Use of Care Bundles in Use of Care Bundles in Critical CareCritical Care

Paul Fulbrook, AustraliaPaul Fulbrook, Australia

Professor of Nursing, Australian Catholic UniversityProfessor of Nursing, Australian Catholic UniversityNursing Director Research, The Prince Charles Hospital, BrisbaneNursing Director Research, The Prince Charles Hospital, Brisbane

Adjunct Professor, University of TasmaniaAdjunct Professor, University of TasmaniaNational Vice President ACCCNNational Vice President ACCCN

Outline� Background and definition� The ventilator bundle� Measures

– Compliance and reliability� The sepsis bundles

– Compliance– Outcomes– Current trials

Background: EBP

• It is imperative that those working in critical care environments examine their practice to ensure that it is evidence-based and of a high quality.

• The importance of knowledge-based practice is constantly emphasised, with the expectation that evidence-based decision-making and practice is at the heart of all healthcare (Bonell, 1999).

• Evidence-based practice emphasises the use of existing research.

Care Bundle: DefinitionCare Bundle: Definition� A small set [no more than 5] of A small set [no more than 5] of

evidence-based interventions for a evidence-based interventions for a defined patient segment/population and defined patient segment/population and care setting that, when implemented care setting that, when implemented together, will result in significantly better together, will result in significantly better outcomes than when implemented outcomes than when implemented individuallyindividually

Resar et al. IHI Innovation Series 2012.Resar et al. IHI Innovation Series 2012.

Care bundlesCare bundles• The ‘bundle’ concept was developed

originally in 2001 by IHI• First described in nursing literature (Fulbrook &

Mooney, 2003)• Goal: to improve critical care processes• Each element relatively independent• Used within a defined population in one

location

Evidence-based practice� Care bundles provide a method for establishing best

clinical practice, which is evidence-based.

� Individual components of each care bundle should be well defined and based on strong science; usually level one or two research.

� By grouping evidence-based practices together, within a single clinical protocol that guides patient management (or process), the overall quality of care given to critically ill patients will improve.

Levels of evidenceLevel of evidence Criteria

1 Strong evidence from at least one systematic review ofmultiple well designed randomised controlled trials

2 Strong evidence from at least one properly designedrandomised controlled trial of appropriate size

3 Evidence from well designed trials without randomisation,single group pre-post, cohort, time series or matched casecontrol studies

4 Evidence from well designed experimental studies frommore than one centre or research group

5 Opinions of respected authorities, based on clinicalevidence, descriptive studies or reports of expertcommittees

(after Moore et al., 1995)

� There should be sufficient evidence to support each element of the bundle

� Each element should be applied to most, if not all, patients

The ventilator care bundleThe ventilator care bundle

•• Deep vein thrombosis (DVT) prophylaxisDeep vein thrombosis (DVT) prophylaxis•• Gastric ulceration prophylaxisGastric ulceration prophylaxis•• Head of bed elevation (>30Head of bed elevation (>30oo))•• Sedation hold (sedation vacation) and assessment of Sedation hold (sedation vacation) and assessment of

readiness to weanreadiness to wean� In 2010, a 5In 2010, a 5 thth element was added: element was added:

– Daily oral care with chlorhexidineDaily oral care with chlorhexidine

� Good evidence of association between VCB Good evidence of association between VCB implementation and VAP reduction implementation and VAP reduction (Lawrence & Fulbrook, 2011)(Lawrence & Fulbrook, 2011)

� Feedback improves compliance but HOBE compliance Feedback improves compliance but HOBE compliance relatively poor relatively poor (Lawrence & Fulbrook, 2012)(Lawrence & Fulbrook, 2012)

Process and outcomeProcess and outcome� Process measures are easier to

measure than outcome measures, and can be used to provide immediate feedback to clinicians

� However, it is important to link process measures with their counterpart outcome measures

� “Outcome measures . . . are ultimately what patients care about” (Berenholtz et al., 2002)

OutcomesOutcomes

Whilst the individual components of a care bundle each has a strong evidence base,

there is, as yet, limited evidence that demonstrates that clustering components in

this way improves patient outcome

(Fulbrook & Mooney, 2003)

All or nothing measurementAll or nothing measurement� Partial compliance not measuredPartial compliance not measured� All elements must be complied with All elements must be complied with

unless medically contra-indicatedunless medically contra-indicatedManagement of early severe sepsis• USA, UK, New Zealand & Australia (n = 2,461)• Scenario-based questionnaire (MC) given to ED, acute medicine, and critical care doctors

• Based on 6-hour resuscitation bundle• Only 2 respondents complied with all SSC guideline recommendations• 81% identified reasons they could not implement at least one of the recommendations

Reade et al. Emerg Med J 2010; 27: 110-115.

ComplianceCompliance

Patient 1 on day 1 = non-compliant = 0%Patient 1 on day 8 = compliant = 100%

Component 1

e.g. DVT prophylaxis

2 e.g. sedation

hold

3 e.g. GU

prophylaxis

4 e.g. head of

bed Patient 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 Day 1 Y Y Y Y N N N N N N Y Y Y N Y Y Day 2 Y Y Y Y N N N N N N Y Y Y N Y Y Day 3 Y N Y Y N N N N Y Y Y Y Y Y Y Y Day 4 Y Y Y Y Y Y N N Y N Y Y Y Y N N Day 5 Y N Y Y Y Y N N Y Y Y Y Y Y Y Y Day 6 Y Y Y N Y N N N N N Y Y Y Y Y Y Day 7 Y Y Y Y Y N Y N Y Y Y Y Y Y Y N Day 8 Y Y Y Y Y N Y N Y Y Y Y Y Y Y Y Care Bundle Compliance. Y = compliance, N = non compliance

ICU compliance day 1 = 0%ICU compliance day 8 = 50%

14/16 = 87.5%

Total =90/128

= 70.3%

9/16 = 56.3%

OverallCompliance

All or NothingCompliance

Overall element compliance DVT 29/32 = 90.6% SH 9/32 = 28.1% GUP 25/32 = 78.1% HOBE 27/32 = 84.4%

Run charts

Care Bundle Component Compliance

0

50

100

150

1 2 3 4 5 6 7 8

Audit Days

Com

plia

nce

(%) DVT

prophylaxis

Sedation hold

GU prophylaxis

Bed headelevation

Run chart: daily compliance

Care Bundle Compliance

0

20

40

60

80

100

120

1 2 3 4 5 6 7 8

Audit Days

Co

mp

lian

ce (

%)

Overall compliance

Target compliance

All or nothing compliance

Cumulative reliabilityT

Compliance and outcome• Assessing compliance

provides limited information in itself.

• Run charts can also be used to demonstrate the impact of interventions on outcome measures.

• Assessing compliance against a local outcome measure is more likely to demonstrate the impact of a new intervention.

• To demonstrate effective change data would need to be collected for at least six months, and preferably a year.

Run chart demonstrating effect of compliance

on an outcome measure

Audit feedback

Bundle Reliability� Measures of reliability inform teams about the extent

of error-free operation of their clinical processes. � This measure tells teams how often every element

included in the relevant bundle was completed in each patient during the review period.

� Not following every indicated bundle element in each patient constitutes an error.

� Defined as the % of cases for whom all applicable bundle elements are completed.

� The suggested reliability goal for compliance is 95% – Numerator: The number of relevant cases that completed all

elements of the bundle– Denominator: The total number of relevant cases

ReliabilityReliability

Component 1

e.g. DVT prophylaxis

2 e.g. sedation

hold

3 e.g. GU

prophylaxis

4 e.g. head of

bed Patient 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 Day 1 Y Y Y Y N N N N N N Y Y Y N Y Y Day 2 Y Y Y Y N N N N N N Y Y Y N Y Y Day 3 Y N Y Y N N N N Y Y Y Y Y Y Y Y Day 4 Y Y Y Y Y Y N N Y N Y Y Y Y N N Day 5 Y N Y Y Y Y N N Y Y Y Y Y Y Y Y Day 6 Y Y Y N Y N N N N N Y Y Y Y Y Y Day 7 Y Y Y Y Y N Y N Y Y Y Y Y Y Y N Day 8 Y Y Y Y Y N Y N Y Y Y Y Y Y Y Y Care Bundle Compliance. Y = compliance, N = non compliance On days 4 & 5, only patient 1 was compliant with all 4 elementsOn days 7 & 8, patients 1 and 3 were compliant with all 4 elementsOverall bundle reliability = 6/32 = 18.8%.8%

If reliability < 95% in any month, teams should identify which specific bundle elements are not being reliably executed.

Surviving SepsisSurviving SepsisA global program to:A global program to:

Reduce mortali ty rates in Reduce mortali ty rates in severe sepsissevere sepsis

• Sepsis Resuscitation Bundle: Tasks that should begin immediately, but must be done within 6 hours for patients with severe sepsis or septic shock.

• Sepsis Management Bundle: Tasks that should begin immediately, but must be done within 24 hours for patients with severe sepsis or septic shock.

Nursing considerationsNursing considerationsAitken et al 2011Aitken et al 2011

� Delphi study based on Surviving Sepsis Campaign Delphi study based on Surviving Sepsis Campaign guidelinesguidelines

� GRADE used to rate quality of evidenceGRADE used to rate quality of evidence� 63 recommendations designed to complement SSC 63 recommendations designed to complement SSC

guidelinesguidelines� Research needed urgentlyResearch needed urgently

– Recognition of deterioration/diagnosis of sepsis/effect of Recognition of deterioration/diagnosis of sepsis/effect of early resusc measuresearly resusc measures

– Methods of haemodynamic support/assessmentMethods of haemodynamic support/assessment– Type and effect of supportive care e.g. pressure injury Type and effect of supportive care e.g. pressure injury

preventionprevention– Application to paediatricsApplication to paediatrics

Sepsis bundles: meta analysis Sepsis bundles: meta analysis Chamberlain et al 2011Chamberlain et al 2011

� 2004-112004-11� 21 sepsis bundle studies: total n 23,43821 sepsis bundle studies: total n 23,438

� Survival benefit:Survival benefit:– 6hr resusc bundle = highest benefit: OR 2.12 (11 6hr resusc bundle = highest benefit: OR 2.12 (11

studies; n = 1,819)studies; n = 1,819)– 24hr bundle = lowest benefit: OR 1.65 (5 studies; n = 24hr bundle = lowest benefit: OR 1.65 (5 studies; n =

16,521)16,521)– Combined bundle benefit: OR 1.74 (12 studies; n = Combined bundle benefit: OR 1.74 (12 studies; n =

5,098)5,098)– ScvOScvO2 2 >> 70% benefit: OR 2.05 (5 studies; n = 11,530) 70% benefit: OR 2.05 (5 studies; n = 11,530)

Compliance & Mortality 2013Compliance & Mortality 2013Authors

Setting Design n Measures Reported compliance Mortality

Almeida et al., 2013

Portugal1 ICU

Prospective cohort

300 • 6h compliance• Compliance of each element• Day vs night compliance

• All or nothing compliance 2%• Element compliance range 4-100%.• Element compliance better at night

Hospital mortality• day 40%• night 34%

Chou et al., 2013

Taiwan1 ICU

Before and after study: 4 phase (modified bundle)

55, 30, 26, 53

• 6- and 24h compliance• All or nothing for both bundles

• Before 20% (phase 1)• After 79.2% (phase 4)

Hospital mortality• Before 43.6%• After 10.0%• (lactate mortality OR 2.2)

Miller et al., 2013

USA11 hospitals 18 ICUs

Prospective cohort (3 phase)

4329 • Combined 6h- and 24h bundle• “All-or-none compliance…non-compliance with any single element interpreted as non-compliance with bundle”

• Baseline compliance 4.9%• Post-implementation compliance 73.4%

Hospital mortality

i) Bundle Compliant• Baseline 21.2%• Post-implementation 8.7%

ii) Bundle non-compliant• Baseline 21.7%• Post-implementation 9.7%

Zhi-qiang et al., 2013

China11 hospitals

Prospective cohort

218 • 6- and 24h compliance• Overall compliance: “when patients met all relevant targets”

• 6h 5.5% (n = 12)• 24h 17.4% (n = 38)

• 28 day mortality 33.0%• 6h comp 25.0%; non-comp 33.5%• 24h comp 26.3%; non-comp 34.4%

2012 Sepsis Guidelines2012 Sepsis Guidelines

� Dellinger RP, Levy MM, Rhodes A, et al. Dellinger RP, Levy MM, Rhodes A, et al. Surviving Sepsis Campaign: International Surviving Sepsis Campaign: International guidelines for management of severe sepsis guidelines for management of severe sepsis and septic shock: 2012. and septic shock: 2012.

Critical Care Medicine. 2013 Feb;41(2):580-Critical Care Medicine. 2013 Feb;41(2):580-637637..

Sepsis bundle modificationsSepsis bundle modifications� The ‘old’ The ‘old’ Sepsis Resuscitation Bundle Sepsis Resuscitation Bundle

was modified into two new bundles:was modified into two new bundles: – Severe Sepsis 3-Hour Resuscitation Severe Sepsis 3-Hour Resuscitation

BundleBundle– 6-Hour Septic Shock Bundle6-Hour Septic Shock Bundle

� The ‘old’The ‘old’ Sepsis Management Bundle Sepsis Management Bundle has been eliminatedhas been eliminated– New: New: Other Supportive TherapiesOther Supportive Therapies

Severe Sepsis 3-hour Severe Sepsis 3-hour Resuscitation BundleResuscitation Bundle

� Measure lactate levelMeasure lactate level� Obtain blood cultures (before giving Obtain blood cultures (before giving

antibiotics)antibiotics)� Administer broad spectrum antibioticsAdminister broad spectrum antibiotics� Administer 30 mL/kg crystallloid for Administer 30 mL/kg crystallloid for

hypotension or lactate hypotension or lactate >> 4 mmol/L (36 4 mmol/L (36 mg/dL)mg/dL)

6-hour Septic Shock Bundle6-hour Septic Shock Bundle1.1. Apply vasopressors if hypotensive (when Apply vasopressors if hypotensive (when

non-responsive to initial fluid resusc) to non-responsive to initial fluid resusc) to maintain MAP maintain MAP >> 65 mm Hg 65 mm Hg

2.2. Persistent hypotension despite volume Persistent hypotension despite volume resusc or initial lactate resusc or initial lactate >> 4 mmol/L 4 mmol/L

– Maintain CVP (target Maintain CVP (target >> 8 mm Hg) 8 mm Hg)– Maintain central venous OMaintain central venous O22 saturation (target saturation (target >>

70%)70%)3.3. Remeasure lactate if initial lactate elevated Remeasure lactate if initial lactate elevated

(target = normal; arterial (target = normal; arterial < < 1.6 mmol/L, 1.6 mmol/L, venous venous << 2.2 mmol/L) 2.2 mmol/L)

Severe Sepsis Bundles: Other Severe Sepsis Bundles: Other Supportive TherapiesSupportive Therapies

1.1. Blood product administrationBlood product administration2.2. Maintain adequate glycaemic controlMaintain adequate glycaemic control3.3. Mechanical ventilation of sepsis-induced Mechanical ventilation of sepsis-induced

ARDSARDS4.4. Sedation, analgesia, and neuromuscular Sedation, analgesia, and neuromuscular

blockadeblockade5.5. DVT and peptic ulcer disease prophylaxisDVT and peptic ulcer disease prophylaxis6.6. NutritionNutrition7.7. Setting goals of careSetting goals of care

3 Major Sepsis Trials� ProCESS (Protocolized Care

for Early Septic Shock)– USA– 3 arms

� ARISE (Australian Resuscitation In Sepsis Evaluation Randomised Controlled Trial)– Australia, commenced 2008– 2 arms

� ProMISe (Protocolised Management in Sepsis Trial)– UK– 2 arms

� Adult ED patients� Inclusion criteria based on

Rivers et al (2001). Early goal-directed therapy in the treatment of sepsis and septic shock. New Eng J Med 345:1368-77.

ProCESS� Protocolized Care for Early Septic Shock � 5-year RCT study� 31 USA emergency departments (n = 1341)� Three approaches to sepsis care

– Early goal directed therapy� Sepsis Resuscitation Bundle

– Protocolized standard care� Does not mandate central lines, inotropes or blood

transfusion– Usual care

ProCESS Outcomes March 2014� ProCESS set out to determine whether a specific

protocol would increase survival in patients with septic shock.

� It showed that patient survival was essentially the same in all three treatment groups– 60-day mortality

� EGDT (n = 439) 21%� Protocol (n = 446) 18.2%� Standard care (n = 456) 18.9%

– 90-day and 1-year mortality ns� indicating that sepsis patients in these clinical

settings were receiving effective care

N Eng J Med (May 2014) 370(18): 1683-93

Surviving Sepsis Campaign: Response to ProCESS

� The SSC will determine (any) appropriate revisions to the bundle elements when the ARISE and ProMISE study results are available.

� ProCESS does not address the protocolized management of patients with severe sepsis without septic shock

� The ProCESS results have no impact on the 3-hour bundle � Regarding the SSC 6-hour bundle (2):

– A companion paper appears to support a mean initial arterial pressure (MAP) target of 65 mm Hg, which is one of the indicators in this bundle. (5)

– The ProCESS paper does not address repeating lactate measures in patients with elevated lactate while literature supports doing so.

– When measured, the first ScvO2 was 71 ± 13%, which is another of the indication of the bundle.

– The majority of the patients in the usual care (56.5%) and protocol-based standard care arms (57.9%) of ProCESS had central lines inserted as part of clinical care.

– The 6-hour bundle currently asks only that CVP be measured and that a venous blood gas be sent from that line to obtain ScvO2. SSC recognizes that alternate means of obtaining results exist and will address specific ways of including those data in future iterations of the quality improvement database.

ARISE Outcomes� 51 centres in Australia & NZ (n = 1600)

– 90-day mortality� EGDT 18.6% (n = 796 )� Usual care 18.8% (n = 804 )

– No significant differences in:� Survival time� In-hospital mortality� Duration of organ support� Hospital LOS

– EGDT group� More vasopressor infusion� More blood transfusions� More dobutamine

� Conclusion– EGDT in ED patients with early septic shock does not reduce all-cause 90-

day mortality

N Eng J Med (Oct 2014) 371(16): 1496-505.

ARISE 6h differences (p < 0.001)EGDT Usual care

CVC insertion 714 (90.0%) 494 (61.9%)

Fluid volume 1.9 l 1.7lVasopressor infusion 66.6% 57.8%

Red cell transfusion 13.6% 7.0%

Dobutamine 15.4% 2.6%MAP at 6h* 76.5 mm Hg 75.3 mm HgED LOS 1.4 h 2.0 h

Also, between 6-72 hours, more EGDT patients received vasopressor infusion (58.8% vs 51.5% p = 0.004) and dobutamine (9.5% vs 5.0% p < 0.001)

* p = 0.04

ProMISE� ICNARC� 56 UK EDs, n = 1260

– 2 arms: EGDT & usual care– Early signs of sepsis & septic shock– Primary outcome 90-day mortality– Cost effectiveness

� Data collection Feb 2011 – July 2014– Closed 24/7/14 – in follow-up

� Awaiting results….

ConclusionsConclusions� Some evidence that increased compliance reduces morbidity Some evidence that increased compliance reduces morbidity

and mortalityand mortality� Various bundle modifications and interpretation of what Various bundle modifications and interpretation of what

constitutes ‘compliance’ constitutes ‘compliance’ � Bundle reliability not reported, but generally very low if Bundle reliability not reported, but generally very low if

calculated calculated � Inconsistent standards of reporting – makes comparison difficultInconsistent standards of reporting – makes comparison difficult� Early days: no studies reported using ‘new’ sepsis bundlesEarly days: no studies reported using ‘new’ sepsis bundles� Evidence leaning towards non-effectiveness of 6h EGDTEvidence leaning towards non-effectiveness of 6h EGDT� 3h sepsis bundle not evaluated3h sepsis bundle not evaluated

Dr Bronagh Blackwood UKProf Maureen Coombs NZA/Prof Sharon Irving USAProf Ruth Kleinpell USA

Bronte Martin, AustraliaProf Claire Rickard, AustraliaKathleen Vollman USAProf Ged Williams, UAE