS score and the dichotomous
variable of baseline pulse oximetry level above 90 per cent could not be tested, as all
infants had a pulse oximetry level above 90 per cent. The association between actual
122
baseline pulse oximetry level and total PIBBS score could be tested; however, this
association was not significant (n = 105, Kendall’s r = .093, p = .205).
There was a weak, significant relationship between the dependent variable of total
PIBBS score and behavioural state prior to feeding (n = 105, Spearman’s r = .204, p
= .037) as well as feeding-readiness behaviours (n = 104, Spearman’s r = .429, p ≤
.0001). However, the association between transition between behavioural states and
total PIBBS did not reach the cut-off for statistical significance (n = 105, Spearman’s
r = .185, p = .059).
Significant associations at the bivariate level between items in the PTSR and total
PIBBS score were placed into a regression model to determine each item’s
independent influence on the outcome of total PIBBS. The overall model was
significant explaining 25.7 per cent of variation in the total PIBBS score.
Table 28: Multi-variable model of total Preterm Infant Breastfeeding Behaviour
(PIBBS) and PTSR variables
Unstandardised
Coefficients Standardised Coefficient 95% CI for B
B Std
Error Beta t sig Lower bound
Upper bound
Behavioural state prior to feed 0.042 0.363 0.012 0.116 0.908 -0.679 0.763 Transition between behavioural states 0.107 0.485 0.022 0.22 0.827 -0.856 1.069 Feeding-readiness behaviours 1.725 0.511 0.365 3.375 0.001 0.71 2.74 Actual baseline respiration rate 0.09 0.054 0.155 1.681 0.096 -0.016 197 Lowest pulse oximetry level during care 0.204 0.086 0.220 2.378 0.019 0.034 0.374
In this model, there was statistical evidence that total PIBBS score was associated
with feeding-readiness behaviour and lowest pulse oximetry during cares (Table 28).
The PTSR item of feeding-readiness behaviour had the strongest association with
total PIBBS score with its individual contribution to the model being 9.2 per cent.
This item had a positive relationship such that the more the feeding-readiness
behaviour exhibited the greater the total PIBBS score. Similarly, lowest pulse
oximetry during care also had a positive relationship such that the higher the pulse
123
oximetry during cares higher the total PIBBS score. This item’s unique contribution
to the model was 4.6 per cent. However, the other items such as baseline pulse
oximetry and behavioural state prior to feed although significant at the bivariate
level, did not contribute to the model. Interestingly, when the model was adjusted for
age at birth, this variable was not significant and excluded from the model. One
explanation for this may be that the variation explained by age at birth was already
accounted for in the model.
Inter-Rater Reliability Testing
5.2.10.4.1 Sample Characteristics
Four research nurses/midwives collected data at 20 breastfeeding events; two
research nurses/midwives observed twelve breastfeeding events at the RBWH, while
two research nurses/midwives observed eight breastfeeding events at the MMH
(Table 29). All data collection for testing inter-rater reliability occurred in SCN and
involved preterm infants. The median age at birth of the observed infants was 31.6
weeks PCA (26.29–35.29). The condition of the infant at birth was measured by
Apgar scores at one and five minutes with median scores and ranges 6.11 (1–8) and
7.83 (3–9), respectively. The median birth weight was 1655 g (347–3075).
Participants were predominately female (70 per cent). The majority of infants (95.0
per cent) have some form of complication during the neonatal period. The most
common complication was suspected sepsis (90 per cent) followed by RDS (65 per
cent). Most infants did not have mechanical ventilation (85 per cent); however,
CPAP was more commonly used with fourteen infants (70 per cent) receiving CPAP
for some time.
Thirteen mothers (76.5 per cent) had a singleton pregnancy, three mothers (17.6 per
cent) had twins and one mother (5.9 per cent) had triplets, for a total of seventeen
mothers and twenty infants.
Singleton infants (n = 13) were either delivered by caesarean section (53.8 per cent)
or spontaneous vaginal birth (46.2 per cent) When a multiple birth occurred, the
method of delivery was sometimes different for each infant; however, this only
124
occurred for twin births (Table 30) with all infants in the triplet set delivered by
caesarean section. Differences between the numbers of each type of delivery for twin
pregnancies are due to only one of the twins being ready to breastfeed during the
study period.
Table 29: Sample characteristics
Characteristics Preterm Infants
Frequency % Hospital
Mater Mothers’ Hospital Royal Brisbane and Women’s Hospital
8 12
40 60
Gender Male Female
6 14
30 70
Multiple Birth a No Yes
13 7
65 35
Any neonatal complication No Yes
1 19
5 95
Hours of mechanical ventilation a None, did not receive mechanical ventilation Less than or equal to 48 hours 49 hours to 7 days 8 to 14 days 15 to 28 days Over 29 days
14 3 0 1 2 0
70 15 0 5
10 0
Hours of CPAP a None, did not receive mechanical ventilation Less than or equal to 48 hours 49 hours to 7 days 8 to 14 days 15 to 28 days Over 29 days
6 7 1 2 3 1
30 35 5
10 15 5
a = one missing case; CPAP = continuous positive airway pressure
Table 30: Method of twin deliveries (n = 6)
Type of delivery
Twin 1 (n = 4) Twin 2 (n = 2)
Frequency (%) Frequency (%)
Caesarean Section 2 50 1 50
SVD 2 50 1 50
Breech/Instrumental 0 0 0 0 SVD = Spontaneous vaginal delivery
125
The two most common presenting obstetric reasons for preterm delivery were
preterm labour (40 per cent) followed by pre-labour rupture of membranes (30 per
cent). Other cited reasons were pregnancy induced hypertension (PIH) (ten per cent),
Antepartum haemorrhage (ten per cent), intrauterine growth restriction (five per cent)
and other (five per cent).
At the time of data collection, majority of the infants (80 per cent) had had a previous
attempt at breastfeeding, with eleven of these infants (55 per cent) having attempts in
which they attached and sucked well. About one third of mothers had previously
breastfed an infant before. Infant and mother dyads were predominately assisted to
breastfeed by the clinical nurse (75 per cent) looking after the infant.
5.2.10.4.2 Reliability testing of items in the Preterm Sucking Readiness (PTSR) Scale
Item 1: Baseline Respiration Rate is within Normal Limits of 30–60 Breaths Per
Minute. For infants at both the MMH (n = 8) and RBWH (n = 12), there was 100 per
cent agreement between both raters. In MMH group, there was agreement that all
infants were within normal limits for baseline respiration rate. In the RBWH group,
raters agreed in eleven cases that infants were within normal limits for baseline
respiratory rate and that one infant was not within normal limits.
Item 2: Baseline Pulse Oximetry Level above 90 Per Cent. There was 100 per cent
agreement for this item with both the MMH and RBWH raters agreeing that the
baseline oxygen saturation level was above 90 per cent for all infants.
Item 3: Change in Baseline Pulse Oximetry Level during Handling or Care.
Differences between the raters were noted for this item. In the MMH infant group (n
= 8) there was agreement in five cases (62.5 per cent) between the raters that there
was less than ten per cent difference in pulse oximetry level. There was more
agreement at the RBWH where the raters agreed that there was less than ten per cent
difference in pulse oximetry level for ten cases (83.3 per cent).
126
Item 4: Behavioural State Prior to Feed. At the MMH (n = 8) there was very good
agreement in seven cases (87.5 per cent) with only one case in which there was
disagreement between the raters with one rater recording the infant was asleep and
the other rater recording that the infant was drowsy (Kappa = .800). Similarly, at the
RBWH, there was only one case where there was a disagreement with the same
discrepancy noted with each rater recording that the infant was asleep and the other
rater recording that the infant was drowsy (Kappa = .879).
Item 5: Transition between Behavioural States during Handling. At the MMH (n
= 8), there was 100 per cent agreement for this item. Conversely, at the RBWH (n =
12), there was only agreement in seven cases (58.3 per cent). These disagreements
occurred across this scale, with differences between the two raters documented in the
item categories of ‘briefly alert but gets drowsy’, ‘alert during cares then becomes
drowsy’ and ‘remains alert during care and immediate post period’ recorded.
Item 6: Feeding-Readiness Behaviours. At the MMH (n = 8) there was agreement
between the two raters in five cases (62.5 per cent) demonstrating good agreement
(Kappa = .667). Once more, the RBWH had a lower percentage of agreement than
the MMH with seven out the twelve cases (58.3 per cent) demonstrating moderate
agreement (Kappa = .417).
5.2.11 Summary
Three of the PTSR scale items related to infant behaviour showed significant
differences between the preterm and fullterm infant groups. Only two items, feeding-
readiness behaviours and lowest pulse oximetry during handling, explained the
variance in breastfeeding behaviour. The inter-rater reliability of the PTSR items
ranged from moderate to very good.
In conclusion, the results of this study indicate that age continues to be a factor in
clinicians’ decisions to commence breastfeeding. However, age alone cannot decide
if an infant is ready to engage in breastfeeding behaviour. The importance of
assessing behavioural cues as an indication of breastfeeding readiness in the preterm
127
infant population was demonstrated in this study. Further research is needed to
confirm these findings.
128
129
Chapter 6: Discussion
6.1 Introduction
Preterm infants are commenced breastfeeding when the baby is deemed to be ready
by the neonatal care providers looking after the infant. However, the optimal timing
of the introduction of breastfeeding is unclear. Currently, there is little guidance for
health care professionals to decide when to initiate breastfeeding in the preterm
infant population. This thesis aimed to develop and test an instrument to assess an
infant’s readiness to engage in breastfeeding behaviour. The development of the
PTSR scale (Chapter 4) was based on the results of the first two phases which
informed the development of items as well as phase 3 which allowed the formation
of an agreed list of important criteria to determine breastfeeding readiness in the
preterm infant population. Testing of the tool in phase 4 (Chapter 5) established
preliminary validity and reliability of the PTSR.
In this final chapter, findings of the present study are discussed and conclusions are
drawn. The chapter commences with a discussion of the main findings of the study.
The next section provides an overview of how the chosen theoretical model was used
as a framework to guide this study. This is followed by the strengths and limitations
of this study. Finally, this chapter concludes with the presentation of several
recommendations for clinical practice and further research investigating decisions
surrounding the commencement of breastfeeding in the preterm infant population.
6.2 Main Findings
Statistically and clinically significant differences were shown in the phase 1 chart
audit, in relation to the age when infants reached the feeding milestones of first
attempt; first successful and first exclusive suck feeds. With regards to time taken to
reach feeding milestones, it appears that the lower the GA at birth and birth weight,
the more prolonged the time taken to first feeding attempt. Although this study was
beneficial to inform the development of the PTSR scale by documenting the
differences in the age of infants to reach the feeding milestones and the existence of
130
variables that impact the time taken to breast or bottle feed, it did not explore how
the introduction and transition to breast or bottle feeding is managed by clinical staff.
The 21 items found from the phase 2 focus groups provided an understanding of the
criteria that neonatal care staff use in deciding breastfeeding readiness. While age
continued to be referred to by neonatal care providers in phase 2 to decide when to
commence breastfeeding there were also a number of other criteria mentioned.
These items generally focused around the broad criteria of behavioural cues,
behavioural state and medical stability.
Consensus for six out of the 21 criteria for deciding feeding readiness in the preterm
population was achieved in the first two rounds of the Delphi survey. The third round
did not provide any more agreement and therefore it was decided that this would be
the final round.
Moderate to very good inter-rater reliability was found for the PTSR in the phase 4
testing of the newly developed tool. Preliminary validation of content validity was
achieved with all six PTSR items. However criterion related validity was only
established for two items: feeding readiness behaviours and lowest pulse oximetry
during handling. Construct validity was found for the three items relating to infant
behaviour but not for the physiological stability PTSR items.
6.3 Development of the PTSR Scale
At the end of phase 2, a list of 21 criteria for possible inclusion in the PTSR scale
were brought together to form the Feeding Readiness Indicators in Preterm Infants
Survey. Criteria listed in the Feeding Readiness Indicators in Preterm Infants Survey
were predominately taken from the results of phase 2 of the study although additional
criteria were added from phase 1 such as birth weight as it was found to be
mentioned as part of clinicians’ decision making of feeding readiness in other studies
(Kinneer & Beachy, 1994; Siddell & Froman, 1994). Additionally, a small number
of other criteria were also added to the survey from the literature such as caffeine
usage (Hedberg Nygvist & Ewald, 1999) which although not mentioned by
participants in phase 2 had been discussed in the literature. Due to the small sample
131
size in phase 2 where items were generated, it was felt prudent to be inclusive of all
previously determined, relevant criteria so that the larger and more representative
sample in the phase 3 Delphi could evaluate their possible inclusion in the PTSR
scale. This survey was distributed to participants in the Delphi survey to determine
consensus of items that should be included into the PTSR scale.
There were six out of the 21 items that had a CVI above the cut-off of 78 per cent in
the initial round of the Delphi survey. In round two, five of these items remained
above the CVI cut-off with the only one item “less than five to ten per cent
difference in baseline respiratory rate during handling and cares” that did not
continue to demonstrate agreement between round one and two. This item received
even less support as an indicator in the third round so was not included into the
PTSR scale. The explanation for this is unknown, but could indicate that after the
expert panel reviewed the results of round one, they reconsidered its importance as
an indicator of feeding readiness. Also as there was some lost to follow-up,
supporters of this item’s inclusion may not have participated in round two. An
additional criterion related to changes in baseline pulse oximetry achieved consensus
in round two and remained above the cut-off CVI value in round three so was
included into the final PTSR scale. Consequently, the following criteria were
incorporated into the final PTSR: “baseline respiration rate with normal limits”, “less
than five to ten per cent difference in baseline respiratory rate during handling and
cares”, “baseline pulse oximetry greater than ninety per cent”, “less than five to ten
per cent difference in baseline pulse oximetry level during handling”, “infant quietly
alert just prior to feed time” and “infant showing interest at the breast”.
As seemed to occur sometimes in mixed method studies (Creswell, et al., 2008) there
were contradictory findings for three criteria “ability to move smoothly between
behavioural (arousal) states”, “sucking regularly on dummy, fingers ” and “presence
of rooting reflex”. The quantitative findings of the Delphi survey indicate that these
criteria should not be included in the PTSR scale as their CVI had not reached the
pre-determined cut-off of 78 per cent. On the other hand, the qualitative findings
from the focus groups and open-ended comments in the Delphi survey suggest that
they should be included. As these three criteria were seen as being able to be
merged with items in which consensus was reached, it was decided that these criteria
132
should be included in the PTSR (Onwuegbuzie, et al., 2010). Further support for
their inclusion has come from the literature which includes three surveys
investigating what criteria neonatal care providers use in deciding feeding readiness.
Two of these studies (Pamela Dodrill, et al., 2008; Kinneer & Beachy, 1994)
provided clear support that these items are used in clinical practice with the third
study (Siddell & Froman, 1994) indicating that infant behaviour was considered in
deciding feeding readiness.
6.4 Criteria to Commence Nutritive Suck Feeds
In the development of the PTSR scale many criteria were highlighted as being
potentially important to the decision of when to initiate breastfeeding in the preterm
infant population. Although not all criteria were used as items in the final version of
the PTSR scale they were seen to be able to be categorised as representing three main
criteria: age, medical and physiological stability as well as infant behaviour.
6.4.1 Gestational Age
In phase 1 of this study preterm infants were shown to commence nutritive suck
feeding across a wide range of ages from 29.86 to 41.73 weeks PCA. Similarly in
phase 2, participants’ discussed that while they considered GA in their decision, they
used age only as a guide with circumstances existing in which they may decide to
feed outside their usual GA or age range in which they would routinely commence
feeding. Previous clinical studies have also demonstrated a considerable variation in
the age at which preterm infants are initiated to nutritive suck feeding despite GA
being used as a criterion to determine preterm infants’ feeding readiness (Pamela
Dodrill, et al., 2008; Kinneer & Beachy, 1994). Participants in phase 2 of the current
study also discussed what age criteria if any they used. While 34 weeks was
mentioned as the time when many staff would consider initiating breastfeeding, there
were also other age criteria mentioned including 32 weeks, 30-32 weeks and no
minimum age to commence.
133
The inconsistencies found in the use of age as a criterion in this study have also been
found in the literature. Historically, the GA of 34 weeks PCA is cited as the age
criterion to be used in practice, as previous studies have suggested that it is not until
this time when suck, swallow and breathing coordination is mature enough to allow
safe feeding (Breton & Steinwender, 2008). Alternate studies that have
demonstrated infants as young as 28 weeks PCA as capable of being initiated at the
breast (Hedberg Nygvist, et al., 1999) as well as other studies showing benefits of
earlier introduction of nutritive suck feeding (McCain, et al., 2002; Simpson, et al.,
2002) have appeared to change some neonatal care providers’ perceptions of when to
commence feeding. This was demonstrated by a recent study (Pamela Dodrill, et al.,
2008) of neonatal nurseries (n=29) in Queensland, Australia which found that
neonatal nurseries used the following GAs as criteria to determine feeding readiness:
> 30 weeks, > 32 weeks, 34 weeks, 34 - 36 weeks and no age criteria which was
consistent with the findings of this study.
It was surprising that GA did not achieve consensus in phase 3 for inclusion in PTSR
scale as it was discussed extensively in the focus groups and in the literature
surrounding the development of bottle feeding behaviour. One explanation could be
the fact that other criteria such as physiological stability or infant’s ability to
transition between behavioural states may already indicate the maturity of the infant
that is reflected in GA. Another explanation could be that age is considered but it is
not a strong criterion when making the decision to allow infants to breastfeed.
Participants stated that they would allow an infant to commence suck feeds outside
their stated minimum age or age range if the infant was physiologically stable and
exhibiting feeding readiness behaviours. Furthermore, the inconsistency in defining
the best age to commence nutritive suck feeding amongst neonatal care providers
may in part explain why age as a criterion did not reach consensus in phase 3 of the
study.
6.4.2 Medical and physiological stability
There were a number of physiological and medical factors associated with the time
taken to first attempt at breastfeeding demonstrated in phase 1 that were not seen as
criteria in an instrument to assess breastfeeding readiness by participants in phases 2
134
and 3. For example, in phase 1 of this study, the variation in time taken to when
infants had their first attempt at nutritive suck feeding was explained by GA at birth,
birth weight, and whether they had received a particular intervention such as
mechanical ventilation, CPAP and oxygen therapy. Despite these results being
consistent with the literature (McGrath & Braescu, 2004; Pickler, et al., 2005;
Pridham, et al., 1993) participants did not see them as criteria. However, it is worthy
to note that other explanatory variables described in the literature were not found to
be significant in phase 1 including gender, suspected infection and treatment of
apnoea of prematurity (Hedberg Nygvist & Ewald, 1999).
Criteria to assess preterm infants’ medical and physiological stability was seen by
participants in phases 2 and 3 as a vital part of the neonatal care providers’ decision
making to commence breastfeeding. However, there were differences between
participants on how this criterion was best defined, resulting in many items being
suggested as methods to assess an infant’s medical and physiological stability in
relation to their feeding readiness. The dilemma of determining what physiological
stability means in terms of feeding readiness exists not only in this study and clinical
practice but also in the literature. Numerous authors (Burklow, et al., 2002;
McCain, 2003; McGrath & Braescu, 2004; Pickler, 2005) discuss the need for the
infant to be physiologically or medically stable in general terms to ensure infant
safety during the feed but how this should be defined to determine if an infant is
ready to breastfeed was not clearly indicated with only few studies (Pamela Dodrill,
et al., 2008; Kinneer & Beachy, 1994) providing limited examples of how clinical
staff have defined it in relation to feeding readiness in the past.
6.4.3 Infant Behaviour
A number of ‘feeding readiness’ behaviours have been highlighted in the literature
including: empty sucking, sucking on fingers, sucking on hands, sucking on tongue,
swipes at mouth, hand to mouth, tonguing and rooting (Cagan, 1995; McGrath &
Braescu, 2004; White-Traut, et al., 2002). However, there is limited evidence to
support the use of these feeding readiness behaviours to predict feeding efficiency
(White-Traut, et al., 2005). The lack of strong evidence in the literature to support
the use of feeding readiness behaviours may be reflected in the contradictory
135
findings of phase 3 in which there was support in qualitative data but lack of
sustained consensus of the importance of specific feeding readiness behaviours to the
decision to initiate breastfeeding. The contradictory findings may also be explained
by some participants believing that access to a dummy or fingers will affect how
relevant this criteria is as an indicator of feeding readiness and therefore some
participants may have decided that although it could be used as a sign of
breastfeeding readiness, it is not an important criteria as its absence would not
preclude the commencement of breastfeeding. Interestingly, feeding readiness
behaviours when tested in the final phase of this study were found to be the best
predictor of feeding success.
Another important highlight in phases 2 and 3 was the behavioural state of the infant
prior to the feeding. Infant behavioural state has been positively correlated with
readiness and availability to engage in oral-nipple feeding opportunities (Medoff-
Cooper, et al., 2000). While a positive relationship between level of alertness during
a feed and successful outcome has also been established (McCain, 1997; McGrath &
Medoff-Cooper, 2002) the evidence is inconclusive with McGrath & Medoff-Cooper
(McGrath & Medoff-Cooper, 2002) demonstrating no significant effect for the
robustness of alertness prior to feeding on indicators of nutritive sucking
competence.
6.5 Testing of the PTSR Scale
Testing of the PTSR scale in phase 4 was preliminary but did establish a number of
different types of validity and reliability. However, as described in Chapter 5, not all
common methods of establishing validity and reliability were able to be utilised due
to the nature of the PTSR. For example, performing factor analysis was not
appropriate to the limited number of numbers representing different dimensions of
feeding readiness.
136
6.5.1 Content Validity
Initial content validity of the PTSR scale was first established in the Delphi survey in
which a panel of expert neonatal care providers across Australia and New Zealand
were used to gain consensus on which criteria were most important to deciding
whether an infant was ready to attempt breastfeeding. Moreover, content validity
was further established when a second group of expert neonatal care providers
reviewed the final instrument to ensure that the PTSR scale incorporated the
important aspects to determine breastfeeding readiness before the scale was used in
the observational study.
6.5.2 Construct Validity
The final three items that related to infants’ behaviour demonstrated construct
validity by being able to distinguish between groups with the preterm group
displaying significantly lower behavioural states prior to the feed, less stability in
transitioning between behavioural states and less feeding readiness behaviours than
fullterm infants. Conversely, the first three items in the PTSR scale that related to
the physiological stability could not distinguish between preterm and fullterm infants
with both groups of infants demonstrating comparable physiological stability. The
lack of discrimination between the two groups in relation to physiological stability
was not unexpected as it would be a requirement that both groups would need to
demonstrate physiological stability prior to an attempt at breastfeeding. However, as
physiological stability is vital to infants’ safety, it may be a more important indicator
in only those who are not stable to commence feeding. All infants in this current
study were deemed physiologically stable before commencement of breastfeeding.
6.5.3 Criterion-Related Validity
The ability of the PTSR items to establish criterion-related validity was mixed. The
first three items on physiological stability were again poor determinants of the
outcomes of milk intake or breastfeeding behaviour. However, a relationship
between the last three behavioural items and the degree of breastfeeding behaviour
(PIBBS) was established. At the bivariate level, there was a significant, positive
137
relationship between the total PIBBS score and the items of behavioural state and
feeding-readiness behaviours. The items of feeding readiness behaviour and lowest
pulse oximetry during handling accounted for 25.7 per cent of variance in total
PIBBS score. In this multivariable model, the item, feeding readiness behaviours
(9.2 per cent) made the greatest contribution to explain the variance in total PIBBS
score. Interestingly, when GA was added into the model, it did not contribute any
further in explaining breastfeeding behaviour.
In relation to the other outcome variable of milk intake, there was no relationship
found with any item in the PTSR scale. This was unexpected as it was assumed that
if a relationship was established with total breastfeeding behaviour score, a
relationship would also exist with milk intake. This assumption is based on studies
that have shown a strong correlation between breastfeeding behaviour and milk
transfer with test weighing being established as an accurate estimate of milk intake in
volumes greater than 20 millilitres (Meier, Engstrom, Fleming, Streeter, &
Lawrence, 1996; Savenije & Brand, 2006; Scanlon, Alexander, Serdula, Davis, &
Bowman, 2002). The reason for milk intake not being able to detect variation in
feeding readiness indicators in this study is not known. One explanation could be the
volume of milk ingested was too low to be accurately recorded. Very small volumes
of milk may have been too modest to obtain an accurate measurement, due to certain
mothers in the study producing colostrum instead of mature milk. While colostrum
is nutrient rich, it is only produced in small amounts with mothers usually expressing
between a few drops and five millilitres per feed. Another explanation may be that
as data collection occurred during the first attempts at breastfeeding, preterm infants’
ability to suck may have been poor or ineffective. Therefore, it is difficult to
establish in this study whether indeed there was no relationship between milk intake
and the items of the PTSR scale or if fact there was a relationship between the two
but not observed due to measurement issues of the outcome of milk intake.
6.5.4 Inter-Rater Reliability
The PTSR scale was overall found to have good to very good inter-rater reliability
between two research nurses with only one item found to have moderate agreement.
Interestingly, the use of the same item at the other hospital showed 100 per cent
138
agreement between two raters. One explanation for the difference between the two
hospital results may be a lack of understanding regarding that particular item by the
research nurses in the hospital with the lower inter-rater reliability indicating that
further training may be required. Alternatively, it may also reflect that the research
nurses were observing the same infant at a slightly different point in time.
6.6 Strengths and Limitations
The use of a sequential, exploratory mixed method design was found to be a
constructive and pragmatic approach to the development and testing of the PTSR
scale. This design was well suited for instrument development as it allowed the
researcher to first explore the topic, to develop the items and then test these items
through establishing the instrument’s validity and reliability (Creswell, et al., 2008;
Tashakkori & Teddlie, 2003). The major strength of this study design was the use of
both qualitative and quantitative approaches as it provided greater content coverage
and alternate levels of analysis that could not be undertaken if a single method was
used. However, given the paucity of information available to begin to investigate
how neonatal care providers decided feeding readiness, these studies were
preliminary in nature.
This study has made significant, unique contributions to the body of knowledge of
feeding readiness in the preterm infant population. Firstly, this study provided a
greater understanding on the concept of feeding readiness, including what criteria are
currently used by neonatal care providers in clinical practice. Of particular
significance was the ability of this study to determine a list of agreed criteria
considered important to deciding feeding readiness. No study has previously gained
consensus on which criteria are the most important to determine feeding readiness
with previous studies only listing what criteria were being currently used (Pamela
Dodrill, et al., 2008; Kinneer & Beachy, 1994; Siddell & Froman, 1994). Obtaining
an agreed list of criteria was important for the development of the instrument but it
might also provide direction for clinical staff on which criteria they should consider
until extensive testing of the PTSR has been completed. Another contribution of this
study was the development and preliminary testing that is specific to the
139
breastfeeding population with the few instruments that have only recently become
available being conducted in the bottle feeding population.
One limitation of this study was that the decision to commence breastfeeding was
explored only from the neonatal care providers’ perspective and did not take into
account the role of the mother in this process. While mothers may not make the
decision to initially commence breastfeeding, mothers are imperative to the
breastfeeding process and consideration needs to be undertaken to how they may be
involved. Moreover, once breastfeeding has been established mothers could be
taught aspects of the PTSR scale to facilitate better understanding when their infant
is ready to feed. There is limited research on mothers’ use of breastfeeding tools
however; Hedberg et al. (Hedberg Nygvist, et al., 1996) has shown that mothers and
nurses who were assessed exhibited equal degrees of assessment of breastfeeding
behaviour.
Another limitation was the smaller than expected sample size in phase 2 of this
study. Although the focus groups provided an abundance of information in relation
to the criteria that neonatal care providers utilise to determine an infant’s readiness to
commence breastfeeding, it is unknown whether the sample size was sufficient to
reveal all possible criteria. The method of focus groups relies on group discussion to
react and build upon the responses of other participants and with a small sample size
this may not have been fully utilised and therefore there may have been additional
criteria not discussed in these groups. However, this limitation was minimised in the
subsequent study. In phase 3, a larger sample (n=43) were asked whether they had
any additional criteria to the list that was formed after the focus groups (phase 2).
As often seen in instrument development, phase 3 of the research incorporated
repeated expert reviews using the Delphi technique to ensure that items were added,
removed and revised via an informed process (Chang, et al., 2010; Wilson, Koziol-
McLain, Garrett, & Sharma, 2010). This repeated expert review process is not
uncommon in instrument development as it ensures that the final set of items
represent the construct of interest (Onwuegbuzie, et al., 2010). Furthermore, the
qualitative data provided by the expert panel in the open-ended comments section
were used to validate the quantitative findings as well as explore any
140
misinterpretation and possible inclusion of any additional items not revealed during
the literature review and earlier phases of the study (Onwuegbuzie, et al., 2010).
However, the conflicting results of the Delphi survey and focus groups may reflect
that the cut-off value established prior to the commencement of the study may have
been too high.
The final major limitations relate to the testing of the instrument in phase 4. The
introduction of breastfeeding was dependent on clinical care providers to initiate
breastfeeding based on current clinical practice. This was performed to ensure the
safety and wellbeing of the infant; however, it was found that some nursery staff did
not notify research nurses for all possible data collection opportunities. Anecdotally
the research nurses reported that some staff used the criteria of 34 weeks gestation
and would not allow an infant to breastfeed until that point in time. There may have
been some infants who were able to feed at earlier age; however, this is not
completely known. In addition, this study only captured one breastfeeding event so
was unable to document changes in breastfeeding readiness over time or within
infant characteristics. Utilising the instrument over time would have allowed the
researcher to further validate the instrument. Time and funding did not allow
longitudinal data collection to be conducted during this study. Lastly, pre-term
infant test weighing pre and post breastfeed may not be sensitive enough to measure
small milk volumes consumed in the early transition period to breastfeeding.
Although, it may have just been the findings of this study, researchers need to
carefully consider this outcome as a sole measure of feeding success.
6.7 Implications for Research and Practice
6.7.1 Theoretical Implications
Overall, the synactive theory of development was found to be a useful framework to
understand neonatal care providers’ decision making with regards to the initiation of
breastfeeding behaviour. Proponents of the theory have posited that preterm infants
must be able to integrate the five subsystems described in theory to ensure safety and
efficiency during feeding events (McGrath, 2004; Medoff-Cooper & Ray, 1995;
141
Pickler, 2005; Premji, et al., 2004). Consequently, it was expected that the items in
the newly developed PTSR scale would be guided by this theory through the
assessment of all five subsystems: autonomic, motor, state, attentional or interactive
and self-regularity (Als, 1986).
Only four of the five subsystems can be clearly seen as represented in the final items
of the newly developed PTSR scale. The autonomic subsystem was seen as very
important in order to ensure the safety of the infant and there was great consideration
given in both the focus groups and Delphi survey in how best to assess autonomic
stability to engage in breastfeeding with a number of criteria including HR,
respiration rate, temperature and pulse oximetry all put forward for possible
inclusion in the instrument. Measures of the state subsystem were also discussed
extensively in phases 2 and 3 of this study, with participants stating that an infant’s
ability to wake and remain awake at feed times are important to their decision. The
item that assesses the degree to which feeding-readiness behaviours are displayed is
seen as reflecting the attentional or interactive subsystem as it measures how well the
infant is able to communicate their feeding readiness to their caregiver through their
behaviour. The importance of the self-regularity subsystem was not specified in the
PTSR scale as an item, but rather in the wording which asks for some of the items to
be assessed during handling or cares. Assessment of items during handling
illustrates an infant’s ability to meet challenges to their physiological stability with
infants not ready to breastfeed showing greater signs of decompensation. The
importance of the four subsystems highlighted in this thesis is also reflected in the
literature with a number of studies investigating the independent effect of particular
aspects of each subsystem on either feeding readiness or nutritive sucking behaviour
(Chen, Wang, & Chi, 2000; McCain, 2003; McCain, et al., 2005; McGrath &
Braescu, 2004; Thoyre & Carlson, 2003; White-Traut, et al., 2002).
The only subsystem that was not directly included in the PTSR scale was the motor
subsystem. The motor subsystem refers to the infant’s ability to maintain adequate
tone and posture to engage in feeding behaviour (Holditch-Davis, et al., 2003).
There was no clear explanation as to why participants did not view this subsystem as
important enough to be directly represented in the PTSR scale. The simplest
explanation would be that neonatal care providers do not consider the motor
142
subsystem as an important factor to determine feeding readiness. However, as
functioning of the motor subsystem is required to breastfeed, this is unlikely, with a
more likely explanation being that they considered other criteria to have incorporated
a measurement of maturity of the motor subsystem (Holditch-Davis, et al., 2003;
McGrath & Braescu, 2004). For example, feeding-readiness behaviours were seen as
a reflection of the attentional or interactive subsystem; however, motor control would
also be required for the infant to bring their hands up to their mouth and suck to alert
the caregiver of their readiness.
While Al’s synactive theory was found to be valuable to guide the development of
the PTSR scale, it may not reflect all possible criteria required to decide feeding
readiness. This was not seen as a limitation of the theory, as it was not expected to
provide a list of all assessment criteria of preterm infants’ readiness to commence
breastfeeding, but rather guide possible areas of assessment and provide an
understanding as to how infants respond to events such as feeding. It is
recommended that researchers who are interested in understanding factors
influencing the decision to commence breastfeeding consider the synactive theory of
development as a framework to guide their research.
6.7.2 Clinical and Policy Implications
Results from this study confirm that deciding an infant’s readiness to commence
breastfeeding is a complex clinical decision and involves the use of multiple criteria.
This finding may encourage neonatal care providers to look beyond the criteria of
current GA to assess the individual infant in how they are communicating their
readiness to breastfeed. This would hopefully lead to preterm infants being
introduced to breastfeeding in an appropriate and timely manner so that attempts at
breastfeeding have the greatest success.
These results further suggest that once an infant is deemed physiologically stable,
neonatal care providers should focus on infant’s behaviour in their decision to
commence breastfeeding. Although some studies do report that neonatal care
providers do consider an infant’s behaviour (Pamela Dodrill, et al., 2008; Kinneer &
Beachy, 1994; Siddell & Froman, 1994) other studies have shown that decisions to
143
commence were based primarily on established nursery routines, caregiver
preferences and what is convenient for staff at the time (McCain, 2003). Providing
evidence to neonatal care providers on criteria that may successfully determine
feeding readiness may encourage more clinical staff to use these highlighted criteria
in their practice. Furthermore, the provision of criteria within an instrument such as
the PTSR scale may also support neonatal care providers to assess infants’ readiness
in a in a more systematic and consistent manner.
Currently, there is a lack of clear policy or guideline for clinical practice, with most
available guidelines focused on the introduction of breastfeeding in healthy fullterm
infant populations. The development of the PTSR and other such instruments would
assist neonatal care providers to decide when to initiate breastfeeding for preterm
infants that could be incorporated within written policies at not only at institutional
level but also could be incorporated into national and international policies and
guidelines.
Although the PTSR instrument requires further development and testing prior to
being incorporated into a written policy it does highlight the need for a policy to
establish consistent practice surrounding the initiation of breastfeeding in the preterm
infant population. Such research would be a valuable contribution to the evidence
needed for developing policy to not only guide clinical staff of the evidence to
support their decisions but also to ensure more consistent initiation of breastfeeding
practices.
As a relationship between the display of feeding readiness and the degree of
breastfeeding behaviour exhibited has been preliminarily established in this study, it
is suggested that preterm infants and their mothers are supported and provided with
opportunities to engage in feeding readiness behaviour. Strategies that neonatal care
providers may undertake include: providing infants a dummy when awake, permit
infants access to their own hands rather than having them tightly swaddled away, and
encouraging the mother to provide kangaroo care (McGrath, 2004; White-Traut et
al., 2005). While the effect of feeding-readiness behaviour on predicting feeding
success is not clearly understood; it is assumed that providing more opportunities
144
would allow infants to better communicate their readiness to engage in feeding
behaviour.
6.7.3 Recommendations
The findings from this study have highlighted potential areas where
recommendations can be made to enhance and improve future practice and research.
Due to the exploratory nature of this study only provisional recommendations can be
made to practice and policy.
6.7.3.1 Recommendations for Future Practice
• Each neonatal unit should have a written policy to guide clinical staff in
deciding when an infant is ready to breastfeed and to improve consistency in
practice.
• Once medically stable, infants should be routinely assessed for indicators of
feeding readiness until breastfeeding is established.
• The decision to initiate breastfeeding should be based on an assessment of
preterm infant’s readiness to engage in breastfeeding behaviour.
• Preterm infants should be supported and given opportunities to engage in
feeding readiness behaviours.
6.7.3.2 Recommendations for Future Research
• Research is required to further establish the validity and reliability of the
PTSR. Particularly a longitudinal study that examines feeding readiness
along the continuum of the infant’s hospital stay.
• Research is needed that repeats the testing of the PTSR in other settings to
ensure the generalisability of the instrument.
145
• After further testing of the instrument’s validity and reliability, research can
be undertaken to investigate the PTSR’s clinical utility and usefulness in
practice.
• Future studies need to explore the role of the mother in the decision to
commence breastfeeding.
• Future researchers need to carefully consider the use of test weighing as an
estimate of milk intake during the early period of transition to breastfeeding.
6.8 Conclusion
This thesis contributes significantly to understanding how neonatal care providers
make clinical decisions surrounding when to initiate breastfeeding in the preterm
infant population. Although many of the criteria highlighted by neonatal care
providers in this study were reflected in existing work on the initiation and
management of nutritive suck feeding this study also extends current knowledge by
providing a list of agreed criteria to commence breastfeeding. Most importantly, this
study has developed an instrument that is specifically designed to assess a preterm
infant’s readiness to engage in breastfeeding behaviour with no other instrument
currently being available in clinical practice or research.
The use of a sequential, exploratory mixed method design was found to be valuable
in the development and testing of the PTSR scale. This design was able to identify a
number of criteria used by neonatal care providers to make clinical decisions
surrounding the initiation of breastfeeding and consensus was able to be achieved for
the six items comprising the PTSR scale. The validity and reliability of the PTSR
was partially established. Further research to determine whether assessment of a
preterm infant’s readiness to commence breastfeeding is needed to ensure
breastfeeding occurs in a timely manner to maximise success in breastfeeding.
146
147
Bibliography
Als, H. (1986). A synactive model of neonatal behavioral organization: framework for the assessment of neurobehavioral development in the premature infant and for support of infants and parents in the neonatal intensive care environment. Physical & Occupational Therapy in Pediatrics, 6(3/4), 3-53.
Als, H. (1998). Developmental care in the newborn intensive care unit. Current Opinion in Pediatrics, 10(2), 138-142.
Als, H., Butler, S., Kosta, S., & McAnulty, G. (2005). The assessment of preterm infants' behavior (APIB): furthering the understanding and measurement of neurodevelopmental competence in preterm and fullterm infants. Mental Retardation And Developmental Disabilities Research Reviews, 11(1 ), 94-102.
Andrews Espy, K., & Senn, T. E. (2003). Incidence and correlates of breast milk feeding in hospitalized preterm infants. Social Science & Medicine, 57, 1421-1428.
Banning, M. (2008). A review of clinical decision making: models and current research. Journal of Clinical Nursing, 17(2), 187-195.
Bauer, M., Prade, L., Keske-Soares, M., Haëffner, L., & Weinmann, A. (2008). The oral motor capacity and feeding performance of preterm newborns at the time of transition to oral feeding. Brazilian Journal of Medical and Biological Research, 41, 904-907.
Bender, D., & Ewbank, D. (2004). The focus group as a tool for health research: issues in design and analysis.
Bergman, M. M. (2008). The Straw Men of the Qualitative-Quantitative Divide and their influence on Mixed Methods Research. In M. M. Bergman (Ed.), Advances in Mixed Methods Research (pp. 11-22). London: Sage Publications Inc.
Black, K. A., & Hylander, M. A. (2000). Breastfeeding the High Risk Infant: Implications for Midwifery Management. Journal of Midwifery & Women's Health, 45(3), 238-245: 197-201 (242 ref).
Boswell, C., & Cannon, S. (2007). Introduction to nursing research: Incorporating evidence-based practice: Jones & Bartlett Pub.
Breton, S., & Steinwender, S. (2008). Timing introduction and transition to oral feeding in preterm infants: current trends and practice. Newborn & Infant Nursing Reviews, 8(3), 153-159.
148
Buckley, K. M., & Charles, G. E. (2006). Benefits and challenges of transitioning preterm infants to at-breast feedings. International Breastfeeding Journal, 1(1), 13.
Buhler, K. E. B., & Limongi, S. C. O. (2004). [Factors associated to oral feeding transition in preterm infants]. Pro-Fono: Revista De Atualizacao Cientifica, 16(3 ), 301-310.
Burklow, K. A., McGrath, A. M., Valerius, K. S., & Rudolph, C. (2002). Relationship between feeding difficulties, medical complexity, and gestational age. Nutrition in Clinical Practice, 17(6), 373-378.
Burns, Y., Rogers, Y., Neil, M., Brazier, K., Croker, A., Behnke, L., et al. (1987). Development of oral function in pre-term infants. Physiotherapy Practice, 3, 168-178.
Cagan, J. (1995). Feeding readiness behavior in preterm infants. [abstract]. Neonatal Network, 14(2), 82.
Cahill, J. B., & Wagner, C. L. (2002a). Challanges in breastfeeding: neonatal concerns, second of two parts. Contemporary Pediatrics, 19(5), 113-114.
Cahill, J. B., & Wagner, C. L. (2002b). Challenges in breastfeeding: maternal considerations - first of two parts. Contemporary Pediatrics, 19(5), 94-96, 99-106, 108 passim (123 ref).
Casaer, P., Daniels, H., Devlieger, H., De Cock, P., & Eggermont, E. (1982). Feeding behaviour in preterm neonates. Early human development, 7(4), 331-346.
Case-Smith, J., Cooper, P., & Scala, V. (1989). Feeding efficiency of premature neonates. The American Journal of Occupational Therapy, 43(4).
Chang, A., Gardner, G., Duffield, C., & Ramis, M. (2010). A Delphi study to validate an Advanced Practice Nursing tool. Journal of Advanced Nursing, 66(10), 2320-2330.
Chen, C. H., Wang, T. M., Chang, H. M., & Chi, C. S. (2000). The effect of breast- and bottle-feeding on oxygen saturation and body temperature in preterm infants. Journal of human lactation : official journal of International Lactation Consultant Association, 16(1), 21-27.
Chen, C. H., Wang, T. M., & Chi, C. S. (2000). Individualized developmental care in the newborn intensive care unit. Acta Paediatrica Taiwanica = Taiwan Er Ke Yi Xue Hui Za Zhi, 41(3), 119-122.
Chertok, I. R., Schneider, J., & Blackburn, S. (2006). A pilot study of maternal and term infant outcomes associated with ultrathin nipple shield use. Journal of Obstetric, Gynecologic and Neonatal Nursing, 35(2), 265-272.
149
Church, P. T., Keller, C., Gilbert, J., McCourt, M., McAvoy, I., Perrin, E., et al. (2006). Serial Neonatal Oral Motor Assessment Scale as a measure of feeding readiness. Developmental Medicine & Child Neurology, 48, 39-39.
Colaizy, T. T., & Morriss, F. H. (2008). Positive effect of NICU admission on breastfeeding of preterm US infants in 2000 to 2003. Journal of Perinatology, 28(7), 505-510.
Collins, K., Onwuegbuzie, A., & Sutton, I. (2006). A model incorporating the rationale and purpose for conducting mixed methods research in special education and beyond. Learning Disabilities: A Contemporary Journal, 4(1), 67-100.
Cousins, R. (1999). Breast feeding the preterm infant in the special care baby unit: the first feed. Journal of Neonatal Nursing, 5(6), 10, 12-13 (13 ref).
Creswell, J. W. (2009). Research design: Qualitative, quantitative, and mixed methods approaches: Sage Publications, Inc.
Creswell, J. W., Piano Clark, V., & Garrett, V. (2008). Methodological issues in conducting mixed methods research designs. In M. M. Bergman (Ed.), Advances in Mixed Methods Research (pp. 66-85). Thousand Oaks California Sage Publications.
Crosson, D. D., & Pickler, R. H. (2004). An integrated review of the literature on demand feedings for preterm infants. Advances in Neonatal Care, 4(4), 216-225.
Cunha, M., Barreiros, J., Gonçalves, I., & Figueiredo, H. (2009). Nutritive sucking pattern--From very low birth weight preterm to term newborn. Early human development, 85(2), 125-130.
da Costa, S. P., & Bos, A. F. (2008). Sucking and swallowing in infants and diagnostic tools. Journal of Perinatology, 28(4), 247-257.
Daniels, H., Casaer, P., Devlieger, H., & Eggermont, E. (1986). Mechanisms of feeding efficiency in preterm infants. Journal of pediatric gastroenterology and nutrition, 5(4), 593.
Daniels, H., Devlieger, H., Casaer, P., Callens, M., & Eggermont, E. (1986). Nutritive and non-nutritive sucking in preterm infants. Journal of developmental physiology, 8(2), 117.
Delaney, A. L., & Arvedson, J. C. (2008). Development of swallowing and feeding: prenatal through first year of life. Developmental Disabilities Research Reviews, 14(2), 105-117.
Devries, J. I. P., Visser, G. H. A., & Prechtl, H. F. R. (1985). The emergence of fetal behavior: II. Quantitative aspects. [10.1016/0378-3782(85)90174-4]. Early Hum Dev, 12, 99-120.
150
Dodrill, P., Donovan, T., Cleghorn, G., McMahon, S., & Davies, P. S. W. (2008). Attainment of early feeding milestones in preterm neonates. Journal Of Perinatology: Official Journal Of The California Perinatal Association, 28(8), 549-555.
Dodrill, P., McMahon, S., Donovan, T., & Cleghorn, G. (2008). Current management of transitional feeding issues in preterm neonates born in Queensland, Australia. Early Human Development, 84(10), 637-643.
Dowling, D. A. (1999). Physiological responses of preterm infants to breast-feeding and bottle-feeding with the orthodontic nipple. Nursing Research, 48(2), 78-85.
Drenckpohl, D., Dudas, R., Justice, S., McConnell, C., & Macwan, K. (2009). Outcomes From an Oral Feeding Protocol Implemented in the NICU. ICAN: Infant, Child, & Adolescent Nutrition, 1(1), 6.
Fernandez Díaz, P., & Valdebenito, M. (2007). The Transition From Tube to Nipple in the Premature Newborn. Newborn and Infant Nursing Reviews, 7(2), 114-119.
Flint, A., Davies, M. W., Collie, L., & Palmer-Field, K. (2007). Progression from gastric tube feeds to full suckle feeds and discharge home of preterm infants. Australian Journal of Advanced Nursing, 25(1), 44-48.
Frischknecht, K. (2005). [Developmentally determined readiness for drinking]. Kinderkrankenschwester: Organ Der Sektion Kinderkrankenpflege / Deutsche Gesellschaft Für Sozialpädiatrie Und Deutsche Gesellschaft Für Kinderheilkunde, 24(10), 427-430.
Fujinaga, C., de Oliveira Rodarte, M., Amorim, N., Gonçalves, T., & Scochi, C. (2007). An assessment of premature baby readiness for oral feeding: a descriptive study. Revista Salus-Guarapuava-PR. jul./dez, 1(2), 129-137.
Fujinaga, C., Zamberlan, N., Rodarte, M., & Scochi, C. (2007). Reliability of an instrument to assess the readiness of preterm infants for oral feeding. Pró-Fono Revista de Atualização Científica, 19, 143-150.
Furman, L., & Minich, N. (2004). Efficiency of Breastfeeding as Compared to Bottle-Feeding in Very Low Birth Weight Infants.. [Article]. Journal of Perinatology, 24(11), 706-713.
Furman, L., & Minich, N. M. (2006). Evaluation of Breastfeeding of Very Low Birth Weight Infants: Can We Use the Infant Breastfeeding Assessment Tool?. [Article]. Journal of Human Lactation, 22(2), 175-181.
Gélinas, C., Fillion, L., & Puntillo, K. (2009). Item selection and content validity of the Critical-Care Pain Observation Tool for non-verbal adults. Journal of Advanced Nursing, 65(1), 203-216.
151
Gewolb, I. H., Vice, F. L., Schwietzer-Kenney, E. L., Taciak, V. L., & Bosma, J. F. (2001). Developmental patterns of rhythmic suck and swallow in preterm infants. Developmental medicine and child neurology, 43(1), 27-27.
Hack, M., Estabrook, M. M., & Robertson, S. S. (1985). Development of sucking rhythm in preterm infants. Early human development, 11(2), 133-140.
Hafström, M., & Kjellmer, I. (2000). Non-nutritive sucking in the healthy pre-term infant. Early human development, 60(1), 13-24.
Hall, W. A., Shearer, K., Mogan, J., & Berkowitz, J. (2002). Weighing preterm infants before & after breastfeeding: does it increase maternal confidence and competence? The American Journal of Materinal Child Nursing, 27(6), 318-326.
Hanlon, M. B., Tripp, J. H., Ellis, R. E., Flack, F. C., Selley, W. G., & Shoesmith, H. J. (1997). Deglutition apnoea as indicator of maturation of suckle feeding in bottle-fed preterm infants. Developmental Medicine & Child Neurology, 39, 534-542.
Harding, C. (2009). An evaluation of the benefits of non-nutritive sucking for premature infants as described in the literature. Archives of Disease in Childhood, 94(8), 636-640.
Hardy, D., O’Brien, A., & Gaskin, C. (2004). The Delphi prioritisation of New Zealand mental health nursing clinical indicators- A bi-cultural study. Journal of Advanced Nursing, 46, 95-109.
Hasson, F., Keeney, S., & McKenna, H. (2000). Research guidelines for the Delphi survey technique. Journal of Advanced Nursing, 32(4), 1008-1015.
Hawdon, J., Beauregard, N., Slattery, J., & Kennedy, G. (2000). Identification of neonates at risk of developing feeding problems in infancy. Developmental Medicine & Child Neurology, 42(4), 235-239.
Hedberg Nygvist, K. (2005). Breastfeeding support in neonatal care: an example of the integration of international evidence and experience. Newborn and Infant Nursing Reviews, 5(1), 34-48.
Hedberg Nygvist, K., & Ewald, U. (1999). Infant and maternal factors in the development of breastfeeding behaviour and breastfeeding outcomes in preterm infants. Acta-paediatrica Oslo, Norway 1992, 88(11), 1194-1203.
Hedberg Nygvist, K., Rubertsson, C., Ewald, U., & Sjoden, P.-O. (1996). Development of the preterm infant breastfeeding scale (PIBBS): a study of nurse-mother agreement. Journal of Human Lactation, 12(3), 207-214.
Hedberg Nygvist, K., Sjoden, P., & Ewald, U. (1999). The development of preterm infants' breastfeeding behaviour. Early Human Development, 55, 247-264.
Hedberg Nyqvist, K., Farnstrand, C., Edebol Eeg-Olofsson, K., & Ewald, U. (2001). Early oral behaviour in preterm infants during breastfeeding: An
152
electromyographic study. Acta Paediatrica, International Journal of Paediatrics, 90(6), 658-663.
Hedberg Nyqvist, K., Rubertsson, C., Ewald, U., & Sjoden, P. O. (1996). Development of the Preterm Infant Breastfeeding Behavior Scale (PIBBS): a study of nurse-mother agreement. Journal of human lactation : official journal of International Lactation Consultant Association, 12(3), 207-219.
Hill, A. S. (2002). Toward a theory of feeding efficiency for bottle-fed preterm infants. Journal of Theory Construction & Testing, 6(1), 75-81.
Holditch-Davis, D., Blackburn Tucker, S., & VanderBerg, K. (2003). Newborn and Infant Neurobehavioral Development. In C. Kenner & J. W. Lott (Eds.), Comprehensive Neonatal Nursing - A physiologic perspective (Third Edition ed., Vol. Chapter 17, pp. 236-284). USA: Saunders - an imprint of Elsivier Science.
Holditch-Davis, D., Miles, M. S., & Belyea, M. (2000). Feeding and nonfeeding interactions of mothers and prematures. Western Journal of Nursing Research, 22(3), 320-334 (340 ref).
Hook Morris, B., Miller-Loncar, C. L., Landry, S. H., Smith, K. E., Swank, P. R., & Denson, S. E. (1999). Feeding, Medical Factors, and Developmental Outcome in Premature Infants. [Article]. Clinical Pediatrics, 38(8), 451.
Howe, K. R. (1988). Against the quantitative-qualitative incompatibility thesis or dogmas die hard. Educational researcher, 17(8), 10.
Howe, T. H., Sheu, C. F., Hsieh, Y. W., & Hsieh, C. H. (2007). Psychometric characteristics of the neonatal oral-motor assessment scale in healthy preterm infants. Developmental Medicine And Child Neurology, 49(12), 915-919.
Jamieson, L., & Mosel-Williams, L. (2003). Focus group methodology:explanatory notes for the novice nurse researcher. Contempory Nurse, 14(3), 271-280.
Johnson, R., Onwuegbuzie, A., & Turner, L. (2007). Toward a definition of mixed methods research. Journal of Mixed Methods Research, 1(2), 112.
Keeney, S., Hasson, F., & McKenna, H. (2001). A critical review of the Delphi technique as a research methodology for nursing. International Journal of Nursing Studies, 38(2), 195-200.
Keeney, S., Hasson, F., & McKenna, H. (2006). Consulting the oracle: ten lessons from using the delphi technique in nursing research. Journal of Advanced Nursing, 53(2), 205-212.
Kinneer, M. D., & Beachy, P. (1994). Nipple feeding premature infants in the neonatal intensive-care unit: factors and decisions. JOGNN Clinical Issues, 23(2), 105-112.
153
Kirk, A. T., Alder, S. C., & King, J. D. (2007). Cue-based oral feeding clinical pathway results in earlier attainment of full oral feeding in premature infants. Journal of Perinatology, 27(9), 572-578.
Lau, C. (2007). Development of oral feeding skills in the preterm infant. Archives de pédiatrie: organe officiel de la Sociéte française de pédiatrie, 14, S35.
Lau, C., Alagugurusamy, R., Schanler, R. J., Smith, E. O., & Schulman, R. J. (2000). Characterization of the developmental stages of sucking in preterm infants during bottle feeding. Acta Padiatr, 89, 846-852.
Lau, C., Smith, E. O., & Schanler, R. J. (2003). Coordination of suck-swallow and swallow respiration in preterm infants. Acta Paediatrica, 92(6), 721-727.
Laws, P. J., Li, Z., & Sullivan, E. A. (2010 ). Australia’s mothers and babies 2008 Perinatal statistics series no. 24. Cat. no. PER 50. Canberra AIHW.
Lee, T. Y., Lee, T. T., & Kuo, S. C. (2009). The experiences of mothers in breastfeeding their very low birth weight infants. Journal of Advanced Nursing, 65(12), 2523-2531.
Lemons, P. (2001). From gavage to oral feedings: just a matter of time. Neonatal Network, 20(3), 7-14.
Lemons, P. K., & Lemons, J. A. (1996). Transition to breast bottle feedings: The premature infant. 15(2), 126-135.
Lessen, R., & Crivelli-Kovach, A. (2007). Prediction of initiation and duration of breast-feeding for neonates admitted to the neonatal intensive care unit. The Journal of Perinatal & Neonatal Nursing, 21(3), 256.
Lo-Biondo-Wood, G., & Haber, J. (2006). Nursing Research: methods and critical appraisal for evidence-based practice (6th ed.). St Louis: Mosby Elsevier.
Ludwig, S. M., & Waitzman, K. A. (2007). Changing feeding documentation to reflect infant-driven feeding practice. Newborn & Infant Nursing Reviews, 7(3), 155-160.
Lundqvist, C., & Hafstrom, M. (1999). Non-nutritive sucking in fullterm and preterm infants studied at term conceptional age. Acta Paediatrica, International Journal of Paediatrics, 88(11), 1287-1289.
Martell, M., Martínez, G., González, M., & Rosselló, J. L. D. (1993). Suction patterns in preterm infants. Journal of Perinatal Medicine-Official Journal of the WAPM, 21(5), 363-369.
Mason-Wyckoff, M., McGrath, J. M., Griffin, T., Malan, J., & White-Traut, R. (2003). Nutrition: Physiologic basis of metabolism and management of enteral and parenteral nutrition. In C. Kenner & J. Wright Lott (Eds.), Comprehensive neonatal nursing: a physiologic persepective (3rd ed ed., pp. 425-447). St Louis: Saunders.
154
McCain, G. C. (1997). Behavioral state activity during nipple feedings for preterm infants. Neonatal Network, 16(5), 43-47.
McCain, G. C. (2003). An evidence-based guideline for introducing oral feeding to healthy preterm infants. Neonatal Network, 22(5), 45.
McCain, G. C., Fuller, E. O., & Gartside, P. S. (2005). Heart rate variability and feeding bradycardia in healthy preterm infants during transition from gavage to oral feeding. Newborn & Infant Nursing Reviews, 5(3), 124-132.
McCain, G. C., Gartside, P. S., & Greenberg, J. M. (2002). A semidemand feeding protocol reduced time to full oral feeding in healthy preterm infants. Evidence-Based Nursing, 5(3), 74(71)-(72).
McCain, G. C., Gartside, P. S., Greenberg, J. M., & Lott, J. W. (2001). A feeding protocol for healthy preterm infants that shortens time to oral feeding. the Journal of Paediatrics, 139(3), 374.
McGrath, J. M. (2003). Pilot-testing of the preterm infant nipple feeding readiness scale (PINFRS)... proceedings of the Communicating Nursing Research Conference and WIN Assembly, "Responding to Societal Imperatives Through Discovery and Innovation", held April 10-12, 2003, Scottsdale, Arizona. Communicating Nursing Research, 36, 289.
McGrath, J. M. (2004). Feeding. In C. Kenner & J. McGrath (Eds.), Developmental Care of Newborns & Infants: a Guide for Health Professionals (pp. 321-342). St Louis: Mosby.
McGrath, J. M. (2005). Factors related to feeding readiness in early born preterm infants. Communicating Nursing Research, 38, 340-340.
McGrath, J. M., & Braescu, A. V. B. (2004). State of the science: feeding readiness in the preterm infant. Journal of Perinatal & Neonatal Nursing, 18(4), 353-370.
McGrath, J. M., & Medoff-Cooper, B. (2002). Alertness and feeding competence in extremely early born preterm infants. Newborn & Infant Nursing Reviews, 2(3), 174-186.
McLafferty, I. (2004). Focus group interviews as a data collecting strategy. Journal of Advanced Nursing, 48(2), 187-194.
Medoff-Cooper, B. (1991). Changes in nutritive sucking patterns with increasing gestational age. Nursing research, 40(4), 245.
Medoff-Cooper, B., McGrath, J. M., & Bilker, W. (2000). Nutritive sucking and neurobehavioral development in preterm infants from 34 weeks PCA to term. MCN: The American Journal of Maternal Child Nursing, 25(2), 64-70.
Medoff-Cooper, B., McGrath, J. M., & Shults, J. (2002). Feeding patterns of fullterm and preterm infants at forty weeks postconceptional age. Journal of developmental and behavioral pediatrics, 23(4), 231-236.
155
Medoff-Cooper, B., & Ray, W. (1995). Neonatal sucking behaviors. Journal of nursing scholarship, 27(3), 195-200.
Medoff-Cooper, B., Verklan, T., & Carlson, S. (1993). The development of sucking patterns and physiologic correlates in very-low-birth-weight infants. Nursing Reseach, 42(2), 100-105.
Medoff-Cooper, B., Weininger, S., & Zukowsky, K. (1989). Neonatal sucking as a clinical assessment tool: preliminary findings. Nursing Reseach, 38(3), 162-165.
Meier, P. P., Engstrom, J. L., Fleming, B. A., Streeter, P. L., & Lawrence, P. B. (1996). Estimating milk intake of hospitalized preterm infants who breastfeed. Journal of human lactation offical journal of International Lactation Consultant Association, 12(1), 21-26.
Meyer-Palmer, M., & VandenBerg, K. (1998). A closer look at Neonatal Sucking. Neonatal Network, 9(2), 77-79.
Miller, J. L., Sonies, B. C., & Macedonia, C. (2003). Emergence of oropharyngeal, laryngeal and swallowing activity in the developing fetal upper aerodigestive tract: an ultrasound evaluation. Early Human Development, 71, 61-37.
Mizuno, K., & Ueda, A. (2003). The maturation and coordination of sucking, swallowing, and respiration in preterm infants. The Journal of Pediatrics, 142(1), 36-40.
Monterosso, L., Kristjanson, L., Sly, P. D., Mulcahy, M., Holland, B. G., Grimwood, S., et al. (2005). The role of the neonatal intensive care nurse in decision making: Advocacy, involvement in ethical decisions and communication. International journal of nursing practice, 11(3), 108-117.
Morrison-Beedy, D., Cote-Arsenault, D., & Fischbeck Feinstein, N. (2001). Maximising results withh focus groups: moderator and analysis issues. Applied Nursing Research, 14(1), 48-53.
Mouradian, L. E., Als, H., & Coster, W. J. (2000). Neurobehavioral functioning of healthy preterm infants of varying gestational ages. Journal Of Developmental And Behavioral Pediatrics: JDBP, 21(6 (Print)), 408-416.
National Health and Medical Research Council, Australian Research Council, & Austraian Vice-Chancellors' Committee. (2007). National Statement of Ethical Conduct in Human Research. Retrieved from www.nhmrc.gov.au.
Neiva, F. C. B., Leone, C., & Leone, C. R. (2008). Non-nutritive sucking scoring system for preterm newborns. Acta Paediatrica (Oslo, Norway: 1992), 97(10), 1370-1375.
Onwuegbuzie, A., Bustamante, R., & Nelson, J. (2010). Mixed Research as a Tool for Developing Quantitative Instruments. Journal of Mixed Methods Research, 4(1), 56.
156
Owen-Smith, A., Sterk, C., McCarty, F., Hankerson-Dyson, D., & DiClemente, R. (2010). Development and Evaluation of a Complementary and Alternative Medicine Use Survey in African-Americans with Acquired Immune Deficiency Syndrome. The Journal of Alternative and Complementary Medicine, 16(5), 569-577.
Pickler, R. H. (2004). A model of feeding readiness for preterm infants. Neonatal Intensive Care, 17(4), 31-36.
Pickler, R. H. (2005). A model of feeding readiness for preterm infants. Neonatal Intensive Care, 18(4), 17-22.
Pickler, R. H., Best, A., & Crosson, D. (2009). The effect of feeding experience on clinical outcomes in preterm infants. Journal of Perinatology, 29(2), 124-129.
Pickler, R. H., Best, A. M., Reyna, B. A., Wetzel, P. A., & Gutcher, G. R. (2005). Prediction of feeding performance in preterm infants. Newborn & Infant Nursing Reviews, 5(3), 116-123.
Pickler, R. H., & Reyna, B. A. (2003). A descriptive study of bottle-feeding opportunities in preterm infants. Advances in Neonatal Care, 3(3), 139-146.
Pineda, R. G. (2011). Predictors of breastfeeding and breastmilk feeding among very low birth weight infants. Breastfeeding Medicine, 6(1), 15-19.
Pinelli, J., & Symington, A. (2005). Non-nutritive sucking for promoting physiologic stability and nutrition in preterm infants. Cochrane Database Of Systematic Reviews (Online)(4), CD001071.
Pinelli, J., Symington, A., & Ciliska, D. (2002). Nonnutritive sucking in high-risk infants: benign intervention or legitimate therapy? JOGNN Clinical Issues, 11(5).
Pinelli, J., & Symington Amanda, J. (2005). Non-nutritive sucking for promoting physiologic stability and nutrition in preterm infants. from http://www.mrw.interscience.wiley.com/cochrane/clsysrev/articles/CD001071/frame.html
Plummer-D'Amato, P. (2008). Focus group methodology Part 1: Considerations for design. International Journal of Therapy and Rehabilitation 15(2), 69-73.
Polit, D., & Beck, C. (2008). Nursing research: Generating and assessing evidence for nursing practice: Lippincott Williams & Wilkins.
Polit, D., Beck, C., & Owen, S. (2007). Is the CVI an acceptable indicator of content validity? Appraisal and recommendations. Research in Nursing & Health, 30(4), 459-467.
Polit, D. F., & Beck, C. T. (2004). Nursing research: principles and methods (7th ed.). Philadelphia: Lippincott, Williams & Wilkins.
157
Polit, D. F., & Beck, C. T. (2006). The content validity index: are you sure you know what's being reported? Critique and recommendations. Research in Nursing & Health, 29(5), 489-497.
Portney, L. G., & Watkins, M. P. (2000). Foundations of clinical research: applications to practice (2nd ed.). Upper Saddle River, New Jersey: Prentice-Hall Inc.
Powell, C. (2003). The Delphi technique: myths and realities. Journal of Advanced Nursing, 41(4), 376-382.
Powers, N. G., Bloom, B., Peabody, J., & Clark, R. (2003). Site of care influences breastmilk feedings at NICU discharge. Journal of perinatology, 23(1), 10-13.
Premji, S. S., McNeil, D. A., & Scotland, J. (2004). Regional neonatal oral feeding protocol: changing the ethos of feeding preterm infants. Journal of Perinatal & Neonatal Nursing, 18(4), 371-384.
Premji, S. S., Paes, B., Jacobson, K., & Chessell, L. (2002). Evidence-based feeding guidelines for very low-birth-weight infants. Advances in Neonatal Care (Elsevier Science), 2(1), 5-18.
Premji, S. S. J. (2000). Feeding practice guidelines for premature infants less than 1500 grams: efficacy and safety. Unpublished Ph.D., McMaster University (Canada).
Pridham, K. F., Sondel, S., Chang, A., & Green, C. (1993). Nipple feeding for preterm infants with bronchopulmonary dysplasia. JOGNN clinical studies, 22(2), 147-155.
Ramezani Badr, F., Nasrabadi, A. N., Yekta, Z. P., & Taleghani, F. (2009). Strategies and criteria for clinical decision making in critical care nurses: a qualitative study. Journal of Nursing Scholarship, 41(4), 351-358.
Riordan, J., Gill-Hopple, K., & Angeron, J. (2005). Indicators of effective breastfeeding and estimates of breast milk intake. Journal of Human Lactation, 21(4), 406-412.
Rommel, N., De Meyer, A., Feenstra, L., & Veereman-Wauters, G. (2003). The complexity of feeding problems in 700 infants and young children presenting to a tertiary care institution. Journal of pediatric gastroenterology and nutrition, 37(1), 75.
Ross, E. S., & Browne, J. V. (2002). Developmental progression of feeding skills: an approach to supporting feeding in preterm infants. Seminars In Neonatology: SN, 7(6), 469-475.
Rossarolla, C., Menon, M., Scochi, C., & Fujinaga, C. (2009). Discriminatory validity of an instrument for evaluating preterm newborns' readiness for oral feeding. Revista da Sociedade Brasileira de Fonoaudiologia, 14, 106-114.
158
Savenije, O. E. M., & Brand, P. L. P. (2006). Accuracy and precision of test weighing to assess milk intake in newborn infants. Archives of Disease in Childhood -- Fetal & Neonatal Edition, 91(5), F330-332.
Scanlon, K. S., Alexander, M. P., Serdula, M. K., Davis, M. K., & Bowman, B. A. (2002). Assessment of infant feeding: the validity of measuring milk intake. Nutrition Reviews, 60(8), 235-251.
Schell, K. (2006). A delphi study of innovative teaching in bacclaureate nursing education. Journal of Nursing Education, 45(11), 439-448.
Shaker, C. S. (1999). Nipple feeding preterm infants: an individualized, developmentally supportive approach. Neonatal Network, 18(3), 15-22.
Shaker, C. S., & Woida, A. M. W. (2007). An evidence-based approach to nipple feeding in a Level III NICU: nurse autonomy, developmental care, and teamwork. Neonatal Network, 26(2), 77-83.
Sheehan, A., Schmied, V., & Barclay, L. (2010). Complex decisions: theorizing women’s infant feeding decisions in the first 6 weeks after birth. Journal of Advanced Nursing, 66(2), 371-380.
Siddell, E. P., & Froman, R. D. (1994). A national survey of neonatal intensive-care units: criteria used to determine readiness for oral feedings. JOGNN: Journal of Obstetric, Gynecologic, & Neonatal Nursing, 23(9), 783-789.
Simpson, C., Schanler, R. J., & Lau, C. (2002). Early introduction of oral feeding in preterm infants. Pediatrics, 110(3), 517-522.
Sisk, P., Quandt, S., Parson, N., & Tucker, J. (2010). Breast Milk Expression and Maintenance in Mothers of Very Low Birth Weight Infants: Supports and Barriers. Journal of Human Lactation, 26(4), 368.
Snyder-Halpern, R., Thompson, C., & Schaffer, J. (2000). Comparison of mailed vs Internet applications of the Delphi technique in clinical informatics research. Paper presented at the Americaqn Medical Informatics Association Annual Symposium.
Spence, K. (1998). Ethical issues for neonatal nurses. Nursing ethics, 5(3), 206.
Spence, K. (2000). The best interest principle as a standard for decision making in the care of neonates. Journal of Advanced Nursing, 31(6), 1286-1292.
St John, E. B., Nelson, K. G., Cliver, S. P., Bishnoi, R. R., & Goldenberg, R. L. (2000). Cost of neonatal care according to gestational age at birth and survival status. Am J Obstet Gynecol, 182(Number 1, Part 1), 170-175.
Sundseth Ross, E., & Browne, J. V. (2002). Developmental progression of feeding skills:an approach to supporting feeding in preterm infants. Seminars in Neonatology, 6(5), 417-424.
159
Sweet, L. (2008a). Birth of a very low birth weight preterm infant and the intention to breastfeed 'naturally'. Women and Birth, 21(1), 13-20.
Sweet, L. (2008b). Expressed breast milk as' connection'and its influence on the construction of'motherhood'for mothers of preterm infants: a qualitative study. International Breastfeeding Journal, 3(1), 30.
Tashakkori, A., & Teddlie, C. (2003). The past and futre of mexed mthods research from data triangulation to mixed model designs. In A. Tashakkori & C. Teddlie (Eds.), Handbook of Mixed Methods in social and behavioural research (pp. 671-703). Thousand Oaks California: Sage Publications.
Tashakkori, A., & Teddlie, C. (2008). The Evolution of Mixed Methods Research. In V. Piano Clark & J. W. Creswell (Eds.), The Mixed Methods Reader (pp. 5-7). Thousand Oaks California: Sage Publications.
Teddlie, C., & Tashakkori, A. (2003). Major issues and controversies in the use of mixed methods in the social and behavioural sciences. In A. Tashakkori & C. Teddlie (Eds.), Handbook of mixed methods in social and behavioral research (pp. 3-51). Thousand Oaks California: Sage Publications.
Thoyre, S. (2009). Dynamic Early Feeding Skills: An Observational System for Coding the Dynamics of Early Infant Feeding. Advances in Neonatal Care, 9(4), 188.
Thoyre, S. M. (2003). Developmental transition from gavage to oral feeding in the preterm infant. Annual Review of Nursing Research, 21, 61-92.
Thoyre, S. M., & Brown, R. L. (2004). Factors contributing to preterm infant engagement during bottle-feeding. Nursing Research, 53(5), 304-313.
Thoyre, S. M., & Carlson, J. (2003). Occurrence of oxygen desaturation events during preterm infant bottle feeding near discharge. Early Human Development, 72(1), 25-36.
Thoyre, S. M., Shaker, C. S., & Pridham, K. F. (2005). The early feeding skills assessment for preterm infants. Neonatal Network, 24(3), 7-16.
Vergara, E. R., & Bigsby, R. (2004). Developmental and Theraputic Interventions in the NICU (1 ed.). Baltimore: Paul H Brooks Publishing Co.
Waltz, C., Strickland, O., & Lenz, E. (2010). Measurement in nursing and health research: Springer Publishing Company.
Weaver, K. A., & Anderson, G. C. (1988). Relationship between integrated sucking pressures and first bottle-feeding scores in premature infants. Journal Of Obstetric, Gynecologic, And Neonatal Nursing: JOGNN / NAACOG, 17(2), 113-120.
Wheeler, J., Chapman, C., Johnson, M., & Langdon, R. (2000). Feeding outcomes and influences within the Neonatal Unit. International journal of nursing practice, 6(4), 196-206.
160
White-Traut, R. C., Berbaum, M. L., Lessen, B., McFarlin, B., & Cardenas, L. (2005). Feeding readiness in preterm infants: the relationship between preterm behavioral state and feeding readiness behaviors and efficiency during transition from gavage to oral feeding. MCN: The American Journal of Maternal Child Nursing, 30(1), 52-59.
White-Traut, R. C., Nelson, M. N., Silvestri, J. M., Vasan, U., Patel, M., & Cardenas, L. (2002). Feeding readiness behaviors and feeding efficiency in response to ATVV intervention... auditory, tactile, visual and vestibular. Newborn & Infant Nursing Reviews, 2(3), 166-173.
Willgerodt, M. A. (2003). Using focus groups to develop culturally relevant instruments. Western Journal of Nursing Research, 25(7), 798-814.
Willis, K., Green, J., Daly, J., Williamson, L., & Bandyopadhyay, M. (2009). Perils and possibilities: achieving best evidence from focus groups in public health research. Australian And New Zealand Journal Of Public Health, 33(2), 131-136.
Wilson, D., Koziol-McLain, J., Garrett, N., & Sharma, P. (2010). A hospital-based child protection programme evaluation instrument: a modified Delphi study. International Journal for Quality in Health Care.
161
Appendices
Appendix A: Summary of Studies in Literature Review
Reference Study Design Sample Outcome Measures
Results
Burns, Roger, Neil, Brazier, Croker, Behnke & Tudehope (1987)
Non-experimental, descriptive study
Australia One group n = 110 infants GA at birth 25–37.5 weeks Male = 51, Female = 59 BW = 600–1499 g GA at start = when infant stabilised, observed weekly
Adaptive and reactive reflexes
Responses could be stimulated several weeks before reaching a level of consistently Development of responses not significantly altered by age at birth, BW, mechanical ventilation or sex. However, presence of IVH significantly affected rooting, non-nutritive sucking and suck-swallow
Miller, Sonies & Macedonia (2003)
Non-experimental descriptive study
USA Two group GP 1 Healthy foetuses: 15–38 weeks GA n = 62 GP 2 At risk for postnatal feeding problems n = 7
Biometric measures of oral, lingual, pharyngeal and laryngeal structures using four sonogram scans
Ingestive behaviour emerged in a sequence of basic to complex movement patterns. Jaw and lip movement progressed from simple mouth opening to repetitive open–close movements. Lingual movement increased in complexity from simple forward thrusting and cupping to anterior–posterior motions. Laryngeal movements varied from shallow flutter-like movements along the lumen to more complex and complete adduction abduction patter. Foetal swallowing primarily occurred in the presence of concomitant oral-facial stimulatory activity. Significant variations in the form and function of the ingestive system occurred in comparison of gestational age-matched controls to at risk cases.
162
Reference Study Design Sample Outcome Measures
Results
Hack, Estabrook & Robertson (1985)
Longitudinal, repeated measures
USA One group Preterm infant n = 6 GA at birth = 28.6 weeks (mean) BW 1110 g (mean) GA at assessment 30–35 weeks; No major co-morbidities
Non-nutritive sucking
Rhythmical sucking bursts were recorded at 28 weeks GA in one infant and by 31–33 weeks GA in the other five. The number of burst per minute increased with age, while the duration of each burst was stable. The pause between bursts decreased. Sucking pace within bursts increased with age, which resulted in an increase in the overall rate of sucking.
Hafstrom & Kjellmer (2000)
Longitudinal correlation study
Sweden One group Preterm infants = 58 GA at birth = 26–35 weeks Low risk healthy infants GA at commencement = ?
Non-nutritive sucking
The typical NNS pattern with bursts of sucking activity separated by pauses was recorded in all infants studied and was already present before 30 weeks gestation. A gradual change over time of their NNS was seen. With increased maturation the sucking activity, sucking frequency, amplitude and burst duration all increased, while the variability of the sucking frequency and the duration of the intervals between bursts declined. PMA was the dominant predictor of the result for both genders, state of activity and weight also influence it to some extent. Girls had more sucking activity and higher sucking frequency than boys. State of activity affected the stability of the rhythm. The weight of the infant influenced both sucking activity and duration of separate bursts.
163
Reference Study Design Sample Outcome Measures
Results
Lundqvist & Hafstrom (1999)
Non-experimental, descriptive study
Sweden Two groups Preterm group n = 35 Fullterm group n = 38 Fullterm infants – commenced study on Day 3 and had used no pacifiers Preterm infants – studied at same PMA as fullterm group – some pacifier use GA at birth for preterm group = ? Co-morbidities of preterm group = ?
Non-nutritive sucking
The preterm group showed a distinct NNS pattern with a higher frequency and lower amplitude. In the preterm group, gender difference was observed with the girls having higher frequency used during early childhood. Experience, gender, maturity and level of tension are suggested as explanatory factors for differences in NNS patterns.
Pinelli & Symington (2005)
Systematic review
21 studies, 15 RCTS
Non-nutritive sucking
NNS was found to decrease significantly the length of hospital stay in preterm infants. The review did not demonstrate a consistent benefit of NNS with respect to the other variables of weight gain, energy intake, heart rate, oxygen saturation, intestinal transit time, age at full oral feeds and behavioural state. However, the review did identify other positive clinical outcomes of NNS including transition from tube to bottle feeds and better bottle-feeding performance. No negative outcomes were recorded.
Medoff-Cooper (1991)
Non-experimental correlation design
USA One group preterm infants n = 66; GA at birth 31 weeks (mean) BW:1379 g (mean) GA at assessment: 32–36 weeks
Nutritive sucking
A modest trend for mean maximum pressure appeared at 34 weeks. No trend was observed for duration of suck or the duration of suck burst. Number of sucks per sucking burst demonstrated somewhat of a decreasing trend from 32 to 36 weeks.
164
Reference Study Design Sample Outcome Measures
Results
Gewolb, Vice, Schwietzer-Kenney, Taciak & Bosma (2001)
Non-experimental correlation study
USA One group Preterm healthy infants n = 20 GA at birth = 26–33 weeks Inclusion criteria: infants were not on oxygen at time of study, did not have significant BPD, or IVH >grade 2. Exclusion criteria: drug abuse, concurrent sepsis and congenital craniofacial abnormality. GA at time of study = 32.1–39.6 weeks Infants studied weekly
Nutritive sucking
The percentage of sucks aggregated into runs increased over time and correlated with increasing PMA. The length of suck runs also correlated significantly with PMA. The stability of sucking rhythm was also directly correlated with increasing suck rate. None of these measures was correlated with PNA. Similarly, increasing PMA but not PNA correlated with a higher percentage of swallows in runs. Stability of swallow rhythm did not change significantly from 32 to 40 weeks PMA.
Hedberg Nygvist, Sjoden & Ewald (1999)
Non-experimental, descriptive study, matched-age pairs
Sweden One group n = 71 infants GA at birth – 26.7–35.9 No BPD or IVH (grade III or IV) BW – 1013–3130 g GA at start = when infant stabilised
Nutritive sucking
Efficient rooting, areolar grasp and latching on observed repeated burst of ≤10 bursts at 28 weeks; Maximum burses of ≥30 sucks at 32 weeks. Nutritive sucking (≥5 ml) at 30.6 weeks.
165
Reference Study Design Sample Outcome Measures
Results
Medoff-Cooper, McGrath & Shults (2002)
Non-experimental, matched groups
USA Three groups (2 preterm + 1 fullterm) Gp 1: GA at birth 24–29 weeks, n = 75 Gp 2: GA at birth 30–32 weeks, n = 68 Gp 3: GA at birth 38–42 weeks, n = 70 Variable numbers of co-morbidities present; Preterm infants were assess at 40 weeks GA, healthy fullterm infants were assessed around the time of birth
Nutritive sucking
There were significant difference amongst infants of different GAs in several of the sucking variables including number of burst, intersuck interval, sucks per burst, suck width, intersuck width, mean maximum pressure and intersuck width/interburst width.
Pickler, Best, Reyna, Wetzel & Gutcher (2005)
Longitudinal, non-experimental study
USA One group of preterm infants n = 95 GA at birth = <32 weeks PMA; GA at assessment = ? Inclusion criteria: had no known gastrointestinal, craniofacial, cardiovascular, neurological or muscular defects.
Nutritive sucking
Sucking activity (number of sucks, sucks/burst and sucks/minute) was predicted by morbidity, maturity, feeding experience and pre-feeding behaviour. Experience at oral feeding had the great effect on changes in the number of sucks, suck/burst and sucks/minute.
Casear, Daniels, Devlieger, De Cock, Eggermont (1982)
Non- experimental, correlation, study
Belgium One group Preterm infants n = 100 GA at birth = 25–37 weeks GA at assessment = ? Co-morbidities? BW = <2000 g
Nutritive sucking – feeding efficiency
Feeding efficiency was greater in infants >34 weeks GA than in those <34 weeks. There was a significant correlation between feeding efficiency and the duration of feeding experience at most ages between 32 and 37 weeks GA.
166
Reference Study Design Sample Outcome Measures
Results
Martell, Martínez, González & Rosselló (1993)
Matched groups; Non-experimental correlational design with repeated measures
Uruguay Two subgroups Gp 1 = 16 breastfed preterm GA at birth = 28–36 weeks GA at assessment = 32–36 weeks Gp 2 = 46 bottle-fed preterm GA at birth 30–35 weeks GA at assessment: 34–38 weeks Co-morbidities = ?
Nutritive sucking – feeding efficiency
The velocity of milk extraction during sucking varied with GA. It was uniform at lower ages, then it became faster in the first minutes and at the 36th week, it was very similar to that of mature neonates. The time elapsed between birth and the commencement of suck feeds was longer in the preterm infants group born at lowers GAs for both groups: breastfed and bottle-fed.
Daniels, Casaer, Devlieger & Eggermont (1986)
Non-experimental, correlational design
Belgium One group 18 preterm infants GA at birth 29–33 weeks BW ranged from 1020 to 2450 g Postnatal age = 0–8 weeks) GA at assessment?
Nutritive sucking – feeding efficiency
Differences in feeding efficiency could not be explained by difference in sucking rate. Quick efficient feeding was characterised by long sucking birth and large amount of milk intake during each sucking movement. Low inefficient feeding was characterised by short, sucking bursts and by a small amount of milk intake with each sucking movement.
Weaver & Anderson G (1988)
Non-experimental correlation design
United States One group 30 preterm infants GA at birth 28 – 36 weeks BW ranged from 1230- 2020 g GA at assessment ? Assessments occurred mean 25 days
Nutritive sucking – feeding efficiency and feeding assessment
Sucking pressures were measured using an electronic suckometer. Feeding ability was measured using a complex and simple research scales. The correlation between the feeding scales was found to be statistically significant but the correlations between feeding scales and sucking pressures were not significantly correlated.
167
Reference Study Design Sample Outcome Measures
Results
Lau, Alagugurusamy, Schanler, Smith & Schulman (2000)
Non-experimental, correlation design
USA Two groups Gp 1: 72 infants GA at birth = 26–29 weeks Mean BW = 1028 Healthy infants Excluded IVH, NEC, major congenital abnormalities, persistent oxygen requirement Gp 2 = 10 healthy fullterm infants GA at start = when infant medically stabilised
Nutritive sucking and feeding efficiency
Significant positive correlations were observed between the five stages of sucking and PMA, the defined feeding outcomes and the number of daily oral feedings. Overall transfer and rate of transfer were enhanced when infants reached the more mature stages of sucking.
Medoff-Cooper, B., Weininger, S. & Zukowsky (1989)
Non-experimental, matched groups
USA Two groups preterm and fullterm Gp 1 preterm = 44 GA at birth: 30.9 weeks (mean) No congenital malformations Other co-morbidities Gp 2: Healthy fullterm infants n = 42
Nutritive sucking
The maximum pressure of the healthy fullterm neonate was higher than the maximum pressure of the preterm infant. Fullterm infants also consumed more formula per suck. In addition, they had more sucks per episode and maintained the pressure longer for a wider suck width.
Medoff-Cooper, McGrath & Bilker (2000)
Non-experimental descriptive correlation design was used
USA One group Preterm infants n = 66 GA at birth = 24–34 Mean BW = 1379 GA at commencement = 34 weeks and 40 weeks Had a range of issues associated with prematurity
Nutritive sucking
There were significant differences in number of sucks, intensity of sucking pressures, average time between sucks and average time between sucking bursts. With maturation, it was noted that the preterm infants were significantly more alert during the sucking protocol from 34 to term. Yet, infant sucking parameters at 34 weeks PCA and term were not significantly related to the Brazelton Neonatal Behavioural Assessment Scale or to the infant’s state.
168
Reference Study Design Sample Outcome Measures
Results
Mizuno & Ueda (2003)
Longitudinal correlational study
Japan One group preterm group n = 24 GA at birth 28–31 weeks No major co-morbidities, neonates studied weekly from 32–36 weeks
Nutritive sucking and SSB coordination
The sucking efficiency significantly increased between 34 and 36 weeks PCA and exceeded 7 ml/min at 35 weeks. There were significant increases in sucking pressure and frequency as well as in duration between 33 and 36 weeks. Although swallowing occurred mostly during pauses in respiration at 32–33 weeks, after 35 weeks swallowing usually occurred at the end of inspiration.
Lau, Smith & Schanler (2003)
Non-experimental, matched groups
USA Two groups Gp 1 healthy preterm infants n = 12 GA at birth = <30 weeks GA at assessment = 34 and 36.7 weeks; no major co-morbidities Gp 2: fullterm infants n = 8
Nutritive sucking and SSB coordination
The rate of milk transfer in the preterm group increased over time and was correlation with mean bolus size and swallowing frequency. Mean bolus size was not correlated with average bolus size and swallowing frequency. Average bolus size was not correlated with swallowing frequency. Preterm infants swallowed preferentially at different phases of respiration than those of their fullterm counterparts.
Daniels, Devlieger, Casaer, Callens & Eggermont (1986)
Non-experimental, correlation design
Belgium One group Preterm infant n = 9 GA at birth = 28–33 weeks GA at assessment = ?
Non-nutritive sucking Nutritive sucking
During nutritive sucking, sucking burst was long
Hedberg Nyqvist, Farnstrand, Edebol Eeg-Olofsson & Ewald (2001)
Non-experimental, correlation design
Sweden One group Preterm infants n = 26 Exclusion criteria: BPD, IVH grade II or IV; any serious illness or abnormality GA at birth =? mean = 35 SD= 1.5 GA assessment = ?
Nutritive sucking
The agreement between EMG data and direct observation was high. The median for mean number of sucks per burst was 8; the longest burst was 28 sucks. The range in mean sucking was 1–1.8 sucks. Suck duration ranges from 0.6 to 1.1 seconds. There was a considerable variation between infants in the extent of mouthing. No association with maturational level appeared for any of the components of behaviour.
169
Reference Study Design Sample Outcome Measures
Results
Hedberg Nygvist & Ewald (1999)
Secondary analysis Non-experimental correlation design
Sweden One group n = 71 infants GA at birth – 26.7–35.9 No BPD or IVH (grade III or IV) BW – 1013–3130 g GA at start = when infant stabilised
Nutritive sucking
Variables associated with efficient infant performance included higher birth weight, less need of ventilator and oxygen treatment, higher haemoglobin level, absence of bottle feeding, no need of apnoea treatment and no suspicion of infection. Infants with a short gestation period achieved full breastfeeding at low postmenstrual and high postnatal age. Low gestational age at birth was associated with early emergence of efficient breastfeeding behaviour and a high incidence of full breastfeeding.
Furman & Minich (2004)
Secondary analysis
USA Two groups BF group = n = 35 Bottle group n = 70 BW = VLBW <1500 g GA birth less than 33 weeks Exclusion criteria: multiple births, positive drug screen, major congenital anomaly or intrauterine infection, maternal social factors such as custody dispute
Nutritive sucking
Breastfeeding as compared with bottle feeding, infants took in smaller volumes, fed less efficiently and spent less time with sucking bursts. For breastfed infants, birth and neonatal factors and prior maternal breastfeeding experience were not associated with feeding efficiency or behaviours.
170
Reference Study Design Sample Outcome Measures
Results
McCain (1997) Secondary analysis correlational study
USA One group Preterm infants = 20 GA at birth = ? GA at assessment = ? Co-morbidities = ?
Nutritive sucking
Nine of the 20 infants were able to successfully nipple their four nipple feedings, the other eleven infants were unable to successfully complete their nipple feedings. The nine successful feeders exhibited significant more awake behaviour during deeding than unsuccessful feeders. The unsuccessful feeders spent 53% of their feeding in sleep state; successful feeders were in sleep states only 30% of their feeding time.
Cagan (1995) Non-experimental descriptive study
USA One group Preterm infants n = 26 GA at birth = ? GA at time of study = 32–33 weeks
Nutritive sucking
Preterm infants between 32 and 33 weeks PCA seem to be able to provide cues that would help caregivers plan individualised schedules. Infants who were fed when they displayed cues at their scheduled feeding had hiccups less of the time during feeding and they received 18.7% fewer nasal-gavage feeding and 18.7% more oral feeding than those infants who did not display cues at their scheduled feeding time.
White-Traut, Nelson, Silvestri, Vasan, Patel & Cardenas (2002)
Experimental design
USA Two groups Gp 1 n = 10 (control) Gp 2 n = 12 (experimental) GA at birth 29–33 weeks BW mean = 1135–2155 g GA at assessment 31–38 weeks
Nutritive sucking
Tested the effects of pre-feeding intervention. Infants in the experimental group were more alert after the intervention and showed more feeding readiness behaviours during the intervention for five out of the eight behaviours. A trend toward decreased feeding time was noticed towards the experimental group.
171
Reference Study Design Sample Outcome Measures
Results
White-Traut, Berbaum, Lessen, McFarlin & Cardenas (2005)
Secondary analysis Predictive, correlation study
USA Two groups Gp 1 n = 10 (control) Gp 2 n = 11 (experimental) GA at birth 29–33 weeks BW mean = 1135–2155 g GA at assessment 31–38 weeks
Nutritive sucking
Infant behavioural state was not a significant predictor of the number of feeding readiness behaviours. The number of feeding readiness behaviours was predictive of feeding efficiency. Group assignment was a marginally significant predictor of feeding. Infants’ gestational at birth and gestational age at entry were identified as significant predictors of the number of feeding readiness behaviours.
Thoyre & Brown (2004)
Non-experimental, longitudinal study
USA One group n = 22 GA at birth = ? GA at start of study = 34 weeks Co-morbidities = ?
Nutritive sucking
Conditions observed during the feeding observations explained most of the variation in engagement. Engagement was more likely to occur during the early phase of feeding, during feeding episodes that began with infant readiness and during feeding episodes with higher mean oxygen saturation during the episode. Feeding episodes with less jiggling of the nipple had a significantly greater amount of engagement.
Pridham (1993) Non-experimental, retrospective correlational study
USA One group Preterm infants n = 55 All infants had BPD GA at birth <32 weeks PCA Exclusion criteria: infants with medical or congenital conditions potentially interfering with feeding including tracheal-oesophageal fistula, cleft palate, gastroschisis, NEC and gastrostomy
Nutritive sucking
Gestational age and days on mechanical ventilation, CPAP, or supplementary oxygen influence age at start and at completion of nipple feeding. The infant’s weight when nipple feeding was introduced was the primary determinant of age at introduction of nipple feeding. The physical condition variable did not influence the transition time. Weight gain between the time of complete nipple feeding and hospital discharge was less for infants who were on supplementary oxygen longer and who were older when completely nipple fed.
172
Reference Study Design Sample Outcome Measures
Results
Hook-Morris, Miller-Loncar, Landry, Smith, Swank & Denson (1999)
Non-experimental longitudinal study
USA One group preterm infants n = 161 BW less than 1600 g Followed until 24 months of age
Nutritive sucking
Results show that when controlling for birth weight and GA, the severity of respiratory complications was significantly related to reaching full nipple feeds. Further, when controlling for the severity of respiratory complication, GA and socioeconomic status, an increased length of time to nipple feeds was significantly associated with a poorer mental outcome.
McCain, Fuller & Gartside (2005)
Experimental study Standard versus semi-demand transition to feeding
USA GA at birth <34 weeks GA at intervention = ? Co-morbidities = ? Exclusion criteria: congenital anomalies, gastrointestinal conditions, neurological diagnoses or IVH > grade 2.
Physiological stability
In the 48 hours after achieving full oral feeding, control infants had significantly more feeding bradycardia episodes than experimental infants. The semi-demand protocol that included use of a pacifier before oral feeding, opportunity to feed orally up to eight times per day and use of infant behaviour to begin and end feedings resulted in the autonomic balance of the experimental infants being more stable as they learned to feed orally.
Chen, Wang, Chang & Chi (2000)
Non-experimental, correlational study
Taiwan One group Preterm n = 25 All babies less than 1800 g GA birth = 29.2 (mean) GA at commencement = 38.5 PCA Co-morbidities = ?
Physiological stability
Oxygen saturation and body temperature of the preterm infants were significantly higher when they were breastfed. There were two episodes of apnoea and 20 episodes of oxygen desaturation during bottle feeding and none during breastfeeding.
173
Reference Study Design Sample Outcome Measures
Results
Dowling (1999) Quasi-experimental, alternating treatment design Bottle feeding versus breastfeeding
USA One group Preterm infants n = 8 GA at birth = ? but mean 30.2 weeks GA Inclusion criteria: weight >1500 g at time of feed, weight appropriate for GA, Absence of craniofacial abnormalities, oxygen dependence and selected mediations.
Nutritive sucking
Statistically significant differences were found in that infants breathed more during sucking bursts for breastfeeding sessions when compared to bottle-feeding sessions and had fewer episodes of oxygen desaturation during breastfeeding. A characteristic sucking waveform associated with organised breathing was observed for some infants during bottle feeding with the orthodontic nipple.
Anderson Behavioural State Scale (ABSS); Brazelton Neonatal Behavioural Assessment Scale (BNBAS); BPD =
bronchopulmonary dysplasia; BW = birth weight; CPAP = continuous positive airway pressure; EMG =
Electromyographic; GA = gestational age; GP = group; IVH = intraventricular haemorrhage; n = sample size;
NEC = necrotising enterocolitis; NNS = non-nutritive sucking; PCA = post-conceptual age; PMA =
postmenstrual age; PNA = postnatal age.
174
175
Appendix B: Search Strategy for Systematic Review
1 CENTRAL search strategy
There were 92 results.
Each keyword was searched for in Title, Abstract or Keywords. There were 92
results for: (preterm or premature) and (feeding or breast or bottle) and (read* or
commence or introduc* or start* or establish*).
2 MEDLINE search strategy
There were 367 results.
S1 preterm or preterm or premature or low birth weight or lowbirth weight or LBW
S2 newborn* or new born* or baby or babies or neonat* or infant*
S3 S1 and S2
S4 (MH ‘Infant, Premature’)S5 S3 or S4
S6 commenc* or start* or begin* or readiness or Introduc*
S7 breast fe* or breastfe* or bottle fe* or bottlefe* or nipple fe* or oral fe*
S8 (MH ‘Bottle Feeding’) or (MH ‘Breast Feeding’) or (MH ‘Feeding Methods’)
S9 (MH ‘Feeding Behavior’) or (MH ‘Sucking Behavior’) or feeding behaviour or
feeding behavior or sucking behaviour or sucking behavior
S10 S7 or S8 or S9
S11 S5 and S6 and S10
3 EMBASE search strategy
72 results
(neonat* OR infant * or newborn OR baby OR babies) AND (preterm OR preterm
OR premature) AND (bottle fe* OR breast fe* OR nipple fe* OR oral fe*) AND
(commenc* OR readiness OR begin* OR introduc*)
4 CINAHL search strategy
There were 161 results.
S1 preterm or preterm or premature or low birth weight or lowbirth weight or LBW
S2 newborn* or new born* or baby or babies or neonat* or infant*
S3 S1 and S2
176
S4 (MH ‘Infant, Premature’)
S5 S3 or S4
S6 commenc* or start* or begin* or readiness or Introduc*
S7 breast fe* or breastfe* or bottle fe* or bottlefe* or nipple fe* or oral fe*
S8 (MH ‘Bottle Feeding’) or (MH ‘Breast Feeding’) or (MH ‘Infant Feeding’)
S9 sucking behaviour or sucking behavior or (MH ‘Sucking Behavior’) or feeding
behavior or feeding behaviour
S10 S7 or S8 or S9
S11 S5 and S6 and S10
5 Health Source search strategy
There were 66 results.
S1 preterm or preterm or premature
S2 newborn* or new born* or baby or babies or neonat* or infant*
S3 S1 and S2
S4 breast fe* or breastfe* or bottle fe* or bottlefe* or nipple fe* or oral fe*
S5 commenc* or start* or begin* or readiness or Introduc*
S6 S3 and S4 and S5
6 Web of science search strategy
There were 150 results.
Topic = (preterm or premature) AND Topic = (infant* or baby or babies or neonat*
or newborn) AND Topic = (breastfe* or bottlefe* or nipplefe* or oral feeding) AND
Topic = (commenc* or start* or readiness or introd* or establish*) Timespan = All
Years. Databases = SCI-EXPANDED, SSCI, A&HCI, CPCI-S, CPCI-SSH.
7 Cochrane search strategy
There were 22 results.
Each keyword was searched for in Title, Abstract or Keywords. There were 22
results for: (preterm or premature) and (feeding or breast or bottle) and (read* or
commence or introduc* or start* or establish*).
177
Appendix C: Characteristics of Studies Excluded from Systematic
Review
Studies excluded from consideration
Reference Reason for exclusion
Bauer (2008) Does not compare methods to determine feeding readiness. This was an observational study involving clinical observation and assessment of feeding readiness and performance of preterm infants during the transition period from gavage to bottle feeding.
Breton (2008) Literature review of introduction and transition to oral feedings.
Bühler (2004) Does not compare methods of determining feeding readiness. An observational study examining factors that impact on commencement and transition to full oral feeding.
Cagan (1995) Does not compare methods of determining feeding readiness. This study is an observational study examining behavioural state and feeding behaviours as indicators of feeding readiness.
Church (2006) Does not compare methods of determining feeding readiness. This observational study examines the inter-rater reliability.
Cunha (2009) Does not compare methods of determining feeding readiness. This study describes and compares the sucking patterns of very low-birth-weight preterm and fullterm infants.
da Costa (2008) Literature review of diagnostic tools to determine feeding readiness and feeding performance.
Dodrill (2008) Does not compare methods of determining feeding readiness. This study involves a retrospective chart audit examining early feeding milestones.
Dodrill (2008a) Does not compare methods of determining feeding readiness. This study involves a survey of staff to investigate and document current transitional feeding practices.
Drenckpohl (2009) Not a randomised or quasi-randomised study. This study uses a historical control to evaluate the implementation of a feeding protocol to initiate and advance feeds. Initiation is commenced at 30 weeks but no assessment is made.
Fernández (2007) Not research but an article that discusses feeding readiness and the transition to suck feeds.
Flint (2007) Does not compare methods of determining feeding readiness. This study involves an observational, retrospective cohort study design in which feeding milestones were examined.
178
Reference Reason for exclusion
Frischknecht (2005) Not a study but an article that describes feeding readiness in preterm infants.
Fujinaga (2007a) Does not compare methods of determining feeding readiness. This is an observational study to test for inter-rater reliability.
Howe (2007) Does not compare methods of determining feeding readiness. This is an observational study design to assess the validity and reliability of the Neonatal Oral Motor Assessment Scale.
Kinneer (1994) Does not compare methods of determining feeding readiness. This study involved a survey of neonatal nurseries to find out how clinicians determine feeding readiness.
Kirk (2007) Not a randomised or quasi-randomised study. This study compares a historical control with a study group. No psychometric testing reported.
Lau (2007) Not primary research but a discussion article on feeding initiation and progression.
Lemons (1996) Not research but an article discussing transition to breast or bottle feeds.
Ludwig (2007) Not research but an article that describes a feeding readiness scale developed by authors as part of their change in feeding documentation.
McCain (2001) This study does not evaluate the use of a feeding readiness indicator independently as the intervention incorporates a period of non-nutritive sucking. The effectiveness of assessing feeding readiness alone on the primary outcomes cannot be established for this study.
McCain (2002) This study does not evaluate the use of a feeding readiness indicator independently as the intervention incorporates a period of non-nutritive sucking. The effectiveness of assessing feeding readiness alone on the primary outcomes cannot be established for this study.
McCain (2003) Not a study but an article that describes an evidence-based guideline for the introduction oral feeding.
McGrath (2002) Does not compare methods of determining feeding readiness. This is an observational study that looks at the association between alertness and ability to engage in nutritive sucking.
McGrath (2003) Does not compare methods of determining feeding readiness. This study describes the content validity as well as an observational, pilot study of a feeding readiness scale.
McGrath (2004) Not research but an article discussing feeding readiness in preterm infants.
McGrath (2005) Does not compare methods of determining feeding readiness. This observational study explores factors associated with feeding readiness.
179
Reference Reason for exclusion
Neiva (2008) Does not compare methods of determining feeding readiness. This study established content validity of non-nutritive sucking scoring system as well as reported the use of the tool within an observational study.
Pickler (2005) Not research. This article describes a theoretical model for feeding readiness in preterm infants.
Pickler (2005a) Does not compare methods of determining feeding readiness and is part of a larger study. This study investigates the relationship between feeding readiness indicators and feeding performance.
Pickler (2009) Does not compare methods of determining feeding readiness. Does not measure feeding readiness. This study examines the effects of feeding experience, maturity and morbidity on clinical milestones.
Premji (2000) Does not compare methods of assessing feeding readiness but investigates the safety and efficacy of implementing a clinical practice guideline for nutritional management compared to no guideline.
Premji (2002) Not research but describes the development of a clinical practice guideline for feeding very low-birth-weight infants.
Premji (2004) Not research but describes the background and implementation of an oral feeding protocol.
Ross (2002) Not research but an article describing the transition from gavage feeds to suck feeds in preterm infants.
Rossarolla (2009) Does not compare methods of determining feeding readiness. This observational study established discriminant validity of the feeding readiness tool developed by Fujinaga.
Shaker (2007) Not research but an article that describes a new feeding protocol.
Siddell (1994) Does not compare methods of determining feeding readiness. This study involved a survey of neonatal nurseries to find out criteria used to determine feeding readiness.
Thoyre (2003) Not research but an article that discusses the transition from gavage to suck feeds.
Thoyre (2005) Not research but an article that describes the early feeding skills assessment checklist.
Thoyre (2009) Does not compare methods of determining feeding readiness. This observational study looks at the validity and reliability of the dynamics of early infant feeding instrument.
White-Traut (2002) Not testing an assessment instrument. The study tested the effects of an auditory, tactile, visual and vestibular (ATVV) intervention on feeding readiness and performance.
White-Traut (2005) Secondary analysis to examine the relationship between behavioural state and the frequency of feeding readiness behaviours.
180
181
Appendix D: List of Variable Fields Exported from the N-CRS
Variable name Type of variable
Baby’s hospital number String
Surname String
First name String
Date of birth Date
Sex Dichotomous
Gestational age at birth Continuous/interval
Birth weight Continuous/interval
APGAR score at one minute Ordinal
APGAR score at five minutes Ordinal
Mechanical ventilation hours Continuous/interval
CPAP hours Continuous/interval
Days of oxygen Continuous/interval
Nasal oxygen Dichotomous
Discharged home with oxygen Dichotomous
Meconium aspirated Dichotomous
Antibiotics suspected sepsis Dichotomous
Antibiotics confirmed sepsis Dichotomous
Caffeine given Dichotomous
Feed commencement date Date
IV out date Date
TPN days Continuous/interval
Lowest weight Continuous/interval
Date birth weight regained Date
Post-conceptual age (PCA) at discharge Continuous/interval
Discharge weight Continuous/interval
Discharge date Date
Length of stay in days Continuous/interval
Feeding method at discharge Categorical
What formula was used Categorical
Antenatal intention for feeding Categorical
Mum’s hospital number String
Mum’s first name String
Date of birth Date
Previous pregnancies Continuous/interval
Previous premature deliveries Continuous/interval
Previous perinatal deaths Continuous/interval
Previous miscarriages/terminations Continuous/interval
Race Categorical
Marital status Categorical
Delivery method Categorical
Maternal diabetes Dichotomous
Gestational diabetes Dichotomous
Uses insulin Dichotomous
Presenting obstetric problem Categorical
182
Variable name Type of variable
Other obstetric problem String
183
184
Appendix E: Chart Audit Form
Parameter Description Codes
UR number Hospital number Xxxxxxx
DOB Baby’s date of birth xx/xx/xxxx
Sex Gender 1 = male 2 = female
K Age at delivery (in weeks) xx
BW Birth weight (in grams) xxxx
Multiple Multiple birth 0 = no 1 = yes
Plural Number of babies 1 = singleton 2 = twins 3 = triplets 4 = quadruplets
Order Birth order 1 = 1st born 2 = 2nd born 3 = 3rd born 4 = 4th born
APGAR 1 min APGAR score at 1 min post-delivery 0–10
APGAR 5 min APGAR score at 5 min post-delivery 0–10
Mech ventilation
Had mechanical ventilation during stay 0 = no 1 = yes
Mech Vent hrs Mechanical ventilation (in hours). Put 0 if none
CPAP Had continuous positive airway pressure during hospital stay.
0 = no 1 = yes
CPAP hrs Continuous positive airway pressure (in hours). Put 0 if none
Additional O2 Required additional O2 (could be via mechanical vent, CPAP, headbox, cot, nasal)
0 = no 1 = yes
Days of O2 Number of days requiring additional O2
Nasal O2 Had nasal oxygen during hospital stay 0 = no 1 = yes
Home O2 Discharged with nasal oxygen 0 = no 1 = yes
185
Parameter Description Codes
Mec Asp Meconium aspiration (aspirated or in liquor) 0 = no 1 = yes
IV antibiotics Used any intravenous antibiotics 0 = no 1 = yes
Infection Confirmed infection 0 = no 1 = yes
Caffeine Commenced on intravenous or oral caffeine 0 = no 1 = yes
Commence caff
Date commenced caffeine xx/xx/xxxx
Finished caff Date caffeine was discontinued xx/xx/xxxx
Total day caff Total number of days on caffeine (in days) Can be calculated
Feeding comm.
Date feeding commence (suck or enteral feeds) xx/xx/xxxx
Breast milk Has the baby had any breast milk during their stay 0 = no 1 = yes
1st breast milk Date first breast milk intake (suck or enteral feeds) xx/xx/xxxx
Method 1st attempt
Method of feeding for first suck feed 1 = breast 2 = bottle
1st attempt Date of first attempt at either breast or bottle feed xx/xx/xxxx
Total enteral Total number of days of naso/orogastric feeds
Required TPN Had TPN during hospital stay 0 = no 1 = yes
Days TPN Total number of days of TPN (days)
Date feed Date of the first breast/bottle feed in which volume had been taken off the enteral feed
xx/xx/xxxx
Age feed Post-conceptual age at first breast/bottle feed in which volume had been taken off enteral feed (weeks)
Day feed Day of life at first breast/bottle feed where there has been volume taken off enteral feed (days)
Feed nasal O2 Had nasal O2 on the day the first breast/bottle feed in which volume had been taken off enteral feed was given
0 = no 1 = yes
Feed add O2 Did they have additional O2 on the day the first breast/bottle feed in which volume had been taken off enteral feed was given
0 = no 1 = yes
Feed caff Did they have caffeine on the day in which volume had been taken off enteral feed
0 = no 1 = yes
Docu feed Had the first feed been described 0 = no 1 = yes
Description feed
Write description of feed e.g. length of feed, quality of sucking etc.
Wt feed Last weight prior to the breast/bottle feed in which volume was taken off enteral feed (grams)
Feed apnoea Date of last recorded apnoea prior to first successful breast/bottle feed
Days apnoea Days since last apnoea Can be calculated
1st full suck feed
Date of first full breast or bottle feed requiring no top ups
xx/xx/xxxx
Breast 3 Number of millilitres of breast milk at Day 3
Breast 7 Number of millilitres of breast milk at Day 7
Breast 14 Number of millilitres of breast milk at Day 14
Breast 21 Number of millilitres of breast milk at Day 21
Formula 3 Number of millilitres of formula at Day 3
Formula 7 Number of millilitres of formula at Day 7
186
Parameter Description Codes
Formula 14 Number of millilitres of formula at Day 14
Formula 21 Number of millilitres of formula at Day 21
BF3 Number of breastfeeds a day including nuzzle feeds at Day 3
BF7 Number of breastfeeds a day including nuzzle feeds at Day 7
BF14 Number of breastfeeds a day including nuzzle feeds at Day 14
BF21 Number of breastfeeds a day including nuzzle feeds at Day 21
Low wt Lowest weight (in grams)
Regain BW Date regained birth weight xx/xx/xxxx
PCA at discharge
Post-conceptual age at discharge (weeks)
Disc wt Discharged weight (in grams)
Date of Dis Date of discharge xx/xx/xxxx
LOS Number of days in hospital (in days)
Plan feed Mother’s plan for either breast or bottle at admission
1 = not recorded 2 = breast 3 = bottle 4 = mother unsure
Feed method Method of feeding at discharge 1 = all BF 2 = EBM via bottle 3 = formula via bottle 4 = BF and bottle with EBM 5 = BF and bottle with formula 6 = Formula via bottle and n/g 7 = Bottle EBM + NG
MUR Mother’s UR number
MUMDOB Mother’s date of birth xx/xx/xxxx
MUMAGE Mother’s age (years)
Prev Preg Number of previous pregnancies
Prem deliv Number of premature deliveries
Perinatal deaths
Number of perinatal deaths
Terminations Number of terminations
Ethic Ethnic group 1 = Caucasian 2 = Australian Aboriginal/Torres Strait Islander 3 = Asian 4 = other
Marital Marital status 1 = lives with father of baby 2 = lives apart from father of baby 3 = unknown 4 = mother died
Delivery method
Method of delivery 1 = LSCS in labour 2 = LSCS without labour 3 = spontaneous vaginal 4 = by forceps 5 = by breech extraction 6 = as an assisted breech 7 = by vacuum extraction 8 = by caesarean section
187
Parameter Description Codes
Maternal Diab Did the mother have diabetes? 0 = no 1 = yes
Gest Diab Was the diabetes gestational? 0 = no 1 = yes
Insulin Was insulin required during pregnancy? 0 = no 1 = yes
Obstetric Problem
Reason for delivery 0 = not stated 1 = preterm prelabour rupture of membranes (PPROM) 2 = preterm labour (<37 weeks) 3 = PIH (BP >140/90 4 = APH/abruption 5 = foetal distress leading to intervention 6 = IUGR 7 = decreased Doppler flow 8 = other
Other problems
Labour pain relief
Pain relief during labour and delivery (first and second stage)
0 = no labour 1 = none 2 = nitrous oxide 3 = epidural 4 = spinal 5 = epidural/spinal 6 = narcotic 7 = caudal 8 = other (specify)
Labour pain relief – 2nd
Additional pain relief 0 = no other/ no labour 1 = nitrous oxide 2 = narcotic 3 = epidural 4 = spinal 5 = epidural/spinal 6 = caudal
Labour pain relief – 3rd
Additional pain relief 0 = no other/ no labour 1 = epidural 2 = spinal 3 = epidural/spinal 4 = caudal 5 = other
Delivery pain relief
Pain relief for delivery (with or without labour); 1 = none 2 = epidural 3 = spinal 4 = epidural/spinal 5 = caudal 6 = general anaesthetic 7 = local to perineum 8 = pudendal 9 = other(specify)
Meth or IV drugs
Methadone or illicit IV drugs used during pregnancy 0 = no 1 = yes 2 = illicit IV drugs
CLD Chronic lung disease which is defined as babies requiring an oxygen requirement at 36 weeks PCA. Calculated from PCA and days on oxygen
0 = No CLD 1 = Yes CLD
188
Instructions for Audit Tool
• Maternal pregnancies: This pregnancy is counted.
• Preterm deliveries: This delivery is included.
• Gestational age is taken to the nearest week (up to 3 days it goes backward; from 4 on it goes forward).
• Date birth weight is regained is sometimes left blank because baby has been discharged.
• Feeding on Days 14 and 21 is sometimes left blank because baby has been discharged.
• Labour pain relief in no particular order i.e. is not in the order that they received it.
• ‘Labour pain relief’ is any pain relief given in the first stage of labour.
• ‘Delivery pain relief’ is any pain relief given for the delivery – with or without labour (an epidural or spinal given in first stage is presumed to be still effective in second stage, so is marked off again).
• In amount of EBM and formula and breastfeeds on Days 3, 7, 14 and 21, field is left blank if baby has been discharged.
• Date regained birth weight if achieved after discharged, I have written ‘after D/C’ – I have not left it blank.
• If the baby is NBM on Day 3, 7, 14 or 21, I have written ‘0’ e.g. 1105222.
• Additional oxygen is any extra oxygen the baby requires – includes vent, CPAP, headbox, nasal and cot.
• First successful breastfeed/bottle feed is defined as the first feed where volume is taken off enteral feeds.
• If baby is discharged on oxygen then the field which asks how many days on additional oxygen I have put the same number as days in hospital (or length of stay) because it was difficult to work out when the baby was weaned from oxygen after discharge.
• If baby is discharged on Day 3, 7, 14 or 21, only the feeds recorded in feed chart for that day are put down – not an estimate of what the baby may have had after discharge on that day.
189
Appendix F: Participant Information Letter for Focus Groups
Dear Fellow Staff Member, Re: Development of an assessment tool to assess feeding readiness in preterm infants My name is Linda Crowe. I am a Clinical Research Nurse from the Mater Nursing Research Centre (NRC) and a PhD student from the School of Nursing, Queensland University of Technology (QUT). As part of my studies under the supervision of Prof Anne Chang, Dr Jenny Fraser and Prof David Tudehope, I am conducting focus groups to explore the initiation of breastfeeding in preterm infants. I am writing to invite you to participate in this activity. This project has been approved by the Director of Neonatology, Prof David Tudehope and the Nursing Director, Neonatology Lynne Elliott. If you agree to be involved, your participation will involve attending a one-hour group discussion (involving 6–10 people) which will be audiotaped. The venue for the focus group will be in the New Life Centre Conference Room, Level 4, at the Mater Mothers Hospital and will be held between 10:30 am and 11:30 am. Sessions on Tuesday, 11th December, 2007 and Wednesday, 6th February, 2008 will concentrate on the initiation of breastfeeding in special care nursery while the session on the Wednesday, 30th January, 2008 will focus on commencement in Intensive Care Nursery. You are only required to attend one of these sessions. Could you please advise Linda Crowe ([email protected]) which session date suits you. Your participation in this project is voluntary. If you do agree to participate, you are free to withdraw from participation at any stage prior to or during the focus group. However, it must be noted that any information that you do provide may still be used, as participants will remain anonymous in the transcript with no record of who stated what information. Your decision to participate will in no way impact upon your current or future relationship with QUT (for example your grades) or with Mater Health Services. This project is not expected to benefit you directly. However, your involvement in this activity will contribute to the knowledge of how breastfeeding is commenced in preterm infants. There are no risks beyond normal day-to-day living associated with your participation in this project. To maintain confidentiality, the names of individual persons will not be identified or associated with comments made during the discussion. Consequently, comments will not be verified by the participants prior to inclusion. Important information to know about the audio recording of the focus group includes:
• tapes will be destroyed once the contents have been transcribed and the transcription has been checked with the audio recording,
• the recording will not be used for any other purpose (e.g. instructional aide),
190
• the recordings and transcript will only be available to Linda Crowe, her research assistant and her supervisors and
• it will not be possible to participate in the project if you are not willing to be audiotaped.
Your presence at the focus group will be taken as consent to participate in this project. Please contact Linda Crowe (ph 3163-8638, e-mail: [email protected]) to have any questions answered or if you require further information about this project. Any concerns or complaints can be directed to my supervisor Prof Anne Chang (ph (07) 3163-8596, e-mail: [email protected]). A summary of the findings of the focus groups will be available if requested by the participants. Your participation will be greatly appreciated. Yours sincerely, Linda Crowe RN M Mid PhD candidate
191
Appendix G: Participant Information Letter for Delphi Survey
Dear Colleagues, Re: Development of an assessment tool to assess feeding readiness in preterm infants My name is Linda Crowe. I am a neonatal nurse from the Mater Health Services and a PhD student from the School of Nursing, Queensland University of Technology (QUT). As part of my studies under the supervision of Prof Anne Chang, Dr Jenny Fraser and Prof David Tudehope, I am conducting a Delphi survey to explore the initiation of breastfeeding in preterm infants. I am writing to invite you to participate in this activity. This project will look at how health-care professionals decide when a preterm infant is ready to commence nutritive breastfeeding. Information gained will be used to form an assessment tool designed to guide health-care professionals in initiation of breastfeeding in the preterm infant population. If you agree to be involved, you will be asked to complete a web-based survey approximately two or three times which will take approximately 5–10 minutes each time. Once all the initial surveys have been completed, returned and analysed, you will then be asked to score the list of the same items but this time will do so with knowledge of your own mean score and the panel’s overall mean scores. Surveys will continue to be sent until an agreed list of decision-making criteria is developed. Usually, no more than three rounds are required to achieve consensus. Your participation in this project is voluntary. If you do agree to participate, you are free to withdraw from participation at any stage prior to or during the survey rounds. However, it must be noted that any information that you do provide may still be used. Your decision to participate will in no way impact upon your current or future relationship with QUT (for example your grades) or with Mater Health Services. This project is not expected to benefit you directly. However, your involvement in this activity will contribute to the knowledge of how breastfeeding is commenced in preterm infants. There are no risks beyond normal day-to-day living associated with your participation in this project. Confidentiality of the information will be maintained and safeguarded. Important information to know about the survey includes:
• names of participants are required to enable the provision of results of preliminary Delphi rounds and to enable subsequent Delphi rounds to be e-mailed; these names will be only known by Linda Crowe and her research assistant
• the survey data will be de-identified for use by Linda Crowe, her research assistant and her supervisors
• Individuals will not be identified when findings are published.
192
Your return of the Delphi Survey will be taken as consent to participate in this project. Please contact Linda Crowe (e-mail: [email protected]) to have any questions answered or if you require further information about this project. Any concerns or complaints can be directed to my supervisor Prof Anne Chang (ph (07) 3163-8596, e-mail: [email protected]). A summary of the findings of the Delphi survey will be available if requested by the participants. Your participation will be greatly appreciated. Yours sincerely, Linda Crowe RN M Mid PhD candidate
193
Appendix H: Staff Information Letters for Observational Study
Information sheet for nursing/midwifery staff working in special care nurseries Project Title: Testing of a newly developed instrument to assess a preterm infant’s readiness to commence breastfeeds. Investigator: Ms Linda Crowe RN MMid PhD candidate [email protected] Supervisors: Prof Anne Chang RN PhD [email protected] Dr Jenny Fraser RN PhD [email protected] Prof David Tudehope MBBS, FRACP [email protected] About the study: My name is Linda Crowe. I am a neonatal nursing research officer from the Mater Mothers Research Centre (MMRC) and a PhD student from the School of Nursing, Queensland University of Technology (QUT). This project is part of my studies under the supervision of Prof Anne Chang, Dr Jenny Fraser and Prof David Tudehope. I am writing to inform you about this project and your possible involvement. Previous research that I have conducted has focused on asking health-care professionals how they determine feeding readiness to develop a set of agreed criteria to assess feeding readiness. This project aims to establish whether this agreed criteria can actually determine if a preterm infant is ready to commence breastfeeding in practice. Your participation: The focus of my research is preterm infants who have been deemed clinically ready to commence breastfeeding. You may become involved if a mother who you are looking after wants to participate in the project. A research nurse/midwife will be employed to collect the data; however, they will not be assisting the mother to breastfeed. Your involvement will include assisting the mother to breastfeed, as you would do in your day-to-day practice. Only the infants will be assessed in relation to their ability to breastfeed. The research nurse/midwife will not be assessing you or how you assist mothers to breastfeed. Arrangements will be made between yourself, the research nurse/midwife and the mother to ensure a time that is convenient is found. Expected benefits: There is not expected to be any direct benefits for babies in your care; however, their involvement in this activity will contribute to the knowledge of how breastfeeding is commenced in preterm infants.
194
Risks: There are no risks beyond normal day-to-day living associated with your participation in this project. Questions/further information: Please contact Linda Crowe (p (07) 3163-6313, e-mail: [email protected]) to have any questions answered or if you require further information about this project. Concerns/complaints regarding the conduct of the project: This study has been approved by Mater Health Services, the RBWH Health Service Districts and QUT Human Research Ethics Committees. Mater Health Services, QUT and the RBWH are committed to researcher integrity and the ethical conduct of research projects. However, if you do have any concerns or complaints about the ethical conduct of the project, you may contact Mater Research Ethics Coordinator on (07) 3163 1585, RBWH HREC Administrator on (07) 3636-5490 or the QUT Research Ethics Coordinator on (07) 3138 2091 or [email protected]. The Research Ethics Coordinator is not connected with the research project and can facilitate a resolution to your concern in an impartial manner. A summary of the findings of this project will be available if requested. Your participation will be greatly appreciated. Yours sincerely, Linda Crowe RN M Mid PhD candidate
195
Information Letter for Lactation Consultants
Project Title: Testing of a newly developed instrument to assess a preterm infant’s readiness to commence breastfeeds. Investigator: Ms Linda Crowe RN MMid PhD candidate [email protected] Supervisors: Prof Anne Chang RN PhD [email protected] Dr Jenny Fraser RN PhD [email protected] Prof David Tudehope MBBS, FRACP [email protected] About the study: My name is Linda Crowe. I am a neonatal nursing research officer from the Mater Mothers Research Centre (MMRC) and a PhD student from the School of Nursing, Queensland University of Technology (QUT). This project is part of my studies under the supervision of Prof Anne Chang, Dr Jenny Fraser and Prof David Tudehope. I am writing to inform you about this project and the possible involvement of babies under your care. I am planning to commence the study within the next week or two. Previous research that I have conducted has focused on asking health-care professionals how they determine feeding readiness in preterm infants to develop a set of agreed criteria. This project aims to establish whether this agreed criteria can actually determine if a preterm infant is ready to commence breastfeeding in practice. The study will involve observing infants immediately prior to and during a breastfeeding event. Data collection will commence once a baby has been deemed ready to commence breastfeeding by clinical staff looking after the baby. Research nurses will collect data in relation to infant characteristics, the newly developed tool, milk intake and exhibited breastfeeding behaviour. To participate in this study, the babies will need to meet the following criteria:
• Born ≤36 weeks PCA,
• Have not had more than five attempts at breastfeeding,
• Inpatients of Mater Mothers Hospital,
• Tolerating enteral feeding,
• Medically stable and
• Their mothers’ wish to breastfeed. Infants who will be excluded from the study are those who were born to substance abuse mothers or who have congenital malformation, syndromes or severe neurological problems.
196
Your participation: The focus of my research is preterm infants who have been deemed clinically ready to commence breastfeeding. You may become involved if a mother who you are looking after wants to participate in the project. A research nurse/midwife will be employed to collect the data; however, they will not be assisting the mother to breastfeed. Your involvement will include assisting the mother to breastfeed, as you would do in your day-to-day practice. Only the infants will be assessed in relation to their ability to breastfeed. The research nurse/midwife will not be assessing you or how you assist mothers to breastfeed. Arrangements will be made between yourself, the research nurse/midwife and the mother to ensure a time that is convenient is found. Expected benefits: There is not expected to be any direct benefits for babies; however, their involvement in this activity will contribute to the knowledge of how breastfeeding is commenced in preterm infants. Risks: There are no risks beyond normal day-to-day living associated with this project. Questions/further information: Please contact Linda Crowe (p (07) 3163-8181, e-mail: [email protected]) to have any questions answered or if you require further information about this project. Concerns/complaints regarding the conduct of the project: This study has been approved by Mater Health Services, the RBWH Health Service Districts and QUT Human Research Ethics Committees. Mater Health Services, QUT and the RBWH are committed to researcher integrity and the ethical conduct of research projects. However, if you do have any concerns or complaints about the ethical conduct of the project, you may contact Mater Research Ethics Coordinator on (07) 3163 1585, RBWH HREC Administrator on (07) 3636-5490 or the QUT Research Ethics Coordinator on (07) 3138 2091 or [email protected]. The Research Ethics Coordinator is not connected with the research project and can facilitate a resolution to your concern in an impartial manner. Your support will be greatly appreciated. Yours sincerely, Linda Crowe RN M Mid PhD candidate
197
Information Letter for Medical Staff
Project Title: Testing of a newly developed instrument to assess a preterm infant’s readiness to commence breastfeeds. Investigator: Ms Linda Crowe RN MMid PhD candidate [email protected] Supervisors: Prof Anne Chang RN PhD [email protected] Dr Jenny Fraser RN PhD [email protected] Prof David Tudehope MBBS, FRACP [email protected] About the study: My name is Linda Crowe. I am a neonatal nursing research officer from the Mater Mothers Research Centre (MMRC) and a PhD student from the School of Nursing, Queensland University of Technology (QUT). This project is part of my studies under the supervision of Prof Anne Chang, Dr Jenny Fraser and Prof David Tudehope. I am writing to inform you about this project and the possible involvement of babies under your care. I am planning to commence this study shortly. Previous research that I have conducted has focused on asking health-care professionals how they determine feeding readiness in preterm infants to develop a set of agreed criteria. This project aims to establish whether this agreed criteria can actually determine if a preterm infant is ready to commence breastfeeding in practice. The study will involve observing infants immediately prior to and during a breastfeeding event. Data collection will commence once a baby has been deemed ready to commence breastfeeding by clinical staff looking after the baby. Research nurses will collect data in relation to infant characteristics, the newly developed tool, milk intake and exhibited breastfeeding behaviour. To participate in this study, the babies will need to meet the following criteria:
• Born ≤36 weeks PCA,
• Have not had more than five attempts at breastfeeding,
• Inpatients of RBWH Grantley Noble Unit,
• Tolerating enteral feeding,
• Medically stable and
• Their mothers’ wish to breastfeed.
• Infants who will be excluded from the study are those who were born to substance abuse mothers or who have congenital malformation, syndromes or severe neurological problems. Your participation: You will not be directly involved in the study; however, mothers of babies in your care might decide to participate. A research nurse/midwife has been employed to
198
collect the data; however, they will not be assisting the mother to breastfeed. Clinical nursing/midwifery staff will assist the mother to breastfeed, if required, as they would do in their day-to-day practice. Expected benefits: There is not expected to be any direct benefits for babies in your care; however, their involvement in this activity will contribute to the knowledge of how breastfeeding is commenced in preterm infants. Risks: There are no risks beyond normal day-to-day living associated with this project. Questions/further information: Please contact Linda Crowe (p (07) 3163-6313, e-mail: [email protected]) to have any questions answered or if you require further information about this project. Alternatively, you may also contact Anndrea Flint CNC Special Care Nursery at the RBWH (ph (07) 3636-7834, e-mail [email protected]). Concerns/complaints regarding the conduct of the project: This study has been approved by Mater Health Services, the RBWH Health Service Districts and QUT Human Research Ethics Committees. Mater Health Services, QUT and the RBWH are committed to researcher integrity and the ethical conduct of research projects. However, if you do have any concerns or complaints about the ethical conduct of the project, you may contact Mater Research Ethics Coordinator on (07) 3163 1585, RBWH HREC Administrator on (07) 3636-5490 or the QUT Research Ethics Coordinator on (07) 3138 2091 or [email protected]. The Research Ethics Coordinator is not connected with the research project and can facilitate a resolution to your concern in an impartial manner. Your support will be greatly appreciated. Yours sincerely, Linda Crowe RN M Mid PhD candidate
199
Appendix I: Information Letters for Parents with Babies in SCN
Information Letter for Parents of Infants in MMH Special Care Nursery
Project Title: Testing of a newly developed instrument to assess a preterm infant’s readiness to commence breastfeeds. Investigator: Ms Linda Crowe RN MMid PhD candidate [email protected] Supervisors: Prof Anne Chang RN PhD [email protected] Dr Jenny Fraser RN PhD [email protected] Prof David Tudehope MBBS, FRACP [email protected] About this study: My name is Linda Crowe. I am a neonatal nursing research officer from the Mater Mothers Research Centre (MMRC) and a PhD student from the School of Nursing, Queensland University of Technology (QUT). This project is part of my studies under the supervision of Prof Anne Chang, Dr Jenny Fraser and Prof David Tudehope. I am writing to invite you to participate in this project. Currently, for preterm infants, breastfeeding is commenced based on health-care professionals’ knowledge and experience of caring for preterm infants. This has led to individuals having their own “checklist” on what they look for to determine feeding readiness. Previous research that I have conducted has focused on asking health-care professionals how they determine feeding readiness in order to develop a set of agreed criteria to assess feeding readiness. This project aims to establish whether these agreed criteria can actually determine if a preterm infant is ready to commence breastfeeding in practice. Your participation: Your participation in this project is voluntary. If you do agree to participate, you are free to withdraw at any stage prior to or during the project. Your decision to participate will in no way impact upon your current or future relationship with Mater Health Services. You and your baby will continue to receive assistance and support in breastfeeding even if you decide not to participate. If you agree to be involved, your participation will involve you breastfeeding your infant while being observed once only by a research nurse/midwife experienced in preterm babies and breastfeeding. Prior to the feed, your baby will be weighed on scales and assessed using the new assessment tool. The assessment tool involves the research nurse/midwife using pre-determined criteria such as the baby’s breathing rate, or how awake the baby is, to assess readiness to commence breastfeeding. The nurse/midwife looking after you and your baby will then assist you to breastfeed.
200
The research nurse/midwife will observe the feed to see how well your baby is able to breastfeed. Your baby will be weighed again to see how much breast milk, if any, they have taken. Your baby may or may not be interested in breastfeeding at this particular time which is not a problem. Whether they breastfeed or not, this study will provide us with valuable information. The breastfeeding session will take place soon after the doctors and nurses/midwives that look after your baby in the nursery have decided it is safe for your baby to attempt to breastfeed. It will take approximately 30–45 minutes of your time to participate and arrangements will be made, so that it also occurs at a time that coincides with your baby’s normal feeding time and is convenient for you. Expected benefits: This project is not expected to benefit you or your baby directly; however, your involvement in this activity will contribute to the knowledge of how breastfeeding is commenced in preterm infants. Risks: There are no risks beyond normal day-to-day living associated with your participation in this project. Confidentiality: The following information is important to you as a participant:
• All results and information collected from this research remain strictly confidential. This is achieved by:
• All identifying information such as your name and hospital number (UR number) will not being recorded on any form except for the cover sheet and you will be given a number code to identify you.
• Information collected in the study will be kept in a locked cabinet in the Mater Mothers Research Centre.
• Identifying information will be kept separate from information collected during the study.
• This study may be published in a journal or presented at a conference. However, no identifying information will be included.
• There will be no financial cost to you for participation in this study.
• This project has been funded by the Golden Casket Fund and the Mater Mothers Research Centre Grant. The funding bodies will not have access to the information obtained during this project.
Consent to participate: We would like to ask you to sign a written consent form to confirm your agreement for you and your baby to participate.
201
Questions/further information about the project: Please contact Linda Crowe (p (07) 3163-6313, e-mail: [email protected]) to have any questions answered or if you require further information about this project. Concerns/complaints regarding the conduct of the project: This study has been approved by the Mater Health Services Human Research Ethics Committee and participants may contact the Mater Research Ethics Coordinator on (07) 3163 1585, should they have any complaints about the conduct of the research, or wish to raise any concerns. The Research Ethics Coordinator may contact the patient representative or hospital ethicist at his/her discretion. A summary of the findings of this project will be available if requested. Your participation will be greatly appreciated. Yours sincerely, Linda Crowe RN M Mid PhD candidate
202
Information Letter for Parents of Infants in RBWH Special Care Nursery
Project Title: Testing of a newly developed instrument to assess a preterm infant’s readiness to commence breastfeeds. Investigator: Ms Linda Crowe RN MMid PhD candidate [email protected] Supervisors: Prof Anne Chang RN PhD [email protected] Dr Jenny Fraser RN PhD [email protected] Prof David Tudehope MBBS, FRACP [email protected] RBWH contact person: Anndrea Flint [email protected] About this study: My name is Linda Crowe. I am a neonatal nursing research officer from the Mater Mothers Research Centre (MMRC) and a PhD student from the School of Nursing, Queensland University of Technology (QUT). This project is part of my studies under the supervision of Prof Anne Chang, Dr Jenny Fraser and Prof David Tudehope. I am writing to invite you to participate in this project. Currently, for preterm infants, breastfeeding is commenced based on health-care professionals’ knowledge and experience of caring for preterm infants. This has led to individuals having their own “checklist” on what they look for to determine feeding readiness. Previous research that I have conducted has focused on asking health-care professionals how they determine feeding readiness in order to develop a set of agreed criteria to assess feeding readiness. This project aims to establish whether these agreed criteria can actually determine if a preterm infant is ready to commence breastfeeding in practice. Your participation: Your participation in this project is voluntary. If you do agree to participate, you are free to withdraw at any stage prior to or during the project. Your decision to participate will in no way impact upon your current or future relationship with RBWH. You and your baby will continue to receive assistance and support in breastfeeding even if you decide not to participate. If you agree to be involved, your participation will involve you breastfeeding your infant while being observed once only by a research nurse/midwife experienced in preterm babies and breastfeeding. Prior to the feed, your baby will be weighed on scales and assessed using the new assessment tool. The assessment tool involves the research nurse/midwife using pre-determined criteria such as the baby’s breathing rate, or how awake the baby is, to assess readiness to commence breastfeeding. The nurse/midwife looking after you and your baby will then assist you to breastfeed.
203
The research nurse/midwife will observe the feed to see how well your baby is able to breastfeed. During the feed, your baby’s vital signs will also be monitored including their heart rate and oxygen level. Your baby will be weighed again to see how much breast milk, if any, they have taken. Your baby may or may not be interested in breastfeeding at this particular time which is not a problem. Whether they breastfeed or not, this study will provide us with valuable information. The breastfeeding session will take place soon after the doctors and nurses/midwives that look after your baby in the nursery have decided it is safe for your baby to attempt to breastfeed. It will take approximately 30–45 minutes of your time to participate and arrangements will be made, so that it also occurs at a time that coincides with your baby’s normal feeding time and is convenient for you. Expected benefits: This project is not expected to benefit you or your baby directly; however, your involvement in this activity will contribute to the knowledge of how breastfeeding is commenced in preterm infants. Risks: There are no risks beyond normal day-to-day living associated with your participation in this project. Confidentiality: The following information is important to you as a participant: All results and information collected from this research remain strictly confidential. This is achieved by:
• All identifying information such as your name and hospital number (UR number) will not being recorded on any form except for the cover sheet and you will be given a number code to identify you.
• Information collected in the study will be kept in a locked cabinet in the Mater Mothers Research Centre.
• Identifying information will be kept separate from information collected during the study.
• This study may be published in a journal or presented at a conference. However, no identifying information will be included.
• There will be no financial cost to you for participation in this study.
• This project has been funded by the Golden Casket Fund and the Mater Mothers Research Centre Grant. The funding bodies will not have access to the information obtained during this project.
Consent to participate: We would like to ask you to sign a written consent form to confirm your agreement for you and your baby to participate.
204
Questions/further information about the project: Please contact Linda Crowe (p (07) 3163-6313, e-mail: [email protected]) to have any questions answered or if you require further information about this project. Alternatively, you may also contact Anndrea Flint CNC Special Care Nursery at the RBWH (p (07) 3636-7834, e-mail [email protected]). Concerns/complaints regarding the conduct of the project: This study has been approved by the RBWH Health Service Districts and QUT Human Research Ethics Committees. QUT and the RBWH are committed to researcher integrity and the ethical conduct of research projects. However, if you do have any concerns or complaints about the ethical conduct of the project, you may contact the RBWH HREC Administrator on (07) 3636-5490 or the QUT Research Ethics Coordinator on (07) 3138 2091 or [email protected]. The Research Ethics Coordinator is not connected with the research project and can facilitate a resolution to your concern in an impartial manner. A summary of the findings of this project will be available if requested. Your participation will be greatly appreciated. Yours sincerely, Linda Crowe RN M Mid PhD candidate
205
Appendix J: Information Letter for Parents of Fullterm Infants in
Postnatal Ward
Project Title: Testing of a newly developed instrument to assess a preterm infant’s readiness to commence breastfeeds. Investigator: Ms Linda Crowe RN MMid PhD candidate [email protected] Supervisors: Prof Anne Chang RN PhD [email protected] Dr Jenny Fraser RN PhD [email protected] Prof David Tudehope MBBS, FRACP [email protected] About this study: My name is Linda Crowe. I am a neonatal nursing research officer from the Mater Mothers Research Centre (MMRC) and a PhD student from the School of Nursing, Queensland University of Technology (QUT). This project is part of my studies under the supervision of Prof Anne Chang, Dr Jenny Fraser and Prof David Tudehope. I am writing to invite you to participate in this project. The research team request your participation in a study about babies’ readiness to commence breastfeeding. Although the focus of the project is on how to determine feeding readiness in preterm infants, fullterm babies are required to participate as a comparison group to the preterm babies. Your participation in this project is voluntary. If you do agree to participate, you are free to withdraw at any stage prior to or during the project. Your decision to participate will in no way impact upon your current or future relationship with Mater Health Services. You and your baby will continue to receive assistance and support in breastfeeding even if you decide not to participate. Unlike fullterm infants who are able to breastfeed soon after birth, preterm infants take time to learn to breastfeed which may take days or weeks after they are born. Breastfeeding is commenced at a time when the baby is deemed to be “ready” by the health-care professionals looking after the baby. Currently, breastfeeding commences for preterm infants based on health-care professionals’ knowledge and experience of caring for preterm infants. This has led to individuals having their own “checklist” on what they look for to determine readiness. Previous research that I have conducted has focused on asking health-care professionals how they determine feeding readiness to develop a set of agreed criteria to assess feeding readiness. This project aims to establish whether these agreed criteria can actually determine if an infant is ready to commence breastfeeding in practice.
206
Your participation: If you agree to be involved, your participation will involve you breastfeeding your infant while being observed once only by a research nurse/midwife experienced in breastfeeding. Prior to the feed, your baby will be weighed on scales and assessed using the new assessment tool. The assessment tool involves the research nurse/midwife using pre-determined criteria such as what is the baby’s heart rate or how awake is the baby to assess readiness to commence breastfeeding. The nurse/midwife looking after you and your baby will then assist you to breastfeed. The research nurse/midwife will observe the feed to see how well your baby is able to breastfeed. Your baby will be weighed again to see how much breast milk if any they have taken. Your baby may or may not be interested in breastfeeding at this particular time which is not a problem. Whether they breastfeed or not, this study will provide us with valuable information. The breastfeeding session will take place while you are in hospital. It will take approximately 30–45 minutes of your time to participate and arrangements will be made, so that it also occurs at a time that coincides with your baby’s normal feeding time and is convenient for you. Expected benefits: This project is not expected to benefit you or your baby directly; however, your involvement in this activity will contribute to the knowledge of how breastfeeding is commenced in preterm infants. Risks: There are no risks beyond normal day-to-day living associated with your participation in this project. Confidentiality: The following information is important to you as a participant: All results and information collected from this research remain strictly confidential. This is achieved by:
• All identifying information such as your name and hospital number (UR number) will not being recorded on any form except for the cover sheet instead you will be given a number code to identify you.
• Information collected in the study will be kept in a locked cabinet in the Mater Mothers Research Centre.
• Identifying information will be kept separate from information collected during the study.
• This study may be published in a journal or presented at a conference. However, no identifying information will be included.
• There will be no financial cost to you for participation in this study.
• This project has been funded by the Golden Casket Fund and the Mater Mothers Research Centre Grant. The funding bodies will not have access to the information obtained during this project.
207
Consent to participate: We would like to ask you to sign a written consent form to confirm your agreement for you and your baby to participate. Questions/further information about the project: Please contact Linda Crowe (p (07) 3163-6313, e-mail: [email protected]) to have any questions answered or if you require further information about this project. Concerns/complaints regarding the conduct of the project: This study has been approved by the Mater Health Services Human Research Ethics Committee and participants may contact the Mater Research Ethics Coordinator on (07) 3163 1585, should they have any complaints about the conduct of the research, or wish to raise any concerns. The Research Ethics Coordinator may contact the patient representative or hospital ethicist at his/her discretion. A summary of the findings of this project will be available if requested. Your participation will be greatly appreciated. Yours sincerely, Linda Crowe RN M Mid PhD candidate
208
209
Appendix K: Parental Consent Form
Project Title: Testing of a newly developed instrument to assess a preterm infant’s readiness to commence breastfeeds. Investigator: Ms Linda Crowe RN MMid PhD candidate [email protected] Supervisors: Prof Anne Chang RN PhD [email protected] Dr Jenny Fraser RN PhD [email protected] Prof David Tudehope MBBS, FRACP [email protected]
I have:
• Read and understood the information package;
• Had any questions or queries answered to my satisfaction;
• Been informed of the possible risks of procedures being conducted;
• Understood that the project is for the purpose of research and not for treatment;
• Been informed that the confidentiality of the information will be maintained and
safeguarded;
• Given permission for access to my medical records, for the purpose of this
research;
• Been assured that I am free to withdraw at any time without comment or penalty;
and
• Agreed that my baby and I will participate in the project.
Signatures:
..................................................
Mother Date
.................................................
Witness Date
..................................................
Investigator (if applicable) Date
210
211
Appendix L: Delphi Surveys
Feeding Readiness Indicators in Preterm Infants Survey for Round 1
Feeding Readiness Indicators in Preterm Infants Survey: first questionnaire This is a survey of staff who work in neonatal nursery, to obtain information regarding their knowledge and attitudes concerning feeding in the nursery. This is the first of a series of questionnaires using the Delphi method. This initial survey is in three sections. Section A enquires personal and professional background. Please click on the appropriate button in each question. Section B enquires knowledge and attitudes. Please click on the appropriate button in each question. Section C accepts comments from respondents. Please enter free text in the comment box. The Delphi technique requires that in the second and any subsequent rounds, we provide you with a comparison of your responses and the mean responses of the other participants. Accordingly, we need to be able to identify the respondents for each questionnaire. Your responses will be kept confidential and all reports of this research will be anonymous. Would you please write your name (e.g., Mary Smith or John Doe) in the space provided below: Name: ______________________________
Section A: Personal and professional background
Question A1: Your gender
□ Male □ Female
Question A2: Your occupation
□ Medical □ Nursing/midwifery □ Allied health
Question A3: What area of the neonatal nursery do you most often work?
□ Intensive care nursery □ High dependency unit □ Special care nursery □ No main area, work in all areas
212
Question A4: How many years experience do you have working with preterm
infants?
□ Less than 1 year □ 1–5 years □ 6–10 years □ 11 or more years
Question A5: How many years experience do you have directly helping mothers
breastfeed preterm infants (i.e. putting babies to the breast)?
□ No experience □ Less than 1 year □ 1–5 years □ 6–10 years □ 11 or more years
Question A6: Are you an IBCLC lactation consultant?
□ Yes □ No
213
Section B: How important do you think the following criteria are?
Instructions For each criteria, click of the box which best reflects the relative importance of each item in how you decide when to initiate the commencement of nutritive breastfeeding in preterm infants. Please feel free to include any comments you may have at the bottom of the survey. Once you have filled in the survey and press send it will be automatically sent to the researcher. Please return the survey by (date inserted).
1 4
Criteria
No
t im
po
rtan
t
Ver
y I
mpo
rtan
t
B1 Baseline heart rate within 120–180 beats per minute
B2 Less than 5–10% difference in baseline heart rate during handling/cares
B3 Baseline respiratory rate within 30–60 breaths per minute prior to commencement of feed
B4 Less than 5–10% difference in baseline respiratory rate during handling/cares
B5 Baseline pulse oximetry oxygen saturation levels greater than 90% prior to commencement of feed
B6 Less than 5–10% difference in baseline pulse oximetry oxygen saturation levels during handling/cares
B7 Infant requiring no supplemental oxygen
B8 Infant not requiring any additional supplemental oxygen during handling/cares
B9 Infant’s oxygen requirement not increased by more than 5–10% of baseline during handling/cares
B10 Baseline body temperature within normal limits prior to commencement of feed
B11 Infant not on caffeine/theophylline medication
B12 Infant has not had a recent bradycardia or apnoea episode
B13 Corrected postnatal age
B14 Infant’s current weight
B15 Sucking regularly on dummy, fingers, feeding tube etc.
B16 Mature non-nutritive sucking pattern
B17 Presence of rooting reflex
B18 Moves smoothly between behavioural (arousal) states
B19 Infant quietly alert just prior to feed time
B20 Infant showing interest at the breast (licking, moving to nipple etc.) during kangaroo care
B21 Infant fussing/demanding feed
Section C: Comments
____________________________________________________________________
214
215
Feeding Readiness Indicators in Preterm Infants Survey for Round 2
Feeding Readiness Indicators in Preterm Infants Survey: second questionnaire The Delphi technique requires that in the second and any subsequent rounds, we provide you with a comparison of your responses and the mean responses of the 42 panel members. Accordingly, we need to be able to identify the respondents for each questionnaire. Your responses will be kept confidential and all reports of this research will be anonymous. Would you please write your name (e.g. Mary Smith or John Doe) in the space provided below. Enter your name here: There are two sections of this questionnaire: Section B enquires knowledge and attitudes. Please click on the appropriate button in each question. Section C accepts comments from respondents. Please enter free text in the comment box.
216
Section B: How important do you think are the following criteria?
Due to some specific comments from panel members, the wording of items B7, B8, B12, B13 and B14 has changed to clarify meaning of these items. This may also change your responses to these items from the first round. For each criteria, click of the box which best reflects the relative importance of each item in how you decide when to initiate the commencement of nutritive breastfeeding in preterm infants. Please feel free to include any comments you may have at the bottom of the survey. Please note: The scale of importance is from 1 (not important) to 4 (very important).
Criteria 1 2 3 4
No
t im
po
rtan
t
Ver
y I
mpo
rtan
t
B1. Baseline heart rate within 120–180 beats per minute
B2. Less than 5–10% difference in baseline heart rate during handling/cares
B3. Baseline respiratory rate within 30–60 breaths per minute prior to commencement of feed
B4. Less than 5–10% difference in baseline respiratory rate during handling/cares
B5. Baseline pulse oximetry oxygen saturation levels greater than 90% prior to commencement of feed
B6. Less than 5–10% difference in baseline pulse oximetry oxygen saturation levels during handling/cares
B7. Infant requiring no supplemental (cot or sub-nasal) oxygen
B8. Infant not requiring any additional supplemental (cot or sub-nasal) oxygen during handling/cares
B9. Infant’s oxygen requirement not increased by more than 5–10% of baseline during handling/cares
B10. Baseline body temperature within normal limits prior to commencement of feed
B11. Infant not on caffeine/theophylline medication
B12. Infant has not had a bradycardia or apnoea episode within the last 24 hours requiring more than stimulation to resolve
B13. Infant attaining a pre-determined minimum corrected postnatal age
B14. Infant’s current weight at time of first attempt to breastfeed
B15. Sucking regularly on dummy, fingers, feeding tube etc.
B16. Mature non-nutritive sucking pattern
B17. Presence of rooting reflex
B18. Moves smoothly between behavioural (arousal) states
B19. Infant quietly alert just prior to feed time
B20. Infant showing interest at the breast (licking, moving to nipple etc.) during kangaroo care
B21. Infant fussing/demanding feed
Section C: Comments
____________________________________________________________________
217
Feeding Readiness Indicators in Preterm Infants Survey for Round 3
Feeding Readiness Indicators in Preterm Infants Survey: third questionnaire The Delphi technique requires that in the second and any subsequent rounds, we provide you with a comparison of your responses and the mean responses of the 42 panel members. Accordingly, we need to be able to identify the respondents for each questionnaire. Your responses will be kept confidential and all reports of this research will be anonymous. Would you please write your name (e.g. Mary Smith or John Doe) in the space provided below. Enter your name here: There are two sections of this questionnaire: Section B enquires knowledge and attitudes. Please click on the appropriate button in each question. Section C accepts comments from respondents. Please enter free text in the comment box.
218
Section B: How important do you think the following criteria are?
Please note, consensus has been reached on eight items and, therefore, these items have not been included in this round of the survey. For each criteria, click of the box which best reflects the relative importance of each item in how you decide when to initiate the commencement of nutritive breastfeeding in preterm infants. Please feel free to include any comments you may have at the bottom of the survey. The scale of importance is from 1 (not important) to 4 (very important)
Criteria 1 4
No
t im
po
rtan
t
Ver
y I
mpo
rtan
t
B1. Baseline heart rate within 120–180 beats per minute
B2. Less than 5–10% difference in baseline heart rate during handling/cares
B4. Less than 5–10% difference in baseline respiratory rate during handling/cares
B7. Infant requiring no supplemental (cot or sub-nasal) oxygen
B8. Infant not requiring any additional supplemental (cot or sub-nasal) oxygen during handling/cares
B9. Infant’s oxygen requirement not increased by more than 5–10% of baseline during handling/cares
B10. Baseline body temperature within normal limits prior to commencement of feed
B12. Infant has not had a bradycardia or apnoea episode within the last 24 hours requiring more than stimulation to resolve
B13. Infant attaining a pre-determined minimum corrected postnatal age
B15. Sucking regularly on dummy, fingers, feeding tube etc.
B16. Mature non-nutritive sucking pattern
B17. Presence of rooting reflex
B18. Moves smoothly between behavioural (arousal) states
219
Appendix M Demographic sheet for observational study
Coded identification number xxxxxx
Hospital 1 = MMH 2 = RBWH
Data collector 1 = LW 2 = JG 3 = AM 4 = RC
DOB xx/xx/xxxx
Gender 1 = male 2 = female
Age at delivery (in weeks) xx
Birth weight (in grams) xxxx
Multiple birth 0 = no 1 = yes
Number of babies 1 = singleton 2 = twins 3 = triplets 4 = quadruplets
Birth order 1 = 1st born 2 = 2nd born 3 = 3rd born 4 = 4th born
APGAR score at 1 minute post-delivery 0–10
APGAR score at 5 minutes post-delivery 0–10
Method of delivery 1 = LSCS in labour 2 = LSCS without labour 3 = spontaneous vaginal 4 = by forceps 5 = by breech extraction 6 = as an assisted breech 7 = by vacuum extraction 8 = by caesarean section
Reason for delivery 0 = not stated 1 = preterm prelabour rupture of membranes (PPROM) 2 = preterm labour (<37 weeks) 3 = PIH (BP >140/90 4 = APH/abruption 5 = foetal distress leading to intervention 6 = IUGR 7 = decreased Doppler flow 8 = other
Neonatal complications unrelated to feeding (circle all that apply)
0 = no complications noted 1 = respiratory distress syndrome 2 = transient tachypnoea of newborn 3 = sepsis definite 4 = sepsis suspected 5 = recurrent apnoea 6 = jaundice requiring phototherapy 7 = chronic lung disease
Time of mechanical ventilation 0 = none, did not receive mechanical ventilation 1 = less or equal to 48 hours 2 = 49 hours to 7 days 3 = 8–14 days 4 = 15–28 days 5 = over 29 days
220
Time of CPAP 0 = none, did not receive CPAP 1 = less or equal to 48 hours 2 = 49 hours to 7 days 3 = 8–14 days 4 = 15 –28 days 5 = over 29 days
Has the baby had any previous attempts at breastfeeding?
0 = no 1 = yes, 1 attempt 2 = yes, 2 attempts 3 = yes, 3 attempts 4 = yes, 4 attempts 5 = yes, 5 attempts
What is the date the baby had its first attempt at breastfeed?
−1 = no previous attempts
What age was the infant at its first attempt? −1 = no previous attempts
Was the any of the attempts successful? 0 = has had no previous attempts 1 = no interest 2 = licked at the breast 3 = attached, occasional suck 4 = attached, sucked well
Unit in which data collection occurred 1 = SCN 2 = HDU/ICN2 3 = ICN
Date of data collection xx/xx/xxxx
Age of infant at data collection (weeks)
Is the baby receiving supplemental oxygen? 0 = no 1 = yes
Has the mother breastfeed an infant before? 0 = no 1 = yes
Who assisted the mother to breastfeed? 1 = clinical nurse looking after infant 2 = nursery lactation consultant 3 = other clinical nurse working in the unit 4 = mum did not receive help
Pre-weight
Post-weight
Difference in weight
Mother’s milk supply 1 = colostrum 2 = transitional milk 3 = mature milk
Is amount of milk enough for baby’s need 1 = baby requiring supplementation of formula/IVT 2 = mum’s supply is just adequate 3 = mum is producing over what baby requires
Is Mum using a nipple shield 0 = no 1 = yes
221
Appendix N: Preterm Infant Breastfeeding Behaviour Scale (PIBBS)
Scale items Maturational steps Score
Rooting Did not root Showed some rooting behaviour Showed obvious rooting behaviour
0 1 2
Areolar grasp (how much of the breast was inside the baby’s mouth)
None, the mouth only touched the nipple Part of the nipple The whole nipple, not the areola The nipple and some of the areola
0 1 2 3
Latched on and fixed to the breast Did not latch on at all so the mother felt it Latched on for ≤5minutes Latched on for 6–10 minutes Latched on for ≥11–15 minutes
0 1 2 3
Sucking No sucking or licking Licking and tasting but no sucking Single suck, occasional short sucking bursts (2–9 sucks) Repeated sucking bursts, occasional long bursts (≥10 sucks) Repeated (≥2) long sucking bursts
0 1 2 3 4
Longest sucking burst 1–5 consecutive sucks 6–10 consecutive sucks 11–15 consecutive sucks 16–20 consecutive sucks 21–25 consecutive sucks ≥26–30 consecutive sucks
1 2 3 4 5 6
Swallowing Swallowing was not noticed Occasional swallowing was noticed Repeated swallowing was notice
0 1 2
Source: Hedberg Nygvist, K., Sjoden, P.-O. & Ewald, U. (1999). The development of preterm infants’ breastfeeding behaviour. Early Human Development, 55, 247–264.
222
223
Appendix O: Preterm Sucking Readiness (PTSR) Scale
Does the infant demonstrate being able to maintain:
1) Baseline respiratory rate within 30–60 breaths/minute Yes � No �
Baseline respiratory rate? _____
2) Baseline pulse oximetry oxygen level greater than 90% Yes � No �
Baseline pulse oximetry oxygen level? _____
3) ≤5–10% difference from baseline pulse oximetry oxygen level during cares or
handling Yes � No�
Lowest pulse oximetry oxygen level during cares? _____
Infant feeding readiness behaviour*
0 1 2 3 4
Behavioural state just prior to disruption at feeding time
Asleep Drowsy Crying Active alert Quietly alert
Transition between behavioural states during cares/handling
Remains asleep
Briefly alert but goes quickly asleep during cares/handling
Briefly alert but gets drowsy during cares/handling
Alert during cares but then becomes drowsy in immediate post-care/handling period
Remains alert during cares and immediate post-care/handling period
Feeding readiness behaviours during cares/handling
Displays no readiness behaviours
Displays occasional feeding readiness behaviours
Displays intermittent feeding readiness behaviours
Displays frequent feeding readiness behaviours
Displays continual feeding readiness behaviours
*Feeding readiness behaviours include behaviours such as sucking on dummy, tube or hand, mouthing, rooting behaviour and showing interest at breast
224
225
Appendix P: Modified Kappa Statistics Tables
Modified Kappa statistics table for 43 participants
Number of experts
Number giving 3 or 4 I-CVI pc k* Evaluation
I-CVI-k* Diff
43 43 1.00 0.00 1.00 Excellent 0.00
43 42 0.98 0.00 0.98 Excellent 0.00
43 41 0.95 0.00 0.95 Excellent 0.00
43 40 0.93 0.00 0.93 Excellent 0.00
43 39 0.91 0.00 0.91 Excellent 0.00
43 38 0.88 0.00 0.88 Excellent 0.00
43 37 0.86 0.00 0.86 Excellent 0.00
43 36 0.84 0.00 0.84 Excellent 0.00
43 35 0.81 0.00 0.81 Excellent 0.00
43 34 0.79 0.00 0.79 Excellent 0.00
43 33 0.77 0.00 0.77 Excellent 0.00
43 32 0.74 0.00 0.74 Excellent 0.00
43 31 0.72 0.00 0.72 Good 0.00
43 30 0.70 0.00 0.70 Good 0.00
43 29 0.67 0.01 0.67 Good 0.00
43 28 0.65 0.02 0.65 Good 0.01
43 27 0.63 0.03 0.62 Good 0.01
43 26 0.60 0.05 0.58 Fair 0.02
43 25 0.58 0.07 0.55 Fair 0.03
43 24 0.56 0.09 0.51 Fair 0.04
43 23 0.53 0.11 0.48 Fair 0.06
43 22 0.51 0.12 0.45 Fair 0.07
43 21 0.49 0.12 0.42 Fair 0.07
43 20 0.47 0.11 0.40 Horrible!! 0.07
43 19 0.44 0.09 0.39 Horrible!! 0.06
43 18 0.42 0.07 0.38 Horrible!! 0.04
43 17 0.40 0.05 0.36 Horrible!! 0.03
43 16 0.37 0.03 0.35 Horrible!! 0.02
43 15 0.35 0.02 0.34 Horrible!! 0.01
43 14 0.33 0.01 0.32 Horrible!! 0.01
43 13 0.30 0.00 0.30 Horrible!! 0.00
43 12 0.28 0.00 0.28 Horrible!! 0.00
43 11 0.26 0.00 0.26 Horrible!! 0.00
43 10 0.23 0.00 0.23 Horrible!! 0.00
43 9 0.21 0.00 0.21 Horrible!! 0.00
43 8 0.19 0.00 0.19 Horrible!! 0.00
43 7 0.16 0.00 0.16 Horrible!! 0.00
43 6 0.14 0.00 0.14 Horrible!! 0.00
43 5 0.12 0.00 0.12 Horrible!! 0.00
43 4 0.09 0.00 0.09 Horrible!! 0.00 I-CVI = Item-level content validity; pc = probability of a chance occurrence was computed using the formula for a binomial random variable, with one specific outcome: pc = [(N!/A!(N-A)]*5N where N = number of experts and A = number agreeing on good relevance; k* = kappa designating agreement on relevance: k* = (I-CVI – pc)/1-pc). Evaluation criteria for kappa, using guidelines described in Cicchetti and Sparrow (1981) and Fleiss (1981): Horrible = k of < .40; Fair = k of 0.40–0.59; Good = k of 0.60–0.74 and Excellent = k >0.74. Cited in Polit, Beck & Owen (2007, p. 465).
226
Modified kappa statistics table for 37 participants
Number of experts
Number giving 3 or 4 I-CVI pc k* Evaluation
I-CVI-k* Diff
37 37 1.00 0.00 1.00 Excellent 0.00
37 36 0.97 0.00 0.97 Excellent 0.00
37 35 0.95 0.00 0.95 Excellent 0.00
37 34 0.92 0.00 0.92 Excellent 0.00
37 33 0.89 0.00 0.89 Excellent 0.00
37 32 0.86 0.00 0.86 Excellent 0.00
37 31 0.84 0.00 0.84 Excellent 0.00
37 30 0.81 0.00 0.81 Excellent 0.00
37 29 0.78 0.00 0.78 Excellent 0.00
37 28 0.76 0.00 0.76 Excellent 0.00
37 27 0.73 0.00 0.73 Good 0.00
37 26 0.70 0.01 0.70 Good 0.00
37 25 0.68 0.01 0.67 Good 0.00
37 24 0.65 0.03 0.64 Good 0.01
37 23 0.62 0.04 0.60 Good 0.02
37 22 0.59 0.07 0.56 Fair 0.03
37 21 0.57 0.09 0.52 Fair 0.04
37 20 0.54 0.12 0.48 Fair 0.06
37 19 0.51 0.13 0.44 Fair 0.07
37 18 0.49 0.13 0.41 Fair 0.08
37 17 0.46 0.12 0.39 Horrible!! 0.07
37 16 0.43 0.09 0.37 Horrible!! 0.06
37 15 0.41 0.07 0.36 Horrible!! 0.04
37 14 0.38 0.04 0.35 Horrible!! 0.03
37 13 0.35 0.03 0.33 Horrible!! 0.02
37 12 0.32 0.01 0.32 Horrible!! 0.01
37 11 0.30 0.01 0.29 Horrible!! 0.00
37 10 0.27 0.00 0.27 Horrible!! 0.00
37 9 0.24 0.00 0.24 Horrible!! 0.00
37 8 0.22 0.00 0.22 Horrible!! 0.00
37 7 0.19 0.00 0.19 Horrible!! 0.00
37 6 0.16 0.00 0.16 Horrible!! 0.00
37 5 0.14 0.00 0.14 Horrible!! 0.00
37 4 0.11 0.00 0.11 Horrible!! 0.00
37 3 0.08 0.00 0.08 Horrible!! 0.00
37 2 0.05 0.00 0.05 Horrible!! 0.00
37 1 0.03 0.00 0.03 Horrible!! 0.00
I-CVI = Item-level content validity; pc = probability of a chance occurrence was computed using the formula for a binomial random variable, with one specific outcome: pc = [(N!/A!(N-A)!]*5N where N = number of experts and A = number agreeing on good relevance; k* = kappa designating agreement on relevance: k* = (I-CVI – pc)/1-pc). Evaluation criteria for kappa, using guidelines described in Cicchetti and Sparrow (1981) and Fleiss (1981): Horrible = k of < .40; Fair = k of 0.40–0.59; Good = k of 0.60–0.74 and Excellent = k > 0.74. Cited in Polit, Beck & Owen (2007, p. 465).
227
Modified Kappa statistics table for 27 participants
Number of experts
Number giving 3 or 4 I-CVI pc k* Evaluation
I-CVI-k* Diff
27 27 1.00 0.00 1.00 Excellent 0.00
27 26 0.96 0.00 0.96 Excellent 0.00
27 25 0.93 0.00 0.93 Excellent 0.00
27 24 0.89 0.00 0.89 Excellent 0.00
27 23 0.85 0.00 0.85 Excellent 0.00
27 22 0.81 0.00 0.81 Excellent 0.00
27 21 0.78 0.00 0.78 Excellent 0.00
27 20 0.74 0.01 0.74 Good 0.00
27 19 0.70 0.02 0.70 Good 0.00
27 18 0.67 0.03 0.65 Good 0.01
27 17 0.63 0.06 0.60 Good 0.02
27 16 0.59 0.10 0.55 Fair 0.04
27 15 0.56 0.13 0.49 Fair 0.07
27 14 0.52 0.15 0.43 Fair 0.08
27 13 0.48 0.15 0.39 Horrible!! 0.09
27 12 0.44 0.13 0.36 Horrible!! 0.08
27 11 0.41 0.10 0.34 Horrible!! 0.06
27 10 0.37 0.06 0.33 Horrible!! 0.04
27 9 0.33 0.03 0.31 Horrible!! 0.02
27 8 0.30 0.02 0.28 Horrible!! 0.01
27 7 0.26 0.01 0.25 Horrible!! 0.00
27 6 0.22 0.00 0.22 Horrible!! 0.00
27 5 0.19 0.00 0.18 Horrible!! 0.00
27 4 0.15 0.00 0.15 Horrible!! 0.00
27 3 0.11 0.00 0.11 Horrible!! 0.00
27 2 0.07 0.00 0.07 Horrible!! 0.00
27 1 0.04 0.00 0.04 Horrible!! 0.00
27 0 0.00 0.00 0.00 Horrible!! 0.00
I-CVI = Item-level content validity; pc = probability of a chance occurrence was computed using the formula for a binomial random variable, with one specific outcome: pc = [(N!/A!(N-A)!]*5N where N = number of experts and A = number agreeing on good relevance. k* = kappa designating agreement on relevance: k* = (I-CVI – pc)/1-pc). Evaluation criteria for kappa, using guidelines described in Cicchetti and Sparrow (1981) and Fleiss (1981): Horrible = k of < .40; Fair = k of 0.40–0.59; Good = k of 0.60–0.74 and Excellent = k > 0.74. Cited in Polit, Beck & Owen (2007, p. 465).
228
229
Appendix Q: Comments from Delphi Survey
Comments from round 1 of feeding readiness indicators in preterm infants
survey
Panel member
Comments
5 I am mainly guided by the individual level of activity and stability during handling.
6
Clearly the decision to commence breastfeeding is based on the balance of a number of factors, with each factor in itself not necessarily ruling out readiness for feeding. The presence of some factors i.e. ‘infant showing interest in the breast’ is a wonderful sign for readiness and when it is present plays a very important role in my decision-making process, but it’s absence does not preclude readiness in an infant that is showing other good signs.
8 I was an IBCLC for 10 years but chose not to requalify at the 10-year mark.
10 A lot of my answers are dependent on how premature the baby is and as my experience is mostly with preterm infants with other problems i.e. cardiac that may be affecting them.
12 I am happy to START with attempting breastfeeding in relatively immature and unstable babies. In these circumstances, most babies will do no more than ‘lick’ the nipple. This may be of benefit to both mother and baby nevertheless.
13 My answers would really depend on a baby’s medical diagnosis and degree of stability.
14
It’s very difficult to quantify the actual assessment that I make. I also assess the whole situation with respect to the mother and baby dyad – has mum had a cuddle/kangaroo cuddle yet? Is there some difficulty in mum’s ability to visit her baby (we have a lot of mums from out of area with other children at home that need her there) etc. Sometimes, I put the baby to the breast more for the maternal emotional support than that the baby has shown any signs of readiness for breast nuzzling/feeding.
16 Careful observation during early nutritive feeds is important as is response to significant changes in physiological stress indicators. Overall ability to remain relatively stable during nutritive feeds more important than individual physiological signs.
18 B13 refers to a preterm baby being at least 32–34 weeks gestation corrected.
19 There is a window of opportunity that needs to be made use of and the baby should get the support it needs to commence suck feeds. This may be increased oxygen, careful pacing etc.
21
We don’t routinely ‘tick off’ criteria in a list like above and tend to take each baby as an individual, i.e. if corrected gestational age in the ball park, frequent alert states and showing signs of interest in breastfeeding, we’ll ‘give them a go’ and see how they manage. They can always still have a gavage feed (or even gavage during B/F) if they’re interested but can’t complete a full breastfeed. In addition, ‘supplemental oxygen’ is a vague term – we wouldn’t B/F a baby in incubator oxygen (i.e. probably acute respiratory illness) but would consider B/F a baby on low flow O2 via nasal cannula if showing above signs of interest (i.e. CLD).
24 As a clinician I work from the premise practice makes perfect, we try to put the babies regularly having skin-to-skin from 30 weeks, feeding should progress from there.
26 It is important that the mother is ready to breastfeed her premature and often tiny baby. Babies speak to us very clearly when they are ready to suck.
27
Once babies are stable enough to commence KMC, we should follow the cues of each individual baby. It is well documented that parameters such as temperature, HR and RR can stabilise with KMC, thus allowing the baby more of a chance to initiate early feeding and its own pace.
29 Sorry, not sure what you mean by B13 and B14 so just ticked ‘1’. In relation to B12, occasional A&B with spontaneous recovery would be OK; however, if requiring stimulation to recover, then not time to initiate BF
31 On our unit, we encourage Kanga cuddles so if the baby shows signs of wanting to suckle we encourage this often the baby is more settled and maintains their temp when being Kanga cuddled.
37 Depends on the individual gender, maturity, state of baby’s health, babies who act older than their gestational age often show feeding cues much earlier.
230
Panel member
Comments
39 Love to see the results of this survey. My answers try to reflect that each baby is an individual and should be treated as such. different parameters would be taken into account for different babies.
41 I think it depends on each individual infant showing signs of interest to feed.
Comments from round 2 of feeding readiness indicators in preterm infants
survey
Participant number Comments
19 Very important to offer suck feeds as soon as possible, with the necessary support.
20 We assess each baby individually; however, we consider time at the breast even if just licks and smells as a learning experience for both mother and baby and very important.
24 Again, I feel it is about giving the mother the opportunities to have skin to skin, so the baby can show interest in feeding if he/she is able. If also assists in the mum gaining confidence in handling and managing her baby.
29
Taking into account the comments from the first survey, I agree, it is important to take each infant and their responses individually. If we only allowed the infant at the breast after reaching a ‘tick-box criteria’, it is likely most prems would never get there! This is probably where critical thinking and assessing each infant’s individual response would come into play.
30 I feel there is a bit difference between and infant fussing and an infant demanding food. Prem babies often fuss and it is not always because they are hungry.
31 As mentioned before, Kanga cuddles are very much encouraged and the babies are often more stable with their temperature and heart rate and Sats.
41
I believe that the baby’s observational behaviours are a better indicator of readiness to feed. In addition, visual observations of the baby’s colour, respiratory effort and alertness are of more value, as equipment may not always provide a true indication of stability. Often a baby’s observations actually stabilise once the baby commences to breastfeed.