BMJ Open · visual field testing, electroretinography), numerical data (e.g. autorefraction,...
Transcript of BMJ Open · visual field testing, electroretinography), numerical data (e.g. autorefraction,...
For peer review only
Assessing the efficacy of the electronic patient record system EDeR: Implementation study - Study protocol
Journal: BMJ Open
Manuscript ID: bmjopen-2012-002478
Article Type: Protocol
Date Submitted by the Author: 11-Dec-2012
Complete List of Authors: Job, Oliver; Eye Clinic Cantonal hospital of Lucerne 6000 Lucerne 16, Switzerland, Bachmann, Lucas; medignition Inc, Schmid, Martin; Eye Clinic Cantonal hospital of Lucerne 6000 Lucerne 16, Switzerland, Thiel, Michael; Eye Clinic Cantonal hospital of Lucerne 6000 Lucerne 16, Switzerland, Ivic, Sandra; medignition Inc,
<b>Primary Subject Heading</b>:
Medical management
Secondary Subject Heading: Ophthalmology
Keywords:
Information management < BIOTECHNOLOGY & BIOINFORMATICS, Information technology < BIOTECHNOLOGY & BIOINFORMATICS, Cataract and refractive surgery < OPHTHALMOLOGY, Glaucoma < OPHTHALMOLOGY, Neuro-ophthalmology < OPHTHALMOLOGY, Medical retina < OPHTHALMOLOGY
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open on June 5, 2020 by guest. P
rotected by copyright.http://bm
jopen.bmj.com
/B
MJ O
pen: first published as 10.1136/bmjopen-2012-002478 on 10 A
pril 2013. Dow
nloaded from
For peer review only
1
Assessing the efficacy of the electronic patient record system EDeR:
Implementation study - Study protocol
Oliver Job 1, Lucas M. Bachmann
2 *, Martin K Schmid
1, Michael A. Thiel
1, Sandra Ivic
2
[1]
Eye Clinic
Cantonal hospital of Lucerne
6000 Lucerne 16, Switzerland
[2]
Medignition Inc. Research Consultants
Postrasse 15, 6300 Zug, Switzerland
* Corresponding Author
Prof . Lucas M. Bachmann, MD, Ph.D. Medignition Inc, Postrasse 15, 6300 Zug, Switzerland
e-mail Adresses:
Oliver Job: [email protected]
Lucas M. Bachmann: [email protected]
Martin K. Schmid: [email protected]
Michael A. Thiel: [email protected]
Sandra Ivic: [email protected]
Page 1 of 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
2
ABSTRACT
Introduction Despite many innovations in information technology, many clinics still rely on
paper based medical records. Critics claim that they are hard to read, because of illegible
handwriting and uncomfortable to use. Moreover, a chronological overview is not always
easily possible, content can be destroyed or get lost. There is an overall opinion that electronic
medical records should solve these problems and improve physicians’ efficiency, patients’
safety and reduce overall costs in practice. However, to date, the evidence supporting this
view is sparse.
Methods and Analysis In this protocol we describe e a study exploring differences in speed
and accuracy when searching clinical information using the paper based patient record or the
EDeR. Designed as a randomised vignette study, we hypothesise that the EDeR increases
efficiency, i.e. reduces time on reading the patient history and looking for relevant
examination results, helps finding mistakes and missing information quicker and more
reliably. In exploratory analyses we aim at exploring factors associated with a higher
performance.
Ethics and Dissemination The ethics committee of the Canton Lucerne, Switzerland,
approved this study. We presume that the implementation of the electronic medical record
software EDeR will have a positive impact on the efficiency of the doctors, which will result
in an increase of consultations per day. We believe that the results of our study will provide a
valid basis to quantify the added value of an electronic medical record system in an
ophthalmologic environment.
238 words
Page 2 of 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
3
Article focus
To compare the gain in
• Efficiency
• Accuracy
When introducing an electronic patient record system vs. paper based records
Key messages
The chosen design
• has the potential to validly address the added value of an electronic patient record
system
• allows extrapolating the economic gain of introducing the system
Strengths and limitations
Study design allows examining the effects
• in a realistic clinical setting
• using relevant clinical problems
• using data of real patients
Economic consequences can only be addressed indirectly.
Page 3 of 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
4
BACKGROUND
Despite substantial effort to promote electronic data management in patient care, a large
proportion of healthcare providers still rely on paper based patient records. The American
Academy of Ophthalmology performed a survey concerning the adoption rate of electronic
medical records within their members and found an adoption rate of only 12%1.
The main problem with paper based patient records is the handwriting and the missing
overview of treatment and patient history. The handwriting is sometimes hard to read for the
physician in charge and the missing overview could potentially lead to needless examinations
and errors in diagnosis, prescription and treatment2, because of the absence of consistent
information on changes in state of health, treatment and the parameters measured. As a result
of these problems, there was the idea to support clinical daily routine with the help of
computers, as it is reality in many fields of service.
In the last decade, there was enormous development and investment in the field of
electronically supported medical records. It seems to be a key strategy to improve health care
in the different specialties3. Electronic medical records are supposed to improve physicians’
workflow and to offer a better overview on patient’s history. Today many authors believe that
this innovation should improve safety and reduce costs4.”A better understanding of the impact
of the electronic medical records on workflows is vital to understanding what the technology
really does offer that is new and unique” 5.
The cantonal hospital of Lucerne has developed new software addressing the special needs of
the ophthalmologic clinic. The idea was to create a tool, which offers a complete overview on
the patient history once a patient has been referred from another ophthalmologist or general
practitioner. The software includes physician’s drawings, imaging, measurement data (e.g.
visual field testing, electroretinography), numerical data (e.g. autorefraction, keratometry, and
biometry) and ophthalmic image data (e.g. fundus photography, optical coherence
Page 4 of 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
5
tomography). These data are essential to support clinical diagnosis to track disease
progression, and to plan treatment9.
In order to explore the impact on productivity of physicians using electronic health records, it
has been shown, that there is a statistically significant increase in the average monthly patient
visit volume of 9 visits per examined provider per month3, or the reduction of time spent on
administration by nursing staff6. The collected data also could be used for clinical research
and clinical studies7. In general, there is an existing hypothesis, that EMR should improve
patient’s safety, but there are neither data nor suggestions available, how this should be
measured. Overall however, there are only few studies, which show evidence, that the
implementation of electronic medical records has a major impact on the efficiency 4 8
,
especially in the field of ophthalmology. There seems to be also the demand for more studies
with prospective and randomized experimental designs (Fontaine et al), instead of surveys.
Possible barriers for a successful use of EMR are costs and time loss for implementation,
issues concerning the security and privacy of the patient’s data, lack in training and support of
the staff using the EMR software4 8
. The main objective should be the elimination of obstacles
for a successful implementation.
In this protocol we describe e a study exploring differences in speed and accuracy when
searching clinical information using the paper based patient record or the EDeR. Designed as
a randomised vignette study, we hypothesise that the EDeR increases efficiency, i.e. reduces
time on reading the patient history and looking for relevant examination results, helps finding
mistakes and missing information quicker and more reliably. In exploratory analyses we aim
at exploring factors associated with a higher performance.
Page 5 of 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
6
METHODS / DESIGN
Study design
The study is planned as a prospective simulated study.
Study location
The study is planned to take place at the cantonal hospital of Lucerne in the clinic of
ophthalmology.
Study population
The study population will be divided in two groups. One group will consist of senior
physicians of the ophthalmologic clinic of the cantonal hospital in Lucerne, Switzerland. The
other group will be formed by junior doctors, who are usually younger and more computer
adepts, but have less clinical experience.
Inclusion criteria
Ophthalmologist of the cantonal hospital Lucerne eye clinic who voluntarily agreed to
participate in the study
Experimental design
The study is planned as a prospective, randomised study using patient vignettes. The aim is to
measure the time needed to solve a given task and the accuracy of the result. For each out of
five different ophthalmologic sub-specialities (Neuro-ophthalmology, Corneal, Glaucoma,
Uveitis and orthoptics) we will create five tasks based on a patient example. Thus, we will
create 25 pairs of cases either represented in paper or electronic form.
We defined the following five task types: i) To detect the medication dose 3 months after
onset of treatment ii) Decision about the adjustment of a treatment based on the impairment
Page 6 of 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
7
or persistent state of data monitored iii) Question about the acquisition of underlying diseases.
iv), Question about the detection of implausible data, which could be important for the further
treatment, v) Question about the completeness of additional information. The Table provides
examples of tasks used for Neuro-Ophthalmology.
The patient model will come from original paper forms to assure that cases are comparable to
the real life situation. The cases, slightly modified, will be transferred from paper into the
electronic patient record system. During the experiment, a research assistant will monitor
physicians and measure the time required to complete each task. Each physician from each
division has to solve all of the tasks. The evaluation will take place at two time points;
immediately after implementation of EDeR and one year after.
Detailed Research Plan
a) Creating the patient vignettes
Of each specialization we will collect five typical patient records on paper.
b) Checking for plausibility and technical feasibility
After creation of the vignettes, two to three individuals will be selected to evaluate the tasks.
They will solve the tasks and afterwards fill out a questionnaire, which will consist of
questions about the difficulty, comprehensibility and technical feasibility of the task.
c) Design matrix
We will generate random sequence coding whether the task is started with the paper or the
electronic version of the patient record. A second random sequence will be generated to code
for the sequence with which tasks will be presented to the participants. These two measures
will eliminate order effects.
Page 7 of 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
8
d) At the assessment
The participant will be sitting in a consultation room and the assessor will be seated next to
him or her to avoid distractions.
We will design a reply form into which each participant enters his or her response to the task.
This reply form will be handed over to the assessor after completion. Time will be taken by
the assessor after completing each entry into this form.
There will be no communication between participant and assessor
e) The role of the assessor
1. The assessor will measure the time the participant needs to complete the reply form for
each task and each mode (paper based, electronic).
2. The assessor will take care of a protocol based procedure.
3. If any inconveniences of technical or administrative manner occur, the assessor will solve
them.
f) End of the assessment
All forms will be collected and the data will be entered in anonymised fashion into a
spreadsheet for further statistical analysis.
Sample size calculation
By implementing a novel electronic health record system it is possible to observe a
productivity gain of the physicians3. We hypothesise that our productivity gain should be one
additional consultation per day assuming that one consultation lasts about 20 minutes and a
physician sees 25 patients daily. Comparing the two methods, paper and electronic medical
record, the overall gain in time should be 60 seconds on average when using the electronic
medical record EDeR.
Page 8 of 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
9
Therefore, we are planning a study of the continuous response variable time from matched
pairs of study subjects. We presume that the difference in the response of matched pairs is
normally distributed with standard deviation 45. If the true difference in the mean time of
matched pairs is 60 seconds, we will need to study 7 pairs of subjects to be able to reject the
null hypothesis that this response difference is zero with probability (power) 0.8. The Type I
error probability associated with this test of this null hypothesis is 0.05.
Data collection
Point of time of data collection
The first data acquisition will take place shortly after the implementation of EDeR. The
defined tasks will be available in paper form and the physicians will fill them out with the
help a medical record, either the paper form or the electronic form.
Additionally, there will be someone from the study personal, who sits with the physicians in
the same room and measure the time they need to complete each task.
Baseline data
Baseline data will consist of information about the physicians’ age, extent of previous training
in ophthalmology, degree of specialization and self-reported level of general computing skills.
Follow-up data
At the second time point, one year after the first assessment, the experiment will be repeated
using the same protocol.
Page 9 of 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
10
Discussion
This paper describes the rationale, methodology and design of a randomised, prospective
simulation study using patient vignettes based on real patients. This study will compare the
paper based medical records with electronic medical record software EDeR in terms of
efficiency, safety and cost reduction in the daily routine of an ophthalmological clinic. The
experiment will consist of 25 tasks, which will be solved by two groups of participants; senior
and junior doctors.
We presume that the implementation of the electronic medical record software EDeR will
have a positive impact on the efficiency of the doctors, which will result in an increase of
consultations per day. We believe that the results of our study will provide a valid basis to
quantify the added value of an electronic medical record system in an ophthalmologic
environment.
Page 10 of 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
11
References
1. Chiang MF, Boland MV, Margolis JW, Lum F, Abramoff MD, Hildebrand PL. Adoption
and perceptions of electronic health record systems by ophthalmologists: an American
Academy of Ophthalmology survey. Ophthalmology 2008;115(9):1591-7; quiz 97 e1-
5.
2. Bates DW, Gawande AA. Improving safety with information technology. N Engl J Med
2003;348(25):2526-34.
3. Cheriff AD, Kapur AG, Qiu M, Cole CL. Physician productivity and the ambulatory EHR
in a large academic multi-specialty physician group. Int J Med Inform;79(7):492-500.
4. Black AD, Car J, Pagliari C, Anandan C, Cresswell K, Bokun T, et al. The impact of
eHealth on the quality and safety of health care: a systematic overview. PLoS
Med;8(1):e1000387.
5. Vishwanath A, Singh SR, Winkelstein P. The impact of electronic medical record systems
on outpatient workflows: a longitudinal evaluation of its workflow effects. Int J Med
Inform;79(11):778-91.
6. Poissant L, Pereira J, Tamblyn R, Kawasumi Y. The impact of electronic health records on
time efficiency of physicians and nurses: a systematic review. J Am Med Inform Assoc
2005;12(5):505-16.
7. Murphy EC, Ferris FL, 3rd, O'Donnell WR. An electronic medical records system for
clinical research and the EMR EDC interface. Invest Ophthalmol Vis Sci
2007;48(10):4383-9.
8. Fontaine P, Ross SE, Zink T, Schilling LM. Systematic review of health information
exchange in primary care practices. J Am Board Fam Med;23(5):655-70.
9. Chiang MF, Boland MV, Brewer A, Epley KD, Horton MB, Lim MC, et al. Special
requirements for electronic health record systems in ophthalmology.
Ophthalmology;118(8):1681-7.
Page 11 of 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
12
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
Lucas M. Bachmann was involved in the conception and design of the study, drafted the
protocol, supervised the revisions and approved the final manuscript. Oliver Job was involved
in conception and design of this study, revised the draft critically for intellectual content and
approved the final revised manuscript. Martin K Schmid and Michael A. Thiel were involved
in the conception of the study, revised the draft critically for intellectual content and approved
the final manuscript. Sandra Ivic was involved in the conception and design of the study,
drafted the protocol and approved the final manuscript.
Acknowledgments
None
Main Sponsors:
Cantonal hospital of Lucerne, 6000 Lucerne 16, Switzerland
Page 12 of 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
13
Table Description of context, task and contrast examined for the example of Neuro-
Ophthalmology
Task 1 Context Regular control visit Task To detect the medication dose after a certain of time after onset of treatment Contrast EdER: 6x daily 3 dragées à 60 mg Mestinon®
Paper record: 15x daily 3 dragées à 60 mg Mestinon®
Task 2 Context Regular control visit Task Decision about the adjustment of a treatment based on the impairment or persistent state of data
monitored - "In a case of Myasthenia gravis, the decision to adjust the steroid dose is made 4 weeks after starting therapy. If there has been a significant improvement in eye position, recorded with the coordimetry of Hess Weiss, the steroids can be tapered off. Is the eye position unchanged, or even worse, the steroid dose should be increased or an additional medication should be introduced. In the present case: Would you increase or decrease the dose of the steroids 4 weeks after onset of the therapy?"
Contrast EdER: There has been a significant improvement in the eye position; therefore the steroids can be tapered off. Paper record: The eye position is unchanged; the steroid dose should be increased.
Task 3 Context Post-op medication Task Question about the detection acquisition of underlying diseases.
Question: "In case of an allergic reaction to Bactrim® the administration of Diamox® often also produces allergic signs due to a cross-reaction. Diamox® should be avoided in such a situation. Is there a contraindication to administer Diamox® after a cataract surgery in the present case?"
Contrast EdER: The section with the patient history, containing information about allergies and underlying diseases, is blank. Although the patients has a sulfonamide allergy. Paper record: There is a short note about an allergic reaction to Bactrim®.
Task 4 Context Error in documentation Task "Do you find any implausible data regarding visual field recordings during the last 3 years of
documentation in the present case of a patient with a pituitary adenoma?"
Contrast EdER: The series of visual fields will show a constant bitemporal field Paper record: The series of visual fields will show one hemianopic visual field within the series of bitemporal fields. This change in just one visual field within an otherwise stable series is implausible and indicates an error of documentation)
Task 5 Context Completeness of clinical information Task In a patient with an idiopathic intracranial hypertension the diagnosis is made by the fact of
(bilateral) papilledema without visual afferent defects and raised intracranial opening pressure measured by lumbar puncture. Is it correct to make the diagnosis of an idiopathic intracranial hypertension in the present case?"
Contrast EdER: The necessary information to diagnose idiopathic intracranial hypertension is present. Paper record: No lumbar puncture, diagnosis not possible, because of missing information.
Page 13 of 13
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
Assessing the efficacy of the electronic patient record system EDeR: Implementation study - Study protocol
Journal: BMJ Open
Manuscript ID: bmjopen-2012-002478.R1
Article Type: Protocol
Date Submitted by the Author: 04-Mar-2013
Complete List of Authors: Job, Oliver; Eye Clinic Cantonal hospital of Lucerne 6000 Lucerne 16, Switzerland, Bachmann, Lucas; medignition Inc, Schmid, Martin; Eye Clinic Cantonal hospital of Lucerne 6000 Lucerne 16, Switzerland, Thiel, Michael; Eye Clinic Cantonal hospital of Lucerne 6000 Lucerne 16, Switzerland, Ivic, Sandra; medignition Inc,
<b>Primary Subject Heading</b>:
Medical management
Secondary Subject Heading: Ophthalmology
Keywords:
Information management < BIOTECHNOLOGY & BIOINFORMATICS, Information technology < BIOTECHNOLOGY & BIOINFORMATICS, Cataract and refractive surgery < OPHTHALMOLOGY, Glaucoma < OPHTHALMOLOGY, Neuro-ophthalmology < OPHTHALMOLOGY, Medical retina < OPHTHALMOLOGY
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open on June 5, 2020 by guest. P
rotected by copyright.http://bm
jopen.bmj.com
/B
MJ O
pen: first published as 10.1136/bmjopen-2012-002478 on 10 A
pril 2013. Dow
nloaded from
For peer review only
1
Assessing the efficacy of the electronic patient record system EDeR:
Implementation study - Study protocol
Oliver Job 1, Lucas M. Bachmann
2 *, Martin K Schmid
1, Michael A. Thiel
1, Sandra Ivic
2
[1]
Eye Clinic
Cantonal hospital of Lucerne
6000 Lucerne 16, Switzerland
[2]
Medignition Inc. Research Consultants
Postrasse 15, 6300 Zug, Switzerland
* Corresponding Author
Prof . Lucas M. Bachmann, MD, Ph.D. Medignition Inc, Postrasse 15, 6300 Zug, Switzerland
e-mail Adresses:
Oliver Job: [email protected]
Lucas M. Bachmann: [email protected]
Martin K. Schmid: [email protected]
Michael A. Thiel: [email protected]
Sandra Ivic: [email protected]
Page 1 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
2
ABSTRACT
Introduction Despite many innovations in information technology, many clinics still rely on
paper based medical records. Critics, however, claim that they are hard to read, because of
illegible handwriting, and uncomfortable to use. Moreover, a chronological overview is not
always easily possible, content can be destroyed or get lost. There is an overall opinion that
electronic medical records should solve these problems and improve physicians’ efficiency,
patients’ safety and reduce overall costs in practice. However, to date, the evidence
supporting this view is sparse.
Methods and Analysis In this protocol we describe a study exploring differences in speed
and accuracy when searching clinical information using the paper based patient record or the
EDeR. Designed as a randomised vignette study, we hypothesise that the EDeR increases
efficiency, i.e. reduces time on reading the patient history and looking for relevant
examination results, helps finding mistakes and missing information quicker and more
reliably. In exploratory analyses we aim at exploring factors associated with a higher
performance.
Ethics and Dissemination The ethics committee of the Canton Lucerne, Switzerland,
approved this study. We presume that the implementation of the electronic medical record
software EDeR will have a positive impact on the efficiency of the doctors, which will result
in an increase of consultations per day. We believe that the results of our study will provide a
valid basis to quantify the added value of an electronic medical record system in an
ophthalmologic environment.
238 words
Page 2 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
3
Article focus
To compare the gain in
• Efficiency
• Accuracy
When introducing an electronic patient record system vs. paper based records
Key messages
The chosen design
• has the potential to validly address the added value of an electronic patient record
system
• allows extrapolating the economic gain of introducing the system
Strengths and limitations
Study design allows examining the effects
• in a realistic clinical setting
• using relevant clinical problems
• using data of real patients
Economic consequences can only be addressed indirectly.
Page 3 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
4
BACKGROUND
Despite substantial effort to promote electronic data management in patient care, a large
proportion of healthcare providers still rely on paper based patient records. The American
Academy of Ophthalmology performed a survey concerning the adoption rate of electronic
medical records within their members and found an adoption rate of only 12% 1.
The main problem with paper based patient records is the handwriting and the missing
overview of treatment and patient history. The handwriting is sometimes hard to read for the
physician in charge and the missing overview could potentially lead to needless examinations
and errors in diagnosis, prescription and treatment 2, because of the absence of consistent
information on changes in state of health, treatment and the parameters measured. As a
solution of these problems, there was the idea to support clinical daily routine with the help of
computers, as it is reality in many fields of service.
In the last decade, there was enormous development and investment in the field of
electronically supported medical records. It seems to be a key strategy to improve health care
in the different specialties3. Electronic medical records are supposed to improve physicians’
workflow and to offer a better overview on patient’s history. Today many authors believe that
this innovation should improve safety and reduce costs 4.”A better understanding of the
impact of the electronic medical records on workflows is vital to understand what the
technology really does offer that is new and unique”5 6
. However, a systematic review
published in 2008 and covering the evidence until 2005 concluded that none of 20 studies
provided a formal cost effectiveness analysis of electronic patient record systems in general. 7
The cantonal hospital of Lucerne has developed new software addressing the special needs of
the ophthalmologic clinic. The idea was to create a tool, which offers a complete overview on
the patient history once a patient has been referred from another ophthalmologist or general
practitioner. The software includes physician’s drawings, imaging, measurement data (e.g.
Page 4 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
5
visual field testing, electroretinography), numerical data (e.g. autorefraction, keratometry, and
biometry) and ophthalmic image data (e.g. fundus photography, optical coherence
tomography). These data are essential to support clinical diagnosis to track disease
progression, and to plan treatment 8.
In order to explore the impact on productivity of physicians using electronic health records, it
has been shown, that there is a statistically significant increase in the average monthly patient
visit volume of 9 visits per examined provider per month 3, or the reduction of time spent on
administration by nursing staff 9. The collected data also could be used for clinical research
and clinical studies 10. In general, there is an existing hypothesis, that EMR should improve
patient’s safety, but there are neither data nor suggestions available, how this should be
measured. Overall however, there are only few studies, which show evidence, that the
implementation of electronic medical records has a major impact on the efficiency4 11
,
especially in the field of ophthalmology. There seems to be also the demand for more studies
with prospective and randomized experimental designs, instead of surveys 11
.
Possible barriers for a successful use of EMR are costs and time loss for implementation,
issues concerning the security and privacy of the patient’s data, lack in training and support of
the staff using the EMR software 4 11
. The main objective should be the elimination of
obstacles for a successful implementation.
In this protocol we describe a study exploring differences in speed and accuracy when
searching clinical information using the paper based patient record or the EDeR. Designed as
a randomised vignette study, we hypothesise that the EDeR increases efficiency, i.e. reduces
time on reading the patient history and looking for relevant examination results, helps finding
mistakes and missing information quicker and more reliably. In exploratory analyses we aim
at exploring factors associated with a higher performance.
Page 5 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
6
METHODS / DESIGN
Study design
The study is planned as a prospective simulated study.
Study location
The study is planned to take place at the cantonal hospital of Lucerne in the clinic of
ophthalmology. The Eye clinic is run by six consultant ophthalmologists, 12 senior
physicians, and 14 junior doctors.
Study population
The study population will be divided into two groups. One group will consist of senior
physicians of the ophthalmologic clinic of the cantonal hospital in Lucerne, Switzerland. The
other group will be formed by junior doctors, who are usually younger and more computer
adepts, but have less clinical experience.
Inclusion criteria
Ophthalmologist of the cantonal hospital Lucerne eye clinic who voluntarily agreed to
participate in the study will be included. Each participant will receive a general introduction
on the usage of the EDeR of half an hour prior to the assessment. The training will cover
aspects of data collection and saving very briefly. The training will be such, that the basic
functions of EDeR are known.
Primary outcome
Gain in speed, when performing the task using the EDeR system
Secondary outcome
Accuracy of replies given using the EDeR system vs. the paper record
Page 6 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
7
Experimental design
The study is planned as a prospective, randomised study using patient vignettes. The aim is to
measure the time needed to solve a given task and the accuracy of the result. For each out of
five different ophthalmologic sub-specialities (Neuro-ophthalmology, Corneal, Glaucoma,
Uveitis and orthoptics) we will create five tasks based on a patient example. Thus, we will
create 25 pairs of cases either represented in paper or electronic form. The tasks are chosen on
the basis of the relevance in clinical practice.
Type of tasks
We defined the following five task types: i) To detect the medication dose 3 months after
onset of treatment ii) Decision about the adjustment of a treatment based on the impairment
or persistent state of data monitored iii) Question about the acquisition of underlying diseases.
iv), Question about the detection of implausible data, which could be important for the further
treatment, v) Question about the completeness of additional information. Table 1 provides
examples of tasks used for Neuro-Ophthalmology.
The patient model will come from original paper forms to assure that cases are comparable to
the real life situation. The full patient record, containing a slight modification for the contrast
examined in a specific task, will be transferred from paper into the electronic patient record
system. During the experiment, a research assistant will monitor physicians and measure the
time required to complete each task. Each physician from each division has to solve all of the
tasks. The evaluation will take place at two time points; immediately after implementation of
EDeR and one year after.
Design Matrix Summary
This is a two group (EDeR vs. paper based) cross-over (randomly assigning the sequence
electronic vs. paper-based per task) study. Time (primary endpoint) and accuracy (secondary
endpoint) are the two outcome variables. Intervention is the EDeR that is compared with the
Page 7 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
8
paper-based patient record. We assess tasks within 5 clinical subgroups (without examining
subgroup effects) and perform the experiment in two groups with different ophthalmologic
expertise (junior doctors vs. senior doctors). Each assessor will complete 50 tasks (i.e. 25
tasks based on the electronic patient record and 25 based on the paper patient record).
Electronic and paper-based vignettes are slightly different for each task. If an assessor checks
the same vignette a second time (e. g. in the electronic version), he or she must read the
content again, because the answer is different to the first time (e. g. in the paper-based
version). The material for the paper-based-group comes from real patients records and the
tasks were selected based on clinical relevance. For Summary see Figure 1
Detailed Research Plan
a) Creating the patient vignettes
Of each specialization we will collect five typical patient records on paper.
b) Checking for plausibility and technical feasibility
After creation of the vignettes, two to three individuals will be selected to evaluate the tasks.
They will solve the tasks and afterwards fill out a questionnaire, which will consist of
questions about the difficulty, comprehensibility and technical feasibility of the task.
c) Design matrix
We will generate random sequence coding whether the task is started with the paper or the
electronic version of the patient record. A second random sequence will be generated to code
for the sequence with which tasks will be presented to the participants. These two measures
will eliminate order effects.
d) At the assessment
Page 8 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
9
The participant will be sitting in a consultation room and the assessor will be seated next to
him or her to avoid distractions.
We will design a reply form into which each participant enters his or her response to the task.
This reply form will be handed over to the assessor after completion. Time will be taken by
the assessor after completing each entry into this form. In total, each participant will go
through 25 pairs of tasks. In case of a technical problem with the electronic system, the
assessment will be suspended and the remaining tasks will be completed after restoration of
the system.
There will be no communication between participant and assessor
e) The role of the assessor
1. The assessor will measure the time the participant needs to complete the reply form for
each task and each mode (paper based, electronic).
2. The assessor will take care of a protocol based procedure.
3. If any inconveniences of technical or administrative manner occur, the assessor will solve
them.
f) End of the assessment
All forms will be collected and the data will be entered in anonymised fashion into a
spreadsheet for further statistical analysis.
Sample size calculation
By implementing a novel electronic health record system it is possible to observe a
productivity gain of the physician 3. We hypothesise that our productivity gain should be one
additional consultation per day assuming that one consultation lasts about 20 minutes and a
physician sees 25 patients daily. Comparing the two methods, paper and electronic medical
Page 9 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
10
record, the overall gain in time should be 60 seconds on average when using the electronic
medical record EDeR.
Therefore, we are planning a study of the continuous response variable time from matched
pairs of study subjects. We presume that the difference in the response of matched pairs is
normally distributed with standard deviation 45. If the true difference in the mean time of
matched pairs is 60 seconds, we will need to study 7 pairs of subjects to be able to reject the
null hypothesis that this response difference is zero with probability (power) 0.8. The Type I
error probability associated with this test of this null hypothesis is 0.05.
At the time of writing this protocol 10 senior physicians and 9 junior doctors were willing to
participate in the study.
Data collection
Point of time of data collection
The first data acquisition will take place shortly after the implementation of EDeR. The
defined tasks will be available in paper form and the physicians will fill them out with the
help of a medical record, either the paper form or the electronic form.
Additionally, there will be someone from the study personnel, who sits with the physicians in
the same room and measure the time they need to complete each task.
Baseline data
Prior to the examination baseline data will be collected of each participant. Baseline data will
consist of information about the physicians’ age, extent of previous training in
ophthalmology, degree of specialization and self-reported level of general computing skills.
Follow-up data
Page 10 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
11
At the second time point, one year after the first assessment, the experiment will be repeated
using the same protocol. In the interval between the two measurements, all physicians will be
mainly working with the EDeR system.
Page 11 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
12
Discussion
This paper describes the rationale, methodology and design of a randomised, prospective
simulation study using patient vignettes based on real patients. This study will compare the
paper based medical records with electronic medical record software EDeR in terms of
efficiency, safety and cost reduction in the daily routine of an ophthalmological clinic. The
experiment will consist of 25 tasks, which will be solved by two groups of participants; senior
and junior doctors.
We presume that the implementation of the electronic medical record software EDeR will
have a positive impact on the efficiency of the doctors, which will result in an increase of
consultations per day. We believe that the results of our study will provide a valid basis to
quantify the added value of an electronic medical record system in an ophthalmologic
environment.
Page 12 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
13
References
1. Chiang MF, Boland MV, Margolis JW, Lum F, Abramoff MD, Hildebrand PL. Adoption
and perceptions of electronic health record systems by ophthalmologists: an American
Academy of Ophthalmology survey. Ophthalmology 2008;115(9):1591-7; quiz 1597
e1-5.
2. Bates DW, Gawande AA. Improving safety with information technology. N Engl J Med
2003;348(25):2526-34.
3. Cheriff AD, Kapur AG, Qiu M, Cole CL. Physician productivity and the ambulatory EHR
in a large academic multi-specialty physician group. Int J Med Inform 2010;79(7):492-
500.
4. Black AD, Car J, Pagliari C, Anandan C, Cresswell K, Bokun T, et al. The impact of
eHealth on the quality and safety of health care: a systematic overview. PLoS Med
2011;8(1):e1000387.
5. Vishwanath A, Singh SR, Winkelstein P. The impact of electronic medical record systems
on outpatient workflows: a longitudinal evaluation of its workflow effects. Int J Med
Inform 2010;79(11):778-91.
6. Vishwanath A, Winkelstein P, Singh SR. The impact of electronic health records on
outpatient flow: an evaluation of pre- and post-implementation expectations. AMIA
Annu Symp Proc 2006:1130.
7. Uslu AM, Stausberg J. Value of the electronic patient record: an analysis of the literature. J
Biomed Inform 2008;41(4):675-82.
8. Chiang MF, Boland MV, Brewer A, Epley KD, Horton MB, Lim MC, et al. Special
requirements for electronic health record systems in ophthalmology. Ophthalmology
2011;118(8):1681-7.
9. Poissant L, Pereira J, Tamblyn R, Kawasumi Y. The impact of electronic health records on
time efficiency of physicians and nurses: a systematic review. J Am Med Inform Assoc
2005;12(5):505-16.
10. Murphy EC, Ferris FL, 3rd, O'Donnell WR. An electronic medical records system for
clinical research and the EMR EDC interface. Invest Ophthalmol Vis Sci
2007;48(10):4383-9.
11. Fontaine P, Ross SE, Zink T, Schilling LM. Systematic review of health information
exchange in primary care practices. J Am Board Fam Med 2010;23(5):655-70.
Page 13 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
14
Competing interests
The authors declare that they have no competing interests.
Conflict of interest
None declared
Authors' contributions
Lucas M. Bachmann was involved in the conception and design of the study, drafted the
protocol, supervised the revisions and approved the final manuscript. Oliver Job was involved
in conception and design of this study, revised the draft critically for intellectual content and
approved the final revised manuscript. Martin K Schmid and Michael A. Thiel were involved
in the conception of the study, revised the draft critically for intellectual content and approved
the final manuscript. Sandra Ivic was involved in the conception and design of the study,
drafted the protocol and approved the final manuscript.
Acknowledgments
None
Main Sponsors:
Cantonal hospital of Lucerne, 6000 Lucerne 16, Switzerland
Page 14 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
15
Table 1 Description of context, task and contrast examined for the example of Neuro-
Ophthalmology
Task 1 Context Regular control visit Task To detect the medication dose after a certain of time after onset of treatment Contrast EdER: 6x daily 3 dragées à 60 mg Mestinon®
Paper record: 15x daily 3 dragées à 60 mg Mestinon®
Task 2 Context Regular control visit Task Decision about the adjustment of a treatment based on the impairment or persistent state of data
monitored - "In a case of Myasthenia gravis, the decision to adjust the steroid dose is made 4 weeks after starting therapy. If there has been a significant improvement in eye position, recorded with the coordimetry of Hess Weiss, the steroids can be tapered off. Is the eye position unchanged, or even worse, the steroid dose should be increased or an additional medication should be introduced. In the present case: Would you increase or decrease the dose of the steroids 4 weeks after onset of the therapy?"
Contrast EdER: There has been a significant improvement in the eye position; therefore the steroids can be tapered off. Paper record: The eye position is unchanged; the steroid dose should be increased.
Task 3 Context Post-op medication Task Question about the detection acquisition of underlying diseases.
Question: "In case of an allergic reaction to Bactrim® the administration of Diamox® often also produces allergic signs due to a cross-reaction. Diamox® should be avoided in such a situation. Is there a contraindication to administer Diamox® after a cataract surgery in the present case?"
Contrast EdER: The section with the patient history, containing information about allergies and underlying diseases, is blank. Although the patients has a sulfonamide allergy. Paper record: There is a short note about an allergic reaction to Bactrim®.
Task 4 Context Error in documentation Task "Do you find any implausible data regarding visual field recordings during the last 3 years of
documentation in the present case of a patient with a pituitary adenoma?"
Contrast EdER: The series of visual fields will show a constant bitemporal field Paper record: The series of visual fields will show one hemianopic visual field within the series of bitemporal fields. This change in just one visual field within an otherwise stable series is implausible and indicates an error of documentation)
Task 5 Context Completeness of clinical information Task In a patient with an idiopathic intracranial hypertension the diagnosis is made by the fact of
(bilateral) papilledema without visual afferent defects and raised intracranial opening pressure measured by lumbar puncture. Is it correct to make the diagnosis of an idiopathic intracranial hypertension in the present case?"
Contrast EdER: The necessary information to diagnose idiopathic intracranial hypertension is present. Paper record: No lumbar puncture, diagnosis not possible, because of missing information.
Page 15 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
16
Figure 1 Study Flow and Design Features
Page 16 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
# Reviewer: M.I. Lopez Galvez Our reply
1 The article only includes the protocol not the results. It is not enough
Thank you. This is the protocol for a study that will be still executed. BMJ open offers publication of protocols and we think that this is very useful to share and discuss methods of studies before their execution.
2 I suggest to publish it with the results with the analysis included There is no conclusion
Please see our reply to query 1
Reviewer: Frank Larkin, Our reply
1 It may be inappropriate to evaluate clinician performance using the electronic record before any necessary learning time using the software has elapsed. I realise that the planned second analysis at one year would probably identify any such effect due to inexperience.
Thank you for the favorable assessment of our proposal.
2 Refs #3-5, 8, 9 are incomplete (year of publication). Thank you – the references have been revised.
3 This manuscript describes the design of a prospective simulated study to examine the factors associated with improved performance using an electronic clinical record versus paper records. They term this a randomised vignette study. The design is based on specific clinical decision points in consultations in a range of ophthalmology sub-specialties. While ophthalmology is itself a small specialty, it does lend itself to electronic patient record keeping better than many others for a number of reasons, including the high proportion of images which are used as diagrams or digital photographs in records (for example fundoscopy findings). My opinion is that the findings in this type of study would be of broad interest in clinical medicine more generally.
Thank you! No reply required.
4 Having indicated my support for this study, it would be of much higher value if the results were included. I encourage the authors to add their findings and discuss these. The manuscript is short as submitted.
Thank you. We would like to refer to our reply to query number 1 of Dr. Lopez Galvez
5 As the authors indicate, there has been little research previously published in this field. More critical analysis of those previous reports which they cite would improve the manuscript.
Thank you – no reply required.
6 Methods & design, P9 line 7: Is it valid to assume standard distribution of any differences in responses to matched vignette pairs?
Thank you. We checked again with our statistician and he confirmed the adequacy of our approach.
7 Time of data collection, P9 line 22: If the authors have not set out to examine for any clinician learning effect with software, why not assess competence with software before the first comparative analysis?
Thank you for this important query. We fully agree and clarified this point in the revised paper. (p.8, 1
st para) “Prior to the examination
baseline data will be collected of each participant.”
Page 17 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
Reviewer Jürgen Stausberg Our reply
1 The authors present a well-thought-out study protocol to assess the electronic patient record system EDeR to be implemented in their ophthalmological department. Main focus of this study is to “explore differences in speed and accuracy when searching clinical information using the paper based patient record or the EDeR”. The authors also talk about patient´s safety, which should be improved by an electronic medical record (EMR), and the possible barriers for a successful use of the EMR.
Thank you for the favorable assessment of our protocol. No reply required.
2 As far as the reviewer understands, the plan is as follows. Patient vignettes are the observational units. The study measures the time needed to solve tasks related to those vignettes. The reviewer assumes that tasks and vignettes are related 1 to 1. Otherwise, if a patient vignette is used for several tasks, the tasks become observational unit. Time is the outcome variable. Intervention is the EMR that is compared with the paper-based patient record. Therefore, the study implements a two group design. The authors assume that the intervention saves 60 seconds time in the mean with a standard deviation of 45 seconds. Instead of randomly assigning vignettes to the groups, the study implements a crossover design in randomly assigning the sequence (EMR first, paper second vs. paper first EMR second) to vignettes. That design is underpinned by a formal correct sample size calculation leading to 7 observational units needed. To make it more complex, the authors include subgroups as well as several observers for each measurement. They include 5 subgroups for clinical specialties related to the patient vignettes and 2 subgroups of level of expertise related to the observers. At least, they will have 2 x 5 x 5 (five tasks each clinical subspecialty) x 2 (EMR and paper) = 500 measurements. To address less reliability having only one observer in each subgroup (experts and beginners) the authors intend to recruit more than the two for the study. To make it more serious, the vignettes are slightly modified according to the group. So if an observer checks the same vignette a second time (e. g. in the EMR), he must read the content again, because the answer is different to the first time (e. g. in the paper-based version). The material for the paper-based-group comes from real patients records; the material for the EMR-group is entered into the EMR specifically for the study.
This is an excellent analysis of the proposed method. Thank you – no reply required.
Page 18 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
3 This long description makes clear that the proposed study protocol characterizes a complex study design. Unfortunately, the complexity is not well addressed in the text. Furthermore, it is unclear whether the design is really appropriate to answer the research question. Therefore I recommend to improve the readability of the text by clarifying the points made above on the one hand and to rethink the study design according to the research question on the other hand.
Thank you for this query – we agree that the description was too brief. We amended the protocol as follows lending substantially from Prof Stausberg’s excellent summary in point 2: (p 8, last. para.) “Design Matrix Summary
This is a two group (EDeR vs. paper based) cross-over (randomly assigning the sequence electronic vs. paper-based per task) study. Time (primary endpoint) and accuracy (secondary endpoint) are the two outcome variables. Intervention is the EDeR that is compared with the paper-based patient record. We assess tasks within 5 clinical subgroups (without examining subgroup effects) and perform the experiment in two groups with different ophthalmologic expertise (junior doctors vs. senior doctors). Each assessor will complete 50 tasks (i.e. 25 tasks based on the electronic patient record and 25 based on the paper patient record). Electronic and paper-based vignettes are slightly different for each task. If an assessor checks the same vignette a second time (e. g. in the electronic version), he or she must read the content again, because the answer is different to the first time (e. g. in the paper-based version). The material for the paper-based-group comes from real patients records and the tasks were selected based on clinical relevance.”
4 It would be helpful if the authors refer to an established schema for study design rather than to describe the details in own words.
Thank you – we are unaware of a similar application and would be grateful, if you could indicate a paper covering the same topic and applying a similar method.
5 However, the study idea of the authors in assessing the efficacy of EDeR is essential and a current and important topic that must be discussed. It offers a good basis to explore other essential usability and efficacy questions regarding the use of EMRs.
Thank you, no reply required.
6 As long as the time needed to fulfill the tasks is the only measurement, the study is solely related to efficacy as the authors denote time saving. Then, the study does not measure accuracy. Consequently, accuracy should be deleted a study goal.
We fully agree that the protocol did not clearly specify that accuracy was a secondary outcome of this study. We amended the protocol as follows. (p. 6. 3 para.) “Primary outcome
Gain in speed, when performing the task using the EDeR system
Secondary outcome
Accuracy of replies given using the EDeR system vs. the paper record
”
7 A figure should be added to explain the several entities in the study (vignettes, tasks, sub-specialties, user groups).
We agree with the Reviewer and added a figure to the amended manuscript.
Page 19 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
8 A major concern with the study design is a potential bias introduced by the unblinded recording of the vignettes into the EMR. The reviewer assumes that the result is an optimized representation of the vignettes in the EMR. This optimized representation will be different from the information available in the EMR in routine use. Comparing the optimized representation of the vignettes in the EMR with an ordinary representation in a paper-based-record is unfair. Addressing this potential bias in the study design is mandatory.
Thank you. The Reviewer raises an important issue. Data managers at the Hospital will enter the complete available patient record into EDeR. Thus, the content will be identical in both exposures. We agree that this aspect was not clearly stated in the initial submission. We therefore revised the paper as follows: (p. 7, 2. para) “The patient model will come from original paper forms to assure that cases are comparable to the real life situation. The full patient record, containing a slight modification for the contrast examined in a specific task, will be transferred from paper into the electronic patient record system.”
9 The authors should clarify whether the participants of the study will be trained or not in the implemented EDeR prior to the assessment.
Thank you. We fully agree and amended the protocol as follows: (p.6, 5 para)“
Ophthalmologist of the cantonal hospital
Lucerne eye clinic who voluntarily agreed to participate in the study will be included. Each participant will receive a general introduction on the usage of the EDeR of half an hour prior to the assessment.”
10 Study population: How big the study population is or will be? For instance, how many ophthalmologists are found in the eye clinic of the cantonal hospital Lucerne? How many junior and senior ophthalmologists are there? How many ophthalmologists already voluntarily agreed to participate in the study? What is the expectation of the recruitment rate?
Thank you for this query. We provide some information in the revised paper. (p.6. 2
nd para)
“The study is planned to take place at the cantonal hospital of Lucerne in the clinic of ophthalmology. The Eye clinic is run by six consultant ophthalmologists, 12 senior physicians, and 14 junior doctors.”
And (p. 9, 3rd para) “At the time of writing this
protocol 10 senior physicians and 9 junior doctors were willing to participate in the study.”
11 For each out of five different ophthalmologic sub-specialties, the authors will create five tasks based on a patient example. Did some literature lead the choice of these five task types? Or how did the authors come to the definition of the task types? What motivated this choice? The authors should specify if these five tasks are the main consultation´s tasks in their ophthalmology department or if these tasks are generally error-prone when using the paper-based form.
Thank you again for this important query. We specified the selection of topics in the amended manuscript as follows: (p.7, 1
st para) “The tasks
are chosen on the basis of the relevance in clinical practice.”
12 Each physician from each division has to solve all of the tasks. So, a physician from a specific division will have to solve <5 or 25?> tasks using paper-based records and <5 or 25?> tasks using the EDeR in a randomized sequence. Please specify if the physicians will have to solve all <10 or 50?> tasks at once, that is without break.
Thank you. We clarified this as follows: (p.8, 3rd
para) “We will design a reply form into which each participant enters his or her response to the task. This reply form will be handed over to the assessor after completion. Time will be taken by the assessor after completing each entry into this form. In total, each participant will go through 25 pairs of tasks.”
Page 20 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
13 “The evaluation will take place at two time points; immediately after implementation of EDeR and one year after.” Please specify if the physicians are already trained in EDeR at the first time point. What about the second time point (one year after)? Also specify if the physicians will be using EDeR during that year or would you keep training them to prepare them to the second evaluation?
Please see our reply to your query 9. Before the first exposure, only a general introduction on the usage of EDeR will be provided that will last about 30 minutes. The training will cover aspects of data collection and saving very briefly. The training will be such, that the basic functions are known. We amended the paper as follows. (p. 6, 3
rd para.) “The training will cover
aspects of data collection and saving very briefly. The training will be such, that the basic functions of the EDeR are known.”
On page 10, 3rd para we also revised as follows:
“At the second time point, one year after the first assessment, the experiment will be repeated using the same protocol. In the interval between the two measurements, all physicians will be mainly working with the EDeR system.”
14 At the assessment/the role of the assessor. Let´s suppose a technical problem occurs after a physician has already performed 8 tasks. This problem unfortunately forces the break of the assessment. Shall the physician start all tasks from the beginning at another time point or may he perform only the last tasks later?
Thank you very much for this important query. We fully agree and added this statement on p.9 1st para: “In case of a technical problem with
the electronic system, the assessment will be suspended and the remaining tasks will be completed after restoration of the system.”
15 Please check, if the sample size estimation holds true in view of the several subgroups as well as the issue of multiple testing introduced with the subgroups.
We will only make statistical comparisons on the main effect. Subgroup analyses will be performed in exploratory analyses. Any subgroup result will be marked in the paper as such and the issue of power will be discussed.
16 Having 20 minutes time for a whole consultation, 60 seconds time reduction with a single task seems unrealistic. Please clarify how the total time reduction for a whole consultation is estimated by observing small tasks.
The reviewer raises an interesting question. We performed a small survey among the senior staff at our institution and there was a consensus that a meaningful difference should be at least 1 Minute / Patient. Therefore, this delta was chosen. We do agree that the advantage in terms of speed will not necessarily come from one single task in clinical practice. But still this could be the case for example when checking the number of interventions a patient received over a longer period of time combined with a judgment of treatment effect. Arguably, there is no strong basis for our statistical assumptions but they seem rational to us. Therefore, we would like to keep the paper unchanged on this point.
17 “We presume that the implementation of the electronic medical record software EDeR will have a positive impact on the efficiency of the doctors, which will result in an increase of consultations per day.” Good and fine, but what is more important: the increase of consultations per day (through decrease of consultation time: 1 min gained) or the improvement of patients healthcare (improved quality of consultation, patient safety)?
Please also see our reply to your query 6. We agree, therefore accuracy is assessed as a secondary outcome in this study. We realize that accuracy is only an indirect measure of “improved quality of consultation and patient safety” and will design further studies investigating this more directly, once this study showed a superiority of EDeR. We kept the manuscript unchanged.
Page 21 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
18 Abstract – Introduction: replace the second sentence with “Critics, however, claim that they are hard to read, because of illegible handwriting, and uncomfortable to use.”
Thank you. We revised the Abstract accordingly.
19 Abstract – Methods and Analysis: there is an 'e' too much in the first sentence.
Well spotted – thank you!
20 Background: second paragraph, last sentence: ‘As a result of these problems’ should be ‘As a solution to these problems’.
Thank you – we revised the paper accordingly.
21 Background: third paragraph, last sentence: the second word 'understanding' should be 'understand'.
Thank you – we revised the paper accordingly.
22 Last paragraph, first sentence: there is an 'e' too much in the first sentence (In this protocol we describe e a study exploring…).
Thank you – we revised the paper accordingly.
23 Methods/Design – Study population, first sentence: change ‘in’ in ‘into’.
Thank you – we revised the paper accordingly.
24 Methods/Design – Data collection – Point of time of data collection: Second part of the second sentence should be rewritten. Did you mean: … with the help ‘of‘ a medical record?
Thank you – we revised the paper accordingly.
25 Even if there is just one table provided in the paper, the authors should give it a number when referencing to it.
Thank you – we revised the paper accordingly.
26 References related to the evaluation of medical informatics interventions are missing.
Thank you – we checked again and added the reference of your excellent 2008 Systematic Review published in J Biomed Inform. to the background section. We also added some additional references to underpin our statements better.
27 The selection of references according to the research question is arbitrary.
Please see our reply to your query 26
28 It is unclear whether a conflict of interest exists due to the involvement of the authors in the development of the EMR-system. Please include a statement related to the involvement of the authors.
Thank you – we fully agree and amended the paper accordingly. We declare to have no conflict of interest. (p.13, last para)
Page 22 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
1
Assessing the efficacy of the electronic patient record system EDeR:
Implementation study - Study protocol
Oliver Job 1, Lucas M. Bachmann 2 *, Martin K Schmid 1, Michael A. Thiel 1, Sandra Ivic 2
[1]
Eye Clinic
Cantonal hospital of Lucerne
6000 Lucerne 16, Switzerland
[2]
Medignition Inc. Research Consultants
Postrasse 15, 6300 Zug, Switzerland
* Corresponding Author
Prof . Lucas M. Bachmann, MD, Ph.D. Medignition Inc, Postrasse 15, 6300 Zug, Switzerland
e-mail Adresses:
Oliver Job: [email protected]
Lucas M. Bachmann: [email protected]
Martin K. Schmid: [email protected]
Michael A. Thiel: [email protected]
Sandra Ivic: [email protected]
Page 23 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
2
ABSTRACT
Introduction Despite many innovations in information technology, many clinics still rely on
paper based medical records. Critics, however, claim that they are hard to read, because of
illegible handwriting, and uncomfortable to use. Moreover, a chronological overview is not
always easily possible, content can be destroyed or get lost. There is an overall opinion that
electronic medical records should solve these problems and improve physicians’ efficiency,
patients’ safety and reduce overall costs in practice. However, to date, the evidence
supporting this view is sparse.
Methods and Analysis In this protocol we describe e a study exploring differences in speed
and accuracy when searching clinical information using the paper based patient record or the
EDeR. Designed as a randomised vignette study, we hypothesise that the EDeR increases
efficiency, i.e. reduces time on reading the patient history and looking for relevant
examination results, helps finding mistakes and missing information quicker and more
reliably. In exploratory analyses we aim at exploring factors associated with a higher
performance.
Ethics and Dissemination The ethics committee of the Canton Lucerne, Switzerland,
approved this study. We presume that the implementation of the electronic medical record
software EDeR will have a positive impact on the efficiency of the doctors, which will result
in an increase of consultations per day. We believe that the results of our study will provide a
valid basis to quantify the added value of an electronic medical record system in an
ophthalmologic environment.
238 words
Page 24 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
3
Article focus
To compare the gain in
• Efficiency
• Accuracy
When introducing an electronic patient record system vs. paper based records
Key messages
The chosen design
• has the potential to validly address the added value of an electronic patient record
system
• allows extrapolating the economic gain of introducing the system
Strengths and limitations
Study design allows examining the effects
• in a realistic clinical setting
• using relevant clinical problems
• using data of real patients
Economic consequences can only be addressed indirectly.
Page 25 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
4
BACKGROUND
Despite substantial effort to promote electronic data management in patient care, a large
proportion of healthcare providers still rely on paper based patient records. The American
Academy of Ophthalmology performed a survey concerning the adoption rate of electronic
medical records within their members and found an adoption rate of only 12%1.%
1.
The main problem with paper based patient records is the handwriting and the missing
overview of treatment and patient history. The handwriting is sometimes hard to read for the
physician in charge and the missing overview could potentially lead to needless examinations
and errors in diagnosis, prescription and treatment2, 2, because of the absence of consistent
information on changes in state of health, treatment and the parameters measured. As a
resultsolution of these problems, there was the idea to support clinical daily routine with the
help of computers, as it is reality in many fields of service.
In the last decade, there was enormous development and investment in the field of
electronically supported medical records. It seems to be a key strategy to improve health care
in the different specialties3. Electronic medical records are supposed to improve physicians’
workflow and to offer a better overview on patient’s history. Today many authors believe that
this innovation should improve safety and reduce costsIt seems to be a key strategy to
improve health care in the different specialties3. Electronic medical records are supposed to
improve physicians’ workflow and to offer a better overview on patient’s history. Today
many authors believe that this innovation should improve safety and reduce costs4.”A better
understanding of the impact of the electronic medical records on workflows is vital to
understanding what the technology really does offer that is new and unique” 5.
4.”A better
understanding of the impact of the electronic medical records on workflows is vital to
understand what the technology really does offer that is new and unique”5 6
. However, a
systematic review published in 2008 and covering the evidence until 2005 concluded that
Page 26 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
5
none of 20 studies provided a formal cost effectiveness analysis of electronic patient record
systems in general. 7
The cantonal hospital of Lucerne has developed new software addressing the special needs of
the ophthalmologic clinic. The idea was to create a tool, which offers a complete overview on
the patient history once a patient has been referred from another ophthalmologist or general
practitioner. The software includes physician’s drawings, imaging, measurement data (e.g.
visual field testing, electroretinography), numerical data (e.g. autorefraction, keratometry, and
biometry) and ophthalmic image data (e.g. fundus photography, optical coherence
tomography). These data are essential to support clinical diagnosis to track disease
progression, and to plan treatment9. 8.
In order to explore the impact on productivity of physicians using electronic health records, it
has been shown, that there is a statistically significant increase in the average monthly patient
visit volume of 9 visits per examined provider per month3, or the reduction of time spent on
administration by nursing staff6. The collected data also could be used for clinical research
and clinical studies7. In general, there is an existing hypothesis, that EMR should improve
patient’s safety, but there are neither data nor suggestions available, how this should be
measured. Overall however, there are only few studies, which show evidence, that the
implementation of electronic medical records has a major impact on the efficiency In order to
explore the impact on productivity of physicians using electronic health records, it has been
shown, that there is a statistically significant increase in the average monthly patient visit
volume of 9 visits per examined provider per month 3, or the reduction of time spent on
administration by nursing staff 9. The collected data also could be used for clinical research
and clinical studies 10. In general, there is an existing hypothesis, that EMR should improve
patient’s safety, but there are neither data nor suggestions available, how this should be
measured. Overall however, there are only few studies, which show evidence, that the
Page 27 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
6
implementation of electronic medical records has a major impact on the efficiency4 84 11,
especially in the field of ophthalmology. There seems to be also the demand for more studies
with prospective and randomized experimental designs (Fontaine et al), instead of surveys.
, especially in the field of ophthalmology. There seems to be also the demand for more studies
with prospective and randomized experimental designs, instead of surveys 11.
Possible barriers for a successful use of EMR are costs and time loss for implementation,
issues concerning the security and privacy of the patient’s data, lack in training and support of
the staff using the EMR software 4 84 11
. The main objective should be the elimination of
obstacles for a successful implementation.
In this protocol we describe e a study exploring differences in speed and accuracy when
searching clinical information using the paper based patient record or the EDeR. Designed as
a randomised vignette study, we hypothesise that the EDeR increases efficiency, i.e. reduces
time on reading the patient history and looking for relevant examination results, helps finding
mistakes and missing information quicker and more reliably. In exploratory analyses we aim
at exploring factors associated with a higher performance.
Page 28 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
7
METHODS / DESIGN
Study design
The study is planned as a prospective simulated study.
Study location
The study is planned to take place at the cantonal hospital of Lucerne in the clinic of
ophthalmology. The Eye clinic is run by six consultant ophthalmologists, 12 senior
physicians, and 14 junior doctors.
Study population
The study population will be divided ininto two groups. One group will consist of senior
physicians of the ophthalmologic clinic of the cantonal hospital in Lucerne, Switzerland. The
other group will be formed by junior doctors, who are usually younger and more computer
adepts, but have less clinical experience.
Inclusion criteria
Ophthalmologist of the cantonal hospital Lucerne eye clinic who voluntarily agreed to
participate in the study will be included. Each participant will receive a general introduction
on the usage of the EDeR of half an hour prior to the assessment. The training will cover
aspects of data collection and saving very briefly. The training will be such, that the basic
functions of EDeR are known.
Primary outcome
Gain in speed, when performing the task using the EDeR system
Secondary outcome
Accuracy of replies given using the EDeR system vs. the paper record
Page 29 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
8
Experimental design
The study is planned as a prospective, randomised study using patient vignettes. The aim is to
measure the time needed to solve a given task and the accuracy of the result. For each out of
five different ophthalmologic sub-specialities (Neuro-ophthalmology, Corneal, Glaucoma,
Uveitis and orthoptics) we will create five tasks based on a patient example. Thus, we will
create 25 pairs of cases either represented in paper or electronic form. The tasks are chosen on
the basis of the relevance in clinical practice.
Type of tasks
We defined the following five task types: i) To detect the medication dose 3 months after
onset of treatment ii) Decision about the adjustment of a treatment based on the impairment
or persistent state of data monitored iii) Question about the acquisition of underlying diseases.
iv), Question about the detection of implausible data, which could be important for the further
treatment, v) Question about the completeness of additional information. The Table 1
provides examples of tasks used for Neuro-Ophthalmology.
The patient model will come from original paper forms to assure that cases are comparable to
the real life situation. The cases, slightly modifiedThe full patient record, containing a slight
modification for the contrast examined in a specific task, will be transferred from paper into
the electronic patient record system. During the experiment, a research assistant will monitor
physicians and measure the time required to complete each task. Each physician from each
division has to solve all of the tasks. The evaluation will take place at two time points;
immediately after implementation of EDeR and one year after.
Design Matrix Summary
This is a two group (EDeR vs. paper based) cross-over (randomly assigning the sequence
electronic vs. paper-based per task) study. Time (primary endpoint) and accuracy (secondary
endpoint) are the two outcome variables. Intervention is the EDeR that is compared with the
Page 30 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
9
paper-based patient record. We assess tasks within 5 clinical subgroups (without examining
subgroup effects) and perform the experiment in two groups with different ophthalmologic
expertise (junior doctors vs. senior doctors). Each assessor will complete 50 tasks (i.e. 25
tasks based on the electronic patient record and 25 based on the paper patient record).
Electronic and paper-based vignettes are slightly different for each task. If an assessor checks
the same vignette a second time (e. g. in the electronic version), he or she must read the
content again, because the answer is different to the first time (e. g. in the paper-based
version). The material for the paper-based-group comes from real patients records and the
tasks were selected based on clinical relevance. For Summary see Figure 1
Detailed Research Plan
a) Creating the patient vignettes
Of each specialization we will collect five typical patient records on paper.
b) Checking for plausibility and technical feasibility
After creation of the vignettes, two to three individuals will be selected to evaluate the tasks.
They will solve the tasks and afterwards fill out a questionnaire, which will consist of
questions about the difficulty, comprehensibility and technical feasibility of the task.
c) Design matrix
We will generate random sequence coding whether the task is started with the paper or the
electronic version of the patient record. A second random sequence will be generated to code
for the sequence with which tasks will be presented to the participants. These two measures
will eliminate order effects.
Page 31 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
10
d) At the assessment
The participant will be sitting in a consultation room and the assessor will be seated next to
him or her to avoid distractions.
We will design a reply form into which each participant enters his or her response to the task.
This reply form will be handed over to the assessor after completion. Time will be taken by
the assessor after completing each entry into this form. In total, each participant will go
through 25 pairs of tasks. In case of a technical problem with the electronic system, the
assessment will be suspended and the remaining tasks will be completed after restoration of
the system.
There will be no communication between participant and assessor
e) The role of the assessor
1. The assessor will measure the time the participant needs to complete the reply form for
each task and each mode (paper based, electronic).
2. The assessor will take care of a protocol based procedure.
3. If any inconveniences of technical or administrative manner occur, the assessor will solve
them.
f) End of the assessment
All forms will be collected and the data will be entered in anonymised fashion into a
spreadsheet for further statistical analysis.
Sample size calculation
By implementing a novel electronic health record system it is possible to observe a
productivity gain of the physicians3.physician
3. We hypothesise that our productivity gain
should be one additional consultation per day assuming that one consultation lasts about 20
Page 32 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
11
minutes and a physician sees 25 patients daily. Comparing the two methods, paper and
electronic medical record, the overall gain in time should be 60 seconds on average when
using the electronic medical record EDeR.
Therefore, we are planning a study of the continuous response variable time from matched
pairs of study subjects. We presume that the difference in the response of matched pairs is
normally distributed with standard deviation 45. If the true difference in the mean time of
matched pairs is 60 seconds, we will need to study 7 pairs of subjects to be able to reject the
null hypothesis that this response difference is zero with probability (power) 0.8. The Type I
error probability associated with this test of this null hypothesis is 0.05.
At the time of writing this protocol 10 senior physicians and 9 junior doctors were willing to
participate in the study.
Data collection
Point of time of data collection
The first data acquisition will take place shortly after the implementation of EDeR. The
defined tasks will be available in paper form and the physicians will fill them out with the
help of a medical record, either the paper form or the electronic form.
Additionally, there will be someone from the study personalpersonnel, who sits with the
physicians in the same room and measure the time they need to complete each task.
Baseline data
Prior to the examination baseline data will be collected of each participant. Baseline data will
consist of information about the physicians’ age, extent of previous training in
ophthalmology, degree of specialization and self-reported level of general computing skills.
Follow-up data
Page 33 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
12
At the second time point, one year after the first assessment, the experiment will be repeated
using the same protocol. In the interval between the two measurements, all physicians will be
mainly working with the EDeR system.
Page 34 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
13
Discussion
This paper describes the rationale, methodology and design of a randomised, prospective
simulation study using patient vignettes based on real patients. This study will compare the
paper based medical records with electronic medical record software EDeR in terms of
efficiency, safety and cost reduction in the daily routine of an ophthalmological clinic. The
experiment will consist of 25 tasks, which will be solved by two groups of participants; senior
and junior doctors.
We presume that the implementation of the electronic medical record software EDeR will
have a positive impact on the efficiency of the doctors, which will result in an increase of
consultations per day. We believe that the results of our study will provide a valid basis to
quantify the added value of an electronic medical record system in an ophthalmologic
environment.
Page 35 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
14
References
1. Chiang MF, Boland MV, Margolis JW, Lum F, Abramoff MD, Hildebrand PL. Adoption
and perceptions of electronic health record systems by ophthalmologists: an American
Academy of Ophthalmology survey. Ophthalmology 2008;115(9):1591-7; quiz 97 e1-
5.
2. Bates DW, Gawande AA. Improving safety with information technology. N Engl J Med
2003;348(25):2526-34.
3. Cheriff AD, Kapur AG, Qiu M, Cole CL. Physician productivity and the ambulatory EHR
in a large academic multi-specialty physician group. Int J Med Inform;79(7):492-500.
4. Black AD, Car J, Pagliari C, Anandan C, Cresswell K, Bokun T, et al. The impact of
eHealth on the quality and safety of health care: a systematic overview. PLoS
Med;8(1):e1000387.
5. Vishwanath A, Singh SR, Winkelstein P. The impact of electronic medical record systems
on outpatient workflows: a longitudinal evaluation of its workflow effects. Int J Med
Inform;79(11):778-91.
6. Poissant L, Pereira J, Tamblyn R, Kawasumi Y. The impact of electronic health records on
time efficiency of physicians and nurses: a systematic review. J Am Med Inform Assoc
2005;12(5):505-16.
7. Murphy EC, Ferris FL, 3rd, O'Donnell WR. An electronic medical records system for
clinical research and the EMR EDC interface. Invest Ophthalmol Vis Sci
2007;48(10):4383-9.
8. Fontaine P, Ross SE, Zink T, Schilling LM. Systematic review of health information
exchange in primary care practices. J Am Board Fam Med;23(5):655-70.
9. Chiang MF, Boland MV, Brewer A, Epley KD, Horton MB, Lim MC, et al. Special
requirements for electronic health record systems in ophthalmology.
Ophthalmology;118(8):1681-7.
References
1. Chiang MF, Boland MV, Margolis JW, Lum F, Abramoff MD, Hildebrand PL. Adoption
and perceptions of electronic health record systems by ophthalmologists: an American
Academy of Ophthalmology survey. Ophthalmology 2008;115(9):1591-7; quiz 1597
e1-5.
2. Bates DW, Gawande AA. Improving safety with information technology. N Engl J Med
2003;348(25):2526-34.
3. Cheriff AD, Kapur AG, Qiu M, Cole CL. Physician productivity and the ambulatory EHR
in a large academic multi-specialty physician group. Int J Med Inform 2010;79(7):492-
500.
4. Black AD, Car J, Pagliari C, Anandan C, Cresswell K, Bokun T, et al. The impact of
eHealth on the quality and safety of health care: a systematic overview. PLoS Med
2011;8(1):e1000387.
Page 36 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
15
5. Vishwanath A, Singh SR, Winkelstein P. The impact of electronic medical record systems
on outpatient workflows: a longitudinal evaluation of its workflow effects. Int J Med
Inform 2010;79(11):778-91.
6. Vishwanath A, Winkelstein P, Singh SR. The impact of electronic health records on
outpatient flow: an evaluation of pre- and post-implementation expectations. AMIA
Annu Symp Proc 2006:1130.
7. Uslu AM, Stausberg J. Value of the electronic patient record: an analysis of the literature. J
Biomed Inform 2008;41(4):675-82.
8. Chiang MF, Boland MV, Brewer A, Epley KD, Horton MB, Lim MC, et al. Special
requirements for electronic health record systems in ophthalmology. Ophthalmology
2011;118(8):1681-7.
9. Poissant L, Pereira J, Tamblyn R, Kawasumi Y. The impact of electronic health records on
time efficiency of physicians and nurses: a systematic review. J Am Med Inform Assoc
2005;12(5):505-16.
10. Murphy EC, Ferris FL, 3rd, O'Donnell WR. An electronic medical records system for
clinical research and the EMR EDC interface. Invest Ophthalmol Vis Sci
2007;48(10):4383-9.
11. Fontaine P, Ross SE, Zink T, Schilling LM. Systematic review of health information
exchange in primary care practices. J Am Board Fam Med 2010;23(5):655-70.
Page 37 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
16
Competing interests
The authors declare that they have no competing interests.
Conflict of interest
None declared
Authors' contributions
Lucas M. Bachmann was involved in the conception and design of the study, drafted the
protocol, supervised the revisions and approved the final manuscript. Oliver Job was involved
in conception and design of this study, revised the draft critically for intellectual content and
approved the final revised manuscript. Martin K Schmid and Michael A. Thiel were involved
in the conception of the study, revised the draft critically for intellectual content and approved
the final manuscript. Sandra Ivic was involved in the conception and design of the study,
drafted the protocol and approved the final manuscript.
Acknowledgments
None
Main Sponsors:
Cantonal hospital of Lucerne, 6000 Lucerne 16, Switzerland
Page 38 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
17
Table 1 Description of context, task and contrast examined for the example of Neuro-
Ophthalmology
Task 1 Context Regular control visit Task To detect the medication dose after a certain of time after onset of treatment Contrast EdER: 6x daily 3 dragées à 60 mg Mestinon®
Paper record: 15x daily 3 dragées à 60 mg Mestinon®
Task 2 Context Regular control visit Task Decision about the adjustment of a treatment based on the impairment or persistent state of data
monitored - "In a case of Myasthenia gravis, the decision to adjust the steroid dose is made 4 weeks after starting therapy. If there has been a significant improvement in eye position, recorded with the coordimetry of Hess Weiss, the steroids can be tapered off. Is the eye position unchanged, or even worse, the steroid dose should be increased or an additional medication should be introduced. In the present case: Would you increase or decrease the dose of the steroids 4 weeks after onset of the therapy?"
Contrast EdER: There has been a significant improvement in the eye position; therefore the steroids can be tapered off. Paper record: The eye position is unchanged; the steroid dose should be increased.
Task 3 Context Post-op medication Task Question about the detection acquisition of underlying diseases.
Question: "In case of an allergic reaction to Bactrim® the administration of Diamox® often also produces allergic signs due to a cross-reaction. Diamox® should be avoided in such a situation. Is there a contraindication to administer Diamox® after a cataract surgery in the present case?"
Contrast EdER: The section with the patient history, containing information about allergies and underlying diseases, is blank. Although the patients has a sulfonamide allergy. Paper record: There is a short note about an allergic reaction to Bactrim®.
Task 4 Context Error in documentation Task "Do you find any implausible data regarding visual field recordings during the last 3 years of
documentation in the present case of a patient with a pituitary adenoma?"
Contrast EdER: The series of visual fields will show a constant bitemporal field Paper record: The series of visual fields will show one hemianopic visual field within the series of bitemporal fields. This change in just one visual field within an otherwise stable series is implausible and indicates an error of documentation)
Task 5 Context Completeness of clinical information Task In a patient with an idiopathic intracranial hypertension the diagnosis is made by the fact of
(bilateral) papilledema without visual afferent defects and raised intracranial opening pressure measured by lumbar puncture. Is it correct to make the diagnosis of an idiopathic intracranial hypertension in the present case?"
Contrast EdER: The necessary information to diagnose idiopathic intracranial hypertension is present. Paper record: No lumbar puncture, diagnosis not possible, because of missing information.
Page 39 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
18
Figure 1 Study Flow and Design Features
Formatted: Font: Times New Roman
Page 40 of 40
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
Assessing the efficacy of the electronic patient record system EDeR: Implementation study - Study protocol
Journal: BMJ Open
Manuscript ID: bmjopen-2012-002478.R2
Article Type: Protocol
Date Submitted by the Author: 11-Mar-2013
Complete List of Authors: Job, Oliver; Eye Clinic Cantonal hospital of Lucerne 6000 Lucerne 16, Switzerland, Bachmann, Lucas; medignition Inc, Schmid, Martin; Eye Clinic Cantonal hospital of Lucerne 6000 Lucerne 16, Switzerland, Thiel, Michael; Eye Clinic Cantonal hospital of Lucerne 6000 Lucerne 16, Switzerland, Ivic, Sandra; medignition Inc,
<b>Primary Subject Heading</b>:
Medical management
Secondary Subject Heading: Ophthalmology, Health informatics
Keywords:
Information management < BIOTECHNOLOGY & BIOINFORMATICS, Information technology < BIOTECHNOLOGY & BIOINFORMATICS, Cataract and refractive surgery < OPHTHALMOLOGY, Glaucoma < OPHTHALMOLOGY, Neuro-ophthalmology < OPHTHALMOLOGY, Medical retina < OPHTHALMOLOGY
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open on June 5, 2020 by guest. P
rotected by copyright.http://bm
jopen.bmj.com
/B
MJ O
pen: first published as 10.1136/bmjopen-2012-002478 on 10 A
pril 2013. Dow
nloaded from
For peer review only
1
Assessing the efficacy of the electronic patient record system EDeR:
Implementation study - Study protocol
Oliver Job 1, Lucas M. Bachmann
2 *, Martin K Schmid
1, Michael A. Thiel
1, Sandra Ivic
2
[1]
Eye Clinic
Cantonal hospital of Lucerne
6000 Lucerne 16, Switzerland
[2]
Medignition Inc. Research Consultants
Postrasse 15, 6300 Zug, Switzerland
* Corresponding Author
Prof . Lucas M. Bachmann, MD, Ph.D. Medignition Inc, Postrasse 15, 6300 Zug, Switzerland
e-mail Adresses:
Oliver Job: [email protected]
Lucas M. Bachmann: [email protected]
Martin K. Schmid: [email protected]
Michael A. Thiel: [email protected]
Sandra Ivic: [email protected]
Page 1 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
2
ABSTRACT
Introduction Despite many innovations in information technology, many clinics still rely on
paper based medical records. Critics, however, claim that they are hard to read, because of
illegible handwriting, and uncomfortable to use. Moreover, a chronological overview is not
always easily possible, content can be destroyed or get lost. There is an overall opinion that
electronic medical records should solve these problems and improve physicians’ efficiency,
patients’ safety and reduce overall costs in practice. However, to date, the evidence
supporting this view is sparse.
Methods and Analysis In this protocol we describe a study exploring differences in speed
and accuracy when searching clinical information using the paper based patient record or the
EDeR. Designed as a randomised vignette study, we hypothesise that the EDeR increases
efficiency, i.e. reduces time on reading the patient history and looking for relevant
examination results, helps finding mistakes and missing information quicker and more
reliably. In exploratory analyses we aim at exploring factors associated with a higher
performance.
Ethics and Dissemination The ethics committee of the Canton Lucerne, Switzerland,
approved this study. We presume that the implementation of the electronic medical record
software EDeR will have a positive impact on the efficiency of the doctors, which will result
in an increase of consultations per day. We believe that the results of our study will provide a
valid basis to quantify the added value of an electronic medical record system in an
ophthalmologic environment.
238 words
Page 2 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
3
Article focus
To compare the gain in
• Efficiency
• Accuracy
When introducing an electronic patient record system vs. paper based records
Key messages
The chosen design
• has the potential to validly address the added value of an electronic patient record
system
• allows extrapolating the economic gain of introducing the system
Strengths and limitations
Study design allows examining the effects
• in a realistic clinical setting
• using relevant clinical problems
• using data of real patients
Economic consequences can only be addressed indirectly.
Page 3 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
4
BACKGROUND
Despite substantial effort to promote electronic data management in patient care, a large
proportion of healthcare providers still rely on paper based patient records. The American
Academy of Ophthalmology performed a survey concerning the adoption rate of electronic
medical records within their members and found an adoption rate of only 12% 1.
The main problem with paper based patient records is the handwriting and the missing
overview of treatment and patient history. The handwriting is sometimes hard to read for the
physician in charge and the missing overview could potentially lead to needless examinations
and errors in diagnosis, prescription and treatment 2, because of the absence of consistent
information on changes in state of health, treatment and the parameters measured. As a
solution of these problems, there was the idea to support clinical daily routine with the help of
computers, as it is reality in many fields of service.
In the last decade, there was enormous development and investment in the field of
electronically supported medical records. It seems to be a key strategy to improve health care
in the different specialties3. Electronic medical records are supposed to improve physicians’
workflow and to offer a better overview on patient’s history. Today many authors believe that
this innovation should improve safety and reduce costs 4.”A better understanding of the
impact of the electronic medical records on workflows is vital to understand what the
technology really does offer that is new and unique”5 6
. However, a systematic review
published in 2008 and covering the evidence until 2005 concluded that none of 20 studies
provided a formal cost effectiveness analysis of electronic patient record systems in general. 7
The cantonal hospital of Lucerne has developed new software addressing the special needs of
the ophthalmologic clinic. The idea was to create a tool, which offers a complete overview on
the patient history once a patient has been referred from another ophthalmologist or general
practitioner. The software includes physician’s drawings, imaging, measurement data (e.g.
Page 4 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
5
visual field testing, electroretinography), numerical data (e.g. autorefraction, keratometry, and
biometry) and ophthalmic image data (e.g. fundus photography, optical coherence
tomography). These data are essential to support clinical diagnosis to track disease
progression, and to plan treatment 8.
In order to explore the impact on productivity of physicians using electronic health records, it
has been shown, that there is a statistically significant increase in the average monthly patient
visit volume of 9 visits per examined provider per month 3, or the reduction of time spent on
administration by nursing staff 9. The collected data also could be used for clinical research
and clinical studies 10. In general, there is an existing hypothesis, that EMR should improve
patient’s safety, but there are neither data nor suggestions available, how this should be
measured. Overall however, there are only few studies, which show evidence, that the
implementation of electronic medical records has a major impact on the efficiency4 11
,
especially in the field of ophthalmology. There seems to be also the demand for more studies
with prospective and randomized experimental designs, instead of surveys 11
.
Possible barriers for a successful use of EMR are costs and time loss for implementation,
issues concerning the security and privacy of the patient’s data, lack in training and support of
the staff using the EMR software 4 11
. The main objective should be the elimination of
obstacles for a successful implementation.
In this protocol we describe a study exploring differences in speed and accuracy when
searching clinical information using the paper based patient record or the EDeR. Designed as
a randomised vignette study, we hypothesise that the EDeR increases efficiency, i.e. reduces
time on reading the patient history and looking for relevant examination results, helps finding
mistakes and missing information quicker and more reliably. In exploratory analyses we aim
at exploring factors associated with a higher performance.
Page 5 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
6
METHODS / DESIGN
Study design
The study is planned as a prospective simulated study.
Study location
The study is planned to take place at the cantonal hospital of Lucerne in the clinic of
ophthalmology. The Eye clinic is run by six consultant ophthalmologists, 12 senior
physicians, and 14 junior doctors.
Study population
The study population will be divided into two groups. One group will consist of senior
physicians of the ophthalmologic clinic of the cantonal hospital in Lucerne, Switzerland. The
other group will be formed by junior doctors, who are usually younger and more computer
adepts, but have less clinical experience.
Inclusion criteria
Ophthalmologist of the cantonal hospital Lucerne eye clinic who voluntarily agreed to
participate in the study will be included. Each participant will receive a general introduction
on the usage of the EDeR of half an hour prior to the assessment. The training will cover
aspects of data collection and saving very briefly. The training will be such, that the basic
functions of EDeR are known.
Primary outcome
Gain in speed, when performing the task using the EDeR system
Secondary outcome
Accuracy of replies given using the EDeR system vs. the paper record
Page 6 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
7
Experimental design
The study is planned as a prospective, randomised study using patient vignettes. The aim is to
measure the time needed to solve a given task and the accuracy of the result. For each out of
five different ophthalmologic sub-specialities (Neuro-ophthalmology, Corneal, Glaucoma,
Uveitis and orthoptics) we will create five tasks based on a patient example. Thus, we will
create 25 pairs of cases either represented in paper or electronic form. The tasks are chosen on
the basis of the relevance in clinical practice.
Type of tasks
We defined the following five task types: i) To detect the medication dose 3 months after
onset of treatment ii) Decision about the adjustment of a treatment based on the impairment
or persistent state of data monitored iii) Question about the acquisition of underlying diseases.
iv), Question about the detection of implausible data, which could be important for the further
treatment, v) Question about the completeness of additional information. Table 1 provides
examples of tasks used for Neuro-Ophthalmology.
The patient model will come from original paper forms to assure that cases are comparable to
the real life situation. The full patient record, containing a slight modification for the contrast
examined in a specific task, will be transferred from paper into the electronic patient record
system. During the experiment, a research assistant will monitor physicians and measure the
time required to complete each task. Each physician from each division has to solve all of the
tasks. The evaluation will take place at two time points; immediately after implementation of
EDeR and one year after.
Design Matrix Summary
This is a two group (EDeR vs. paper based) cross-over (randomly assigning the sequence
electronic vs. paper-based per task) study. Time (primary endpoint) and accuracy (secondary
endpoint) are the two outcome variables. Intervention is the EDeR that is compared with the
Page 7 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
8
paper-based patient record. We assess tasks within 5 clinical subgroups (without examining
subgroup effects) and perform the experiment in two groups with different ophthalmologic
expertise (junior doctors vs. senior doctors). Each assessor will complete 50 tasks (i.e. 25
tasks based on the electronic patient record and 25 based on the paper patient record). The
electronic records are exact representations of an existing paper-based record. Patient records
differ only in respect to one contrast that is changed for a specific task. Thus, besides that
task-specific detail, the remaining content of each pair of patient records is identical. But, if
an assessor checks the same vignette a second time (e. g. in the electronic version), he or she
must read the content of the paper-based record again, because the correct answer, i.e. a
specific lab value, is different. The tasks were selected based on clinical relevance. For
Summary see Figure 1
Detailed Research Plan
a) Creating the patient vignettes
Of each specialization we will collect five typical patient records on paper.
b) Checking for plausibility and technical feasibility
After creation of the vignettes, two to three individuals will be selected to evaluate the tasks.
They will solve the tasks and afterwards fill out a questionnaire, which will consist of
questions about the difficulty, comprehensibility and technical feasibility of the task.
c) Design matrix
We will generate random sequence coding whether the task is started with the paper or the
electronic version of the patient record. A second random sequence will be generated to code
for the sequence with which tasks will be presented to the participants. These two measures
will eliminate order effects.
Page 8 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
9
d) At the assessment
The participant will be sitting in a consultation room and the assessor will be seated next to
him or her to avoid distractions.
We will design a reply form into which each participant enters his or her response to the task.
This reply form will be handed over to the assessor after completion. Time will be taken by
the assessor after completing each entry into this form. In total, each participant will go
through 25 pairs of tasks. In case of a technical problem with the electronic system, the
assessment will be suspended and the remaining tasks will be completed after restoration of
the system.
There will be no communication between participant and assessor
e) The role of the assessor
1. The assessor will measure the time the participant needs to complete the reply form for
each task and each mode (paper based, electronic).
2. The assessor will take care of a protocol based procedure.
3. If any inconveniences of technical or administrative manner occur, the assessor will solve
them.
f) End of the assessment
All forms will be collected and the data will be entered in anonymised fashion into a
spreadsheet for further statistical analysis.
Sample size calculation
By implementing a novel electronic health record system it is possible to observe a
productivity gain of the physician 3. We hypothesise that our productivity gain should be one
additional consultation per day assuming that one consultation lasts about 20 minutes and a
Page 9 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
10
physician sees 25 patients daily. Comparing the two methods, paper and electronic medical
record, the overall gain in time should be 60 seconds on average when using the electronic
medical record EDeR.
Therefore, we are planning a study of the continuous response variable time from matched
pairs of study subjects. We presume that the difference in the response of matched pairs is
normally distributed with standard deviation 45. If the true difference in the mean time of
matched pairs is 60 seconds, we will need to study 7 pairs of subjects to be able to reject the
null hypothesis that this response difference is zero with probability (power) 0.8. The Type I
error probability associated with this test of this null hypothesis is 0.05.
At the time of writing this protocol 10 senior physicians and 9 junior doctors were willing to
participate in the study.
Data collection
Point of time of data collection
The first data acquisition will take place shortly after the implementation of EDeR. The
defined tasks will be available in paper form and the physicians will fill them out with the
help of a medical record, either the paper form or the electronic form.
Additionally, there will be someone from the study personnel, who sits with the physicians in
the same room and measure the time they need to complete each task.
Baseline data
Prior to the examination baseline data will be collected of each participant. Baseline data will
consist of information about the physicians’ age, extent of previous training in
ophthalmology, degree of specialization and self-reported level of general computing skills.
Follow-up data
Page 10 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
11
At the second time point, one year after the first assessment, the experiment will be repeated
using the same protocol. In the interval between the two measurements, all physicians will be
mainly working with the EDeR system.
Page 11 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
12
Discussion
This paper describes the rationale, methodology and design of a randomised, prospective
simulation study using patient vignettes based on real patients. This study will compare the
paper based medical records with electronic medical record software EDeR in terms of
efficiency, safety and cost reduction in the daily routine of an ophthalmological clinic. The
experiment will consist of 25 tasks, which will be solved by two groups of participants; senior
and junior doctors.
We presume that the implementation of the electronic medical record software EDeR will
have a positive impact on the efficiency of the doctors, which will result in an increase of
consultations per day. It can be argued that choosing volunteers for this study could lead to
exaggerated results. Nevertheless we believe that the results of our study will provide a
rational basis to quantify the added value of an electronic medical record system in an
ophthalmologic environment.
Page 12 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
13
References
1. Chiang MF, Boland MV, Margolis JW, et al. Adoption and perceptions of electronic health
record systems by ophthalmologists: an American Academy of Ophthalmology
survey. Ophthalmology 2008;115(9):1591-7; quiz 1597 e1-5.
2. Bates DW, Gawande AA. Improving safety with information technology. N Engl J Med
2003;348(25):2526-34.
3. Cheriff AD, Kapur AG, Qiu M, et al. Physician productivity and the ambulatory EHR in a
large academic multi-specialty physician group. Int J Med Inform 2010;79(7):492-500.
4. Black AD, Car J, Pagliari C, et al. The impact of eHealth on the quality and safety of health
care: a systematic overview. PLoS Med 2011;8(1):e1000387.
5. Vishwanath A, Singh SR, Winkelstein P. The impact of electronic medical record systems
on outpatient workflows: a longitudinal evaluation of its workflow effects. Int J Med
Inform 2010;79(11):778-91.
6. Vishwanath A, Winkelstein P, Singh SR. The impact of electronic health records on
outpatient flow: an evaluation of pre- and post-implementation expectations. AMIA
Annu Symp Proc 2006:1130.
7. Uslu AM, Stausberg J. Value of the electronic patient record: an analysis of the literature. J
Biomed Inform 2008;41(4):675-82.
8. Chiang MF, Boland MV, Brewer A, et al. Special requirements for electronic health record
systems in ophthalmology. Ophthalmology 2011;118(8):1681-7.
9. Poissant L, Pereira J, Tamblyn R, et al. The impact of electronic health records on time
efficiency of physicians and nurses: a systematic review. J Am Med Inform Assoc
2005;12(5):505-16.
10. Murphy EC, Ferris FL, 3rd, O'Donnell WR. An electronic medical records system for
clinical research and the EMR EDC interface. Invest Ophthalmol Vis Sci
2007;48(10):4383-9.
11. Fontaine P, Ross SE, Zink T, et al. Systematic review of health information exchange in
primary care practices. J Am Board Fam Med 2010;23(5):655-70.
Page 13 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
14
Competing interests
The authors declare that they have no competing interests.
Conflict of interest
None declared
Authors' contributions
Lucas M. Bachmann was involved in the conception and design of the study, drafted the
protocol, supervised the revisions and approved the final manuscript. Oliver Job was involved
in conception and design of this study, revised the draft critically for intellectual content and
approved the final revised manuscript. Martin K Schmid and Michael A. Thiel were involved
in the conception of the study, revised the draft critically for intellectual content and approved
the final manuscript. Sandra Ivic was involved in the conception and design of the study,
drafted the protocol and approved the final manuscript.
Acknowledgments
None
Main Sponsors:
Cantonal hospital of Lucerne, 6000 Lucerne 16, Switzerland
Page 14 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
15
Table 1 Description of context, task and contrast examined for the example of Neuro-
Ophthalmology
Task 1 Context Regular control visit Task To detect the medication dose after a certain of time after onset of treatment Contrast EdER: 6x daily 3 dragées à 60 mg Mestinon®
Paper record: 15x daily 3 dragées à 60 mg Mestinon®
Task 2 Context Regular control visit Task Decision about the adjustment of a treatment based on the impairment or persistent state of data
monitored - "In a case of Myasthenia gravis, the decision to adjust the steroid dose is made 4 weeks after starting therapy. If there has been a significant improvement in eye position, recorded with the coordimetry of Hess Weiss, the steroids can be tapered off. Is the eye position unchanged, or even worse, the steroid dose should be increased or an additional medication should be introduced. In the present case: Would you increase or decrease the dose of the steroids 4 weeks after onset of the therapy?"
Contrast EdER: There has been a significant improvement in the eye position; therefore the steroids can be tapered off. Paper record: The eye position is unchanged; the steroid dose should be increased.
Task 3 Context Post-op medication Task Question about the detection acquisition of underlying diseases.
Question: "In case of an allergic reaction to Bactrim® the administration of Diamox® often also produces allergic signs due to a cross-reaction. Diamox® should be avoided in such a situation. Is there a contraindication to administer Diamox® after a cataract surgery in the present case?"
Contrast EdER: The section with the patient history, containing information about allergies and underlying diseases, is blank. Although the patients has a sulfonamide allergy. Paper record: There is a short note about an allergic reaction to Bactrim®.
Task 4 Context Error in documentation Task "Do you find any implausible data regarding visual field recordings during the last 3 years of
documentation in the present case of a patient with a pituitary adenoma?"
Contrast EdER: The series of visual fields will show a constant bitemporal field Paper record: The series of visual fields will show one hemianopic visual field within the series of bitemporal fields. This change in just one visual field within an otherwise stable series is implausible and indicates an error of documentation)
Task 5 Context Completeness of clinical information Task In a patient with an idiopathic intracranial hypertension the diagnosis is made by the fact of
(bilateral) papilledema without visual afferent defects and raised intracranial opening pressure measured by lumbar puncture. Is it correct to make the diagnosis of an idiopathic intracranial hypertension in the present case?"
Contrast EdER: The necessary information to diagnose idiopathic intracranial hypertension is present. Paper record: No lumbar puncture, diagnosis not possible, because of missing information.
Page 15 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
16
Figure legend
Figure 1 Study Flow and Design Features
Page 16 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
1
Assessing the efficacy of the electronic patient record system EDeR:
Implementation study - Study protocol
Oliver Job 1, Lucas M. Bachmann
2 *, Martin K Schmid
1, Michael A. Thiel
1, Sandra Ivic
2
[1]
Eye Clinic
Cantonal hospital of Lucerne
6000 Lucerne 16, Switzerland
[2]
Medignition Inc. Research Consultants
Postrasse 15, 6300 Zug, Switzerland
* Corresponding Author
Prof . Lucas M. Bachmann, MD, Ph.D. Medignition Inc, Postrasse 15, 6300 Zug, Switzerland
e-mail Adresses:
Oliver Job: [email protected]
Lucas M. Bachmann: [email protected]
Martin K. Schmid: [email protected]
Michael A. Thiel: [email protected]
Sandra Ivic: [email protected]
Page 17 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
2
ABSTRACT
Introduction Despite many innovations in information technology, many clinics still rely on
paper based medical records. Critics, however, claim that they are hard to read, because of
illegible handwriting, and uncomfortable to use. Moreover, a chronological overview is not
always easily possible, content can be destroyed or get lost. There is an overall opinion that
electronic medical records should solve these problems and improve physicians’ efficiency,
patients’ safety and reduce overall costs in practice. However, to date, the evidence
supporting this view is sparse.
Methods and Analysis In this protocol we describe a study exploring differences in speed
and accuracy when searching clinical information using the paper based patient record or the
EDeR. Designed as a randomised vignette study, we hypothesise that the EDeR increases
efficiency, i.e. reduces time on reading the patient history and looking for relevant
examination results, helps finding mistakes and missing information quicker and more
reliably. In exploratory analyses we aim at exploring factors associated with a higher
performance.
Ethics and Dissemination The ethics committee of the Canton Lucerne, Switzerland,
approved this study. We presume that the implementation of the electronic medical record
software EDeR will have a positive impact on the efficiency of the doctors, which will result
in an increase of consultations per day. We believe that the results of our study will provide a
valid basis to quantify the added value of an electronic medical record system in an
ophthalmologic environment.
238 words
Page 18 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
3
Article focus
To compare the gain in
• Efficiency
• Accuracy
When introducing an electronic patient record system vs. paper based records
Key messages
The chosen design
• has the potential to validly address the added value of an electronic patient record
system
• allows extrapolating the economic gain of introducing the system
Strengths and limitations
Study design allows examining the effects
• in a realistic clinical setting
• using relevant clinical problems
• using data of real patients
Economic consequences can only be addressed indirectly.
Page 19 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
4
BACKGROUND
Despite substantial effort to promote electronic data management in patient care, a large
proportion of healthcare providers still rely on paper based patient records. The American
Academy of Ophthalmology performed a survey concerning the adoption rate of electronic
medical records within their members and found an adoption rate of only 12% 1.
The main problem with paper based patient records is the handwriting and the missing
overview of treatment and patient history. The handwriting is sometimes hard to read for the
physician in charge and the missing overview could potentially lead to needless examinations
and errors in diagnosis, prescription and treatment 2, because of the absence of consistent
information on changes in state of health, treatment and the parameters measured. As a
solution of these problems, there was the idea to support clinical daily routine with the help of
computers, as it is reality in many fields of service.
In the last decade, there was enormous development and investment in the field of
electronically supported medical records. It seems to be a key strategy to improve health care
in the different specialties3. Electronic medical records are supposed to improve physicians’
workflow and to offer a better overview on patient’s history. Today many authors believe that
this innovation should improve safety and reduce costs 4.”A better understanding of the
impact of the electronic medical records on workflows is vital to understand what the
technology really does offer that is new and unique”5 6
. However, a systematic review
published in 2008 and covering the evidence until 2005 concluded that none of 20 studies
provided a formal cost effectiveness analysis of electronic patient record systems in general. 7
The cantonal hospital of Lucerne has developed new software addressing the special needs of
the ophthalmologic clinic. The idea was to create a tool, which offers a complete overview on
the patient history once a patient has been referred from another ophthalmologist or general
practitioner. The software includes physician’s drawings, imaging, measurement data (e.g.
Page 20 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
5
visual field testing, electroretinography), numerical data (e.g. autorefraction, keratometry, and
biometry) and ophthalmic image data (e.g. fundus photography, optical coherence
tomography). These data are essential to support clinical diagnosis to track disease
progression, and to plan treatment 8.
In order to explore the impact on productivity of physicians using electronic health records, it
has been shown, that there is a statistically significant increase in the average monthly patient
visit volume of 9 visits per examined provider per month 3, or the reduction of time spent on
administration by nursing staff 9. The collected data also could be used for clinical research
and clinical studies 10. In general, there is an existing hypothesis, that EMR should improve
patient’s safety, but there are neither data nor suggestions available, how this should be
measured. Overall however, there are only few studies, which show evidence, that the
implementation of electronic medical records has a major impact on the efficiency4 11
,
especially in the field of ophthalmology. There seems to be also the demand for more studies
with prospective and randomized experimental designs, instead of surveys 11
.
Possible barriers for a successful use of EMR are costs and time loss for implementation,
issues concerning the security and privacy of the patient’s data, lack in training and support of
the staff using the EMR software 4 11
. The main objective should be the elimination of
obstacles for a successful implementation.
In this protocol we describe a study exploring differences in speed and accuracy when
searching clinical information using the paper based patient record or the EDeR. Designed as
a randomised vignette study, we hypothesise that the EDeR increases efficiency, i.e. reduces
time on reading the patient history and looking for relevant examination results, helps finding
mistakes and missing information quicker and more reliably. In exploratory analyses we aim
at exploring factors associated with a higher performance.
Page 21 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
6
METHODS / DESIGN
Study design
The study is planned as a prospective simulated study.
Study location
The study is planned to take place at the cantonal hospital of Lucerne in the clinic of
ophthalmology. The Eye clinic is run by six consultant ophthalmologists, 12 senior
physicians, and 14 junior doctors.
Study population
The study population will be divided into two groups. One group will consist of senior
physicians of the ophthalmologic clinic of the cantonal hospital in Lucerne, Switzerland. The
other group will be formed by junior doctors, who are usually younger and more computer
adepts, but have less clinical experience.
Inclusion criteria
Ophthalmologist of the cantonal hospital Lucerne eye clinic who voluntarily agreed to
participate in the study will be included. Each participant will receive a general introduction
on the usage of the EDeR of half an hour prior to the assessment. The training will cover
aspects of data collection and saving very briefly. The training will be such, that the basic
functions of EDeR are known.
Primary outcome
Gain in speed, when performing the task using the EDeR system
Secondary outcome
Accuracy of replies given using the EDeR system vs. the paper record
Page 22 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
7
Experimental design
The study is planned as a prospective, randomised study using patient vignettes. The aim is to
measure the time needed to solve a given task and the accuracy of the result. For each out of
five different ophthalmologic sub-specialities (Neuro-ophthalmology, Corneal, Glaucoma,
Uveitis and orthoptics) we will create five tasks based on a patient example. Thus, we will
create 25 pairs of cases either represented in paper or electronic form. The tasks are chosen on
the basis of the relevance in clinical practice.
Type of tasks
We defined the following five task types: i) To detect the medication dose 3 months after
onset of treatment ii) Decision about the adjustment of a treatment based on the impairment
or persistent state of data monitored iii) Question about the acquisition of underlying diseases.
iv), Question about the detection of implausible data, which could be important for the further
treatment, v) Question about the completeness of additional information. Table 1 provides
examples of tasks used for Neuro-Ophthalmology.
The patient model will come from original paper forms to assure that cases are comparable to
the real life situation. The full patient record, containing a slight modification for the contrast
examined in a specific task, will be transferred from paper into the electronic patient record
system. During the experiment, a research assistant will monitor physicians and measure the
time required to complete each task. Each physician from each division has to solve all of the
tasks. The evaluation will take place at two time points; immediately after implementation of
EDeR and one year after.
Design Matrix Summary
This is a two group (EDeR vs. paper based) cross-over (randomly assigning the sequence
electronic vs. paper-based per task) study. Time (primary endpoint) and accuracy (secondary
endpoint) are the two outcome variables. Intervention is the EDeR that is compared with the
Page 23 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
8
paper-based patient record. We assess tasks within 5 clinical subgroups (without examining
subgroup effects) and perform the experiment in two groups with different ophthalmologic
expertise (junior doctors vs. senior doctors). Each assessor will complete 50 tasks (i.e. 25
tasks based on the electronic patient record and 25 based on the paper patient record).
Electronic and paper-based vignettes are slightly different for each task. IfThe electronic
records are exact representations of an existing paper-based record. Patient records differ only
in respect to one contrast that is changed for a specific task. Thus, besides that task-specific
detail, the remaining content of each pair of patient records is identical. But, if an assessor
checks the same vignette a second time (e. g. in the electronic version), he or she must read
the content of the paper-based record again, because the correct answer, i.e. a specific lab
value, is different to the first time (e. g. in the paper-based version).. The material for the
paper-based-group comes from real patients records and the tasks were selected based on
clinical relevance. For Summary see Figure 1
Detailed Research Plan
a) Creating the patient vignettes
Of each specialization we will collect five typical patient records on paper.
b) Checking for plausibility and technical feasibility
After creation of the vignettes, two to three individuals will be selected to evaluate the tasks.
They will solve the tasks and afterwards fill out a questionnaire, which will consist of
questions about the difficulty, comprehensibility and technical feasibility of the task.
c) Design matrix
We will generate random sequence coding whether the task is started with the paper or the
electronic version of the patient record. A second random sequence will be generated to code
Page 24 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
9
for the sequence with which tasks will be presented to the participants. These two measures
will eliminate order effects.
d) At the assessment
The participant will be sitting in a consultation room and the assessor will be seated next to
him or her to avoid distractions.
We will design a reply form into which each participant enters his or her response to the task.
This reply form will be handed over to the assessor after completion. Time will be taken by
the assessor after completing each entry into this form. In total, each participant will go
through 25 pairs of tasks. In case of a technical problem with the electronic system, the
assessment will be suspended and the remaining tasks will be completed after restoration of
the system.
There will be no communication between participant and assessor
e) The role of the assessor
1. The assessor will measure the time the participant needs to complete the reply form for
each task and each mode (paper based, electronic).
2. The assessor will take care of a protocol based procedure.
3. If any inconveniences of technical or administrative manner occur, the assessor will solve
them.
f) End of the assessment
All forms will be collected and the data will be entered in anonymised fashion into a
spreadsheet for further statistical analysis.
Sample size calculation
Page 25 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
10
By implementing a novel electronic health record system it is possible to observe a
productivity gain of the physician 3. We hypothesise that our productivity gain should be one
additional consultation per day assuming that one consultation lasts about 20 minutes and a
physician sees 25 patients daily. Comparing the two methods, paper and electronic medical
record, the overall gain in time should be 60 seconds on average when using the electronic
medical record EDeR.
Therefore, we are planning a study of the continuous response variable time from matched
pairs of study subjects. We presume that the difference in the response of matched pairs is
normally distributed with standard deviation 45. If the true difference in the mean time of
matched pairs is 60 seconds, we will need to study 7 pairs of subjects to be able to reject the
null hypothesis that this response difference is zero with probability (power) 0.8. The Type I
error probability associated with this test of this null hypothesis is 0.05.
At the time of writing this protocol 10 senior physicians and 9 junior doctors were willing to
participate in the study.
Data collection
Point of time of data collection
The first data acquisition will take place shortly after the implementation of EDeR. The
defined tasks will be available in paper form and the physicians will fill them out with the
help of a medical record, either the paper form or the electronic form.
Additionally, there will be someone from the study personnel, who sits with the physicians in
the same room and measure the time they need to complete each task.
Baseline data
Page 26 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
11
Prior to the examination baseline data will be collected of each participant. Baseline data will
consist of information about the physicians’ age, extent of previous training in
ophthalmology, degree of specialization and self-reported level of general computing skills.
Follow-up data
At the second time point, one year after the first assessment, the experiment will be repeated
using the same protocol. In the interval between the two measurements, all physicians will be
mainly working with the EDeR system.
Page 27 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
12
Discussion
This paper describes the rationale, methodology and design of a randomised, prospective
simulation study using patient vignettes based on real patients. This study will compare the
paper based medical records with electronic medical record software EDeR in terms of
efficiency, safety and cost reduction in the daily routine of an ophthalmological clinic. The
experiment will consist of 25 tasks, which will be solved by two groups of participants; senior
and junior doctors.
We presume that the implementation of the electronic medical record software EDeR will
have a positive impact on the efficiency of the doctors, which will result in an increase of
consultations per day. WeIt can be argued that choosing volunteers for this study could lead to
exaggerated results. Nevertheless we believe that the results of our study will provide a
validrational basis to quantify the added value of an electronic medical record system in an
ophthalmologic environment.
Page 28 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
13
References
1. Chiang MF, Boland MV, Margolis JW, Lum F, Abramoff MD, Hildebrand PL. Adoption
and perceptions of electronic health record systems by ophthalmologists: an American
Academy of Ophthalmology survey. Ophthalmology 2008;115(9):1591-7; quiz 1597
e1-5.
2. Bates DW, Gawande AA. Improving safety with information technology. N Engl J Med
2003;348(25):2526-34.
3. Cheriff AD, Kapur AG, Qiu M, Cole CL. Physician productivity and the ambulatory EHR
in a large academic multi-specialty physician group. Int J Med Inform 2010;79(7):492-
500.
4. Black AD, Car J, Pagliari C, Anandan C, Cresswell K, Bokun T, et al. The impact of
eHealth on the quality and safety of health care: a systematic overview. PLoS Med
2011;8(1):e1000387.
5. Vishwanath A, Singh SR, Winkelstein P. The impact of electronic medical record systems
on outpatient workflows: a longitudinal evaluation of its workflow effects. Int J Med
Inform 2010;79(11):778-91.
6. Vishwanath A, Winkelstein P, Singh SR. The impact of electronic health records on
outpatient flow: an evaluation of pre- and post-implementation expectations. AMIA
Annu Symp Proc 2006:1130.
7. Uslu AM, Stausberg J. Value of the electronic patient record: an analysis of the literature. J
Biomed Inform 2008;41(4):675-82.
8. Chiang MF, Boland MV, Brewer A, Epley KD, Horton MB, Lim MC, et al. Special
requirements for electronic health record systems in ophthalmology. Ophthalmology
2011;118(8):1681-7.
9. Poissant L, Pereira J, Tamblyn R, Kawasumi Y. The impact of electronic health records on
time efficiency of physicians and nurses: a systematic review. J Am Med Inform Assoc
2005;12(5):505-16.
10. Murphy EC, Ferris FL, 3rd, O'Donnell WR. An electronic medical records system for
clinical research and the EMR EDC interface. Invest Ophthalmol Vis Sci
2007;48(10):4383-9.
11. Fontaine P, Ross SE, Zink T, Schilling LM. Systematic review of health information
exchange in primary care practices. J Am Board Fam Med 2010;23(5):655-70.
Page 29 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
14
Competing interests
The authors declare that they have no competing interests.
Conflict of interest
None declared
Authors' contributions
Lucas M. Bachmann was involved in the conception and design of the study, drafted the
protocol, supervised the revisions and approved the final manuscript. Oliver Job was involved
in conception and design of this study, revised the draft critically for intellectual content and
approved the final revised manuscript. Martin K Schmid and Michael A. Thiel were involved
in the conception of the study, revised the draft critically for intellectual content and approved
the final manuscript. Sandra Ivic was involved in the conception and design of the study,
drafted the protocol and approved the final manuscript.
Acknowledgments
None
Main Sponsors:
Cantonal hospital of Lucerne, 6000 Lucerne 16, Switzerland
Page 30 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
15
Table 1 Description of context, task and contrast examined for the example of Neuro-
Ophthalmology
Task 1 Context Regular control visit Task To detect the medication dose after a certain of time after onset of treatment Contrast EdER: 6x daily 3 dragées à 60 mg Mestinon®
Paper record: 15x daily 3 dragées à 60 mg Mestinon®
Task 2 Context Regular control visit Task Decision about the adjustment of a treatment based on the impairment or persistent state of data
monitored - "In a case of Myasthenia gravis, the decision to adjust the steroid dose is made 4 weeks after starting therapy. If there has been a significant improvement in eye position, recorded with the coordimetry of Hess Weiss, the steroids can be tapered off. Is the eye position unchanged, or even worse, the steroid dose should be increased or an additional medication should be introduced. In the present case: Would you increase or decrease the dose of the steroids 4 weeks after onset of the therapy?"
Contrast EdER: There has been a significant improvement in the eye position; therefore the steroids can be tapered off. Paper record: The eye position is unchanged; the steroid dose should be increased.
Task 3 Context Post-op medication Task Question about the detection acquisition of underlying diseases.
Question: "In case of an allergic reaction to Bactrim® the administration of Diamox® often also produces allergic signs due to a cross-reaction. Diamox® should be avoided in such a situation. Is there a contraindication to administer Diamox® after a cataract surgery in the present case?"
Contrast EdER: The section with the patient history, containing information about allergies and underlying diseases, is blank. Although the patients has a sulfonamide allergy. Paper record: There is a short note about an allergic reaction to Bactrim®.
Task 4 Context Error in documentation Task "Do you find any implausible data regarding visual field recordings during the last 3 years of
documentation in the present case of a patient with a pituitary adenoma?"
Contrast EdER: The series of visual fields will show a constant bitemporal field Paper record: The series of visual fields will show one hemianopic visual field within the series of bitemporal fields. This change in just one visual field within an otherwise stable series is implausible and indicates an error of documentation)
Task 5 Context Completeness of clinical information Task In a patient with an idiopathic intracranial hypertension the diagnosis is made by the fact of
(bilateral) papilledema without visual afferent defects and raised intracranial opening pressure measured by lumbar puncture. Is it correct to make the diagnosis of an idiopathic intracranial hypertension in the present case?"
Contrast EdER: The necessary information to diagnose idiopathic intracranial hypertension is present. Paper record: No lumbar puncture, diagnosis not possible, because of missing information.
Page 31 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
16
Figure 1 Study Flow and Design Features
Page 32 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from
For peer review only
125x90mm (300 x 300 DPI)
Page 33 of 33
For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml
BMJ Open
123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960
on June 5, 2020 by guest. Protected by copyright.
http://bmjopen.bm
j.com/
BM
J Open: first published as 10.1136/bm
jopen-2012-002478 on 10 April 2013. D
ownloaded from