Full Report - Fondation canadienne pour l'amélioration des services ...

The Impact of a Standardized Information System
Between the Emergency Department and the
Primary Care Network: Effects on Continuity and
Quality of Care
February 2003
Dr. Marc Afilalo
Dr. Eddy Lang
Dr. Jean François Boivin
Decision Maker Partners:
Collège des Médecins du Québec
Collège de Médecine Familiale du Québec
Réseau de Recherche en Urgence de Québec
Funding Provided by:
Canadian Health Service Research Foundation
Fonds de la Recherche en Santé du Québec
Ministère de la Santé et des Services Sociaux
Ministère de la Recherche de la Science et de la Technologie
Régie régionale de la Santé et des Services Sociaux
Association des Médecins d’Urgence du Québec
Emergency Department, Sir Mortimer B. Davis, Jewish General Hospital
Information Services Department, Jewish General Hospital

Principal Investigators:
Dr. Marc Afilalo
Emergency department, D-012
SMBD – Jewish General Hospital
3755, Côte St-Catherine road
Montreal, Quebec H3T 1E2
Telephone: (514) 340-8222 ext. 5568
Fax: (514) 340-7519
E-mail: marc.afilalo@mcgill.ca
Dr. Jean François Boivin
Epidemiology department
SMBD – Jewish General Hospital
3755, Côte St-Catherine road
Montreal, Quebec H3T 1E2
Telephone: (514) 340-8222 ext. 3647
E-mail: jean-f.boivin@mcgill.ca
Dr. Eddy Lang
Emergency department, D-027
SMBD – Jewish General Hospital
3755, Côte St-Catherine road
Montreal, Quebec H3T 1E2
Telephone: (514) 340-8222 ext. 5568
Fax: (514) 340-7519
E-mail: eddy.lang@mcgill.ca
This document is available on the Canadian Health Services Research Foundation web site
(www.chrsf.ca).
For more information on the Canadian Health Services Research Foundation, contact the
foundation at:
1565 Carling Avenue, Suite 700
Ottawa, Ontario
K1Z 8R1
E-mail: communications@chsrf.ca
Telephone: (613) 728-2238
Fax: (613) 728-3527
Ce document est disponible sur le site web de la Fondation canadienne de la recherche sur les
services de santé (www.fcrss.ca).
Pour obtenir de plus amples renseignements sur la Fondation canadienne de la recherche sur les
services de santé, communiquez avec la Fondation :
1565, avenue Carling, bureau 700
Ottawa (Ontario)
K1Z 8R1
Courriel : communications@fcrss.ca
Téléphone : (613) 728-2238
Télécopieur : (613) 728-3527

The Impact of a Standardized Information System
Between the Emergency Department and the Primary Care
Network: Effects on Continuity and Quality of Care
Dr. Marc Afilalo 1
Dr. Eddy Lang 1
Dr. Jean François Boivin 2
Antoinette Colacone 1
Dr. Howard Goldstein 3
Dr. Alex Guttman 1
Sophie Lapierre, PhD 4
Alain Vandal, PhD 5
Dr. Steve Rosenthal 1
Dr. Bernard Unger 1
Claudine Giguère 1
1
Emergency Department, Jewish General Hospital, McGill University
2
Epidemiology department, Jewish General Hospital, McGill University
3
Department of Family Medicine, Jewish General Hospital, McGill University
4
École Polytechnique, Université de Montréal
5
Department of Mathematic and Statistic, McGill University
Acknowledgements:
The members of the research team would like to thank the following persons for their
invaluable collaboration during the study:
The 23 family physicians participating in the study: Dr. D Alper, Dr. D Amdursky, Dr. J
Backler, Dr. I Benjamin, Dr. A Bourkas, Dr. A Elbaz, Dr. N Fox, Dr. S Heisler, Dr. P Lipes,
Dr. R Ludman, Dr. M Malus, Dr. L Newman, Dr. J Rivilis, Dr. G Rosenthal, Dr. G Rubin,
Dr. D Steg, Dr. M Zigman, Dr. A Zylbergold, Dr. B ElFassy, Dr. ZM Jast, Dr. D Shahin, Dr.
N Ubani, Dr. M Rajakesari.
Collaborators from other centres: Dr. Bruno Baril, Dr. Yves Rouleau, Marie-Josée April.
The research assistants: Maria Gordon, Elana Ptack, Mady Virgona, Aicha Lewis, Krystyna
Zaluski, Marie Gdalewitch, Perry Gdalevitch, Manon Goulet.
The Information Services Department of the SMBD-Jewish general Hospital: Frédéric
Abergel, Ramin Abedan, Josée Brassard, Helene Jean-Baptiste.
SMBD-Jewish General Hospital, Emergency Department: All emergency physicians and
support staff.

Key Implications for Decision Makers
• There is a great deal of inefficiency and duplication in patient care due to a large gap
in information-sharing between emergency departments and the primary care
network.
• A standardized communication system can provide automatic updates on patient
emergency department care to family physicians using a secure web site.
• While a standardized communication system can bridge gaps in patient information, it
requires the support of all stakeholders: hospital managers, information technology
departments, clerical staff, and emergency and primary healthcare physicians.
• The standardized communication system does not appear to reduce emergency
department overcrowding.
• Patient confidence in the healthcare system is increased by knowing their family
physician has information about their emergency department use.
• Primary care providers also need detailed information about their patients’
hospitalizations as well as their emergency department use.
• Communication tools need to be tailored to each type of healthcare professional, such
as CLSCs, specialists, emergency physicians, etc.
i

Executive Summary
Context
Inadequate communication between the various components of the healthcare system
leads to a lack of co-ordination, a breakdown in continuity of care, and inefficient use of
resources because of duplication. More specifically, there is a crucial gap in informationsharing
between the emergency department and primary care providers. The aim of this
project was to explore these deficiencies and develop a computer-based network that
would serve as a standardized communication system linking the emergency department
and primary care providers.
The project was completed in the three phases. First, a series of focus groups were held at
three Quebec hospitals with emergency and family physicians to better understand the
state of communication between emergency departments and primary care providers.
Second, an Internet-based communication system was designed to provide physicians
with the information they had identified as important. Third, the impact of the
standardized communication system was evaluated in light of four healthcare delivery
outcomes: improved follow-up by family physicians; reduced resource use; family
physician value; and patient value.
Phase One: Focus Groups
These sessions allowed the research team to gauge the importance of sending patient
information to the emergency rooms from the family physician’s office, as well as in the
other direction. The team also determined what information each set of physicians needed
the most, and how an ideal communication tool should work. The focus groups
confirmed that communication between emergency departments and primary care
facilities is poor, resulting in difficulties organizing follow-up care. Physicians also noted
significant duplication of tests and evaluations, due to a lack of information on what was
done in the emergency department.
ii

Phase Two: Development of the Standardized Communication System
On the morning following the emergency department visit of a patient whose family
physician was a participant in the project, an e-mail message would be automatically sent
to this physician indicating that a patient of his (not identified by name) had been in the
emergency department within the previous 24 hours. The e-mail contained a link to a
password-secured web site where the physician could read a detailed report. The report
contains information on the patient’s presentation to the emergency department, the
results of any examinations performed (electrocardiographic, laboratory, and radiologic),
and consultation reports completed in the emergency department by a physician specialist
or consulting service. There is also information on the patient’s final diagnosis,
disposition (including new medications), and follow-up plan. All reports were made
printer-friendly for inclusion in the primary care provider’s dossier.
For information going in the other direction, emergency physicians were asked if they
required information from a primary care provider for any patient who had consented to
participate in the project. As none of the family physicians in the project used electronic
patient records, primary care information could only be obtained during the hours of 8
a.m. to 4 p.m. When the emergency physician wanted information, a detailed request
would be sent by fax from the hospital to the family physician, and the data would be
faxed back to the emergency department. There were very few instances where
emergency physicians took advantage of the system.
Phase Three: Evaluation
Four questions were asked to evaluate the impact of the standardized communication
system:
1. Does the standardized communication system improve continuity of care for
patients cared for in the emergency department by enhancing follow-up of acute
medical problems and reducing duplication?
2. Does the standardized communication system lead to a reduction in resource use
within the emergency department as a result of enhanced follow-up of patients
discharged from the emergency department?
3. Do primary care providers value the standardized communication system?
iii

4. Do patients value the standardized communication system?
1. The standardized communication system showed significant benefits in continuity of
patient care, including enhanced follow-up of patients by their primary care
physicians. Similarly, primary care providers with access to the system reported far
superior knowledge of their patient’s medical circumstances as they related to the
recent emergency visit. The system did not appear to achieve significant reductions in
the duplication of tests and other consultations at the point where primary care
providers followed up with their patients after an emergency department visit.
2. The intervention did not demonstrate any of the anticipated benefits for emergency
department-based evaluations (fewer tests and consultations in the emergency
department and fewer return visits.)
3. The majority of physicians who participated in this project were extremely positive
about the system, perceiving it as a vast improvement over existing methods of
transmitting information on patients who sought emergency department care.
Physicians reported direct benefits to the quality of care provided to their patients, the
physician-patient relationship and satisfaction with clinical practice. Most family
physicians also felt that the scope of the tool should be widened to include
information on their patients’ hospitalizations.
4. The patients recruited into this trial reported that the standardized communication
system resulted in appreciably enhanced transmission of medical information from
the emergency department to their primary care provider. This rapid transfer of
information was perceived as being beneficial to the quality of the patient care that
the patient’s physician was able to provide.
Limitations
While the project was envisioned as a multi-centered trial involving three Quebec
hospitals, a number of significant technological barriers — primarily a lack of
standardization between hospital computer systems — forced the application to be
implemented only at the Sir Mortimer B. Davis Jewish General Hospital.
iv

Context
The viability and integrity of the health care system is possibly the most important preoccupation
of the Canadian population as emphasized in the recent report of the Commission on the Future
of Health Care in Canada 1 . The public perception of our health care delivery is that it is itself
ailing as a result of deep budget cuts and inefficiency. Policy makers are eagerly seeking
creative and innovative solutions that can reduce the frustration that both patients and health care
providers experience when navigating the system. A possible solution to improve the
satisfaction and quality of care is to address the continuity of care problem. Actual dysfunction
of the health care system is often associated with emergency department (ED) overcrowding.
Recent headlines appearing nation-wide have highlighted this crisis in ED care. Issues faced by
EDs across the country and solutions to solve them are of keen interest to policy makers. A
suggested means of improving continuity of care would be through enhanced linkages between
the ED and the primary care network (PCN) which would result in a more appropriate, efficient
and rational utilization of resources in both milieus. Thus, the aim of this project was to explore
the deficiencies in communication between EDs and PCPs, develop a computer-based electronic
network that would serve as a standardized communication system (SCS) linking the ED and
PCPs and measure the impact of such a system.
Unfortunately, very little communication exists between these two sources of care. Physicians
from both the ED and the PCN must treat and manage patients with missing or incomplete
information. Continuity of care between the ED and the PCN depends on improving
communication between these two settings 2 . Knowledge of a patient’s visit to an ED is critical
for PCPs since it allows them to know that their patient has experienced a major change in their
health status. Perhaps more importantly however, it will enable informed and effective followup
of an ED visit and empower the PCP to act as a real coordinator of care 3 . In order to achieve
these fundamental objectives, the PCN must, at a minimum, have knowledge of the visit, the
diagnosis, and the treatment plan. Such information is all too often unavailable to family
physicians in a format that is both sufficiently detailed and accessible in a reasonable time frame
to allow participation in a patient’s medical care. Family physicians have indicated that
continuity of care is difficult in the PCN because of significant problems with communication
between acute care facilities and the PCN 4 . Family physicians report problems with obtaining
1

key information (test results, consult notes and discharge information) in a manner that is
sufficiently timely and complete to allow for continuity of patient care after hospitalization of
ED care. The few EDs that do contact the PCN with information regarding patient visits usually
do so via mail delivery of hand-written notes. Rubinstein et al. 5 evaluated both emergency and
family physicians’ opinions of such mail-delivered summaries. These reports were actually
carbon copies of the treating emergency department physician’s record and were supposed to
contain data on diagnosis, disposition and pertinent lab tests. The perceived impact of this
system according to the physicians using them (emergency physicians and family physicians)
was decreased speciality referrals by both emergency and family physicians as well as decreased
test ordering and increased visits to hospitalized patients by their family physicians. The family
physicians surveyed in the study however reported significant deficiencies in that same system,
including transmission delays, incomplete and scant content as well as illegibility. Reducing or
eliminating these deficiencies may serve to enhance both the perceived and real benefits.
To achieve high levels of continuity of care and quality of care treating physicians depend on
greater access to information from other health care providers. Three aspects of this
communication must be addressed: 1) the content of the information to be communicated, 2) the
format of the information exchanged and 3) the modality for the communication. Van Walraven
and Rokosh 6 surveyed hospital based physicians-in-training and community family physicians
about the importance of various information elements as they might be included in hospital
discharge summaries. The results indicate a strong consensus between hospital and family
physicians. Elements rated as most important are admitting diagnosis, history of presenting
illness, in-hospital therapeutic procedures, complications during hospital stay, consultations in
hospital, discharge diagnoses, all discharge medication information, arranged medical follow-up,
medical or social issues outstanding at patient discharge, and active medical problems at
discharge. Wass 3 observed that following a patient’s visit to an emergency department, general
practitioners wished to receive details of the results of investigations, diagnosis, treatment plan
and follow-up arrangements. As for important format characteristics influencing the quality and
3, 7,8
use of discharge summaries, studies have concluded that the summary needs to be succinct
and delivered quickly 7 .
2

Regarding the modality of communication, various authors suggest that optimal transfer of
information will be achieved through electronic data interchange systems. For example,
Bernstein et al. 7 propose that the transfer of patients care among health care provider is highly
dependant on standardized and preferably electronic instruments. The study of such systems in
clinical medical practice is however in its infancy, lagging far behind the developments in the
technological realm of devices increasingly available to the public at large as well as physicians.
Feied suggests that EDs are ideally suited to play a leadership role in the development of clinical
information systems. His reasoning stems from the integral role that ED care plays in the
healthcare system as a whole, i.e. portal of entry to acute care hospitals and the interface between
hospital and community resources 8 . The benefits of electronic exchanges have also been
demonstrated. Branger et al. 9 surveyed British general practitioners on the advantages of a
recently instituted electronic data interchange which consisted mostly of laboratory data and
discharge summaries for hospitalized patients within their practice. The results were indicative
of a far more reliable mechanism of data transfer than previous methods (i.e., paper-based
communication via mail delivery) and an improvement in the efficiency of patient care as
reported by the general practitioners. While they reported decreased workloads with electronic
exchanges, they also felt they had more accurate knowledge of their patients as a result of their
use. Increased knowledge of patients can have serious implications for policymakers as it has
been associated with a decreased use of resources (i.e., decreased time needed with patient,
decreased requests for laboratory tests) 10 . Overall, electronic exchange of medical information
has been shown to be highly cost-effective. Atkinson 11 reports that two hospitals having
introduced an electronic health information network estimated minimum annual savings at $200,
000 US.
A wide range of elements including the current state of information exchange, unmet needs and
technological innovations have provided the groundwork for this project. In addition, the health
care system in Canada is in the process of moving the care of patients out of the hospital and into
the ambulatory setting. In the province of Quebec, this movement (virage ambulatoire) has
placed an emphasis on integrating family physicians and community care clinics (CLSCs) into
the management of patients with both acute and chronic diseases. Unfortunately, the
implementation of the “virage ambulatoire” has encountered obstacles that may in part be related
3

to an absence of an efficient communication network. This fundamental shift in health care
delivery hinges upon notifying and informing the primary care network about the in-hospital
component of their patients’ care.
Another key element driving the importance of this project is the current revolution we are
experiencing in information technologies, both in the home and in the health care milieu. In the
latter setting, computers have transformed the areas related to research and education for health
care professionals but have been slower to benefit patients in a more direct manner. This study is
of compelling interest to managers because it showcases an intervention that takes advantage of
the explosion in computer-related technologies and uses it to facilitate the continuity of care that
patients receive within hospitals and the PCN (i.e., family physicians, CLSCs).
The intervention developed for this study was designed to enhance communication and
integration within the health care system. The research focus was also to examine the potential
impact on resource utilization, health care provider satisfaction and patient satisfaction. These
effects, derived from enhanced information transfer between these two settings will compel
policy managers to consider reliable modes of communication as a fundamental component of
health service policy and infrastructure.
Implications
Decision makers in health policy
The SCS project lends support to several of the major recommendations contained within the
report issued by the Commission on the Future of Health Care in Canada led by former
Saskatchewan Premier Roy Romanow. Through major recommendations 8 through to 11, the
report focuses on the importance of continued investment and development of Canada’s health
information technology infrastructure. The Romanow report is largely focused upon the
electronic patient file that would provide health care practitioners with authorized access to
diagnoses, treatments and results. The implications of the SCS project extend beyond this
recommendation and emphasize instead the importance of timely information transfer between
acute care and primary care services in the context of acute illness. Specifically, the SCS project
4

demonstrates first hand the benefits of rapid electronic access to patient information by the health
care professionals, involved in patient care both within the hospital and the community.
In an age of technological innovations the barriers, which prevent health care providers from
obtaining rapid access to information that is crucial to decision-making for their patients, should
be falling rapidly. The success of the SCS project confirms that not only is this type of access
possible but that it is welcomed both by patients and health care providers in the primary care
network. Improved access to this information will empower those providers who as a result of a
lack of information are excluded from the care provided to their patients within the hospital
context and hence in any post-hospital care.
The shift of medical care from the inpatient to outpatient setting as well as important
developments in Primary Care reform will see an expanding role for PCPs in the care of acutely
ill patients. Successfully incorporating PCPs into these realms is going to be contingent on
accurate, timely and unhampered access to the information required for informed decisionmaking.
Decision makers in health informatics
Decision-makers in health informatics can gleam important and relevant conclusions from the
SCS project. Most important of these are the existence of significant technological hurdles that
must be faced when implementing a secure, rapid and reliable informational link between
hospitals and community-based health care providers. Any such project needs to take careful
inventory of compatibility issues with existent hospital databases and the application responsible
for making patient information available to healthcare providers based in the community. The
implication for decision makers in health informatics relates to the necessary foresight that is
required early in the development of the health care information systems so as to assure
standardization of it’s components with an eye to assuring seamless integration throughout the
emerging network as a whole.
5

Managers in the hospital-based and primary care systems
The SCS project supports the notion that enabling quality communication linkages between
hospitals and the primary care network should become a high priority for managers in the health
care system. This implication is the reflection of the conclusions reached throughout the three
phases of the SCS project. Phase I established the deficiencies in communication in the three
Quebec hospitals recruited to participate. Phase II proved that a secure, rapid and accessible link
was possible as long as certain technological aspects were sufficiently addressed. Thirdly, Phase
III confirmed the benefit of the SCS in regards to the continuity and quality of patient care
delivered.
Approach
General Approach
As reflected in the research objectives, this project's methodology also followed a triphasic
approach. The first phase consisted of focus groups aimed at determining the key data elements
necessary for informed clinical decision-making by PCP and ED providers. During the second
phase, a standardized communication tool facilitating the transfer of information between
emergency department and the PCPs was developed and implemented. Finally, in the third
phase of the study, family physicians and their patients were recruited to evaluate the impact of
the SCS on several health care outcomes.
Phase 1: Focus Groups
Identification of the informational elements required by both PCPs and ED teams to facilitate
the comprehensive care of patients seeking urgent care in the ED and are then referred back to
the PCPs for follow-up.
A focus group was held at each participating centre Sir Mortimer B. Davis – Jewish General
Hospital (Montreal), Hôtel Dieu de Lévis (Quebec City), and Hôpital Charles LeMoyne
(Longueuil) in order to obtain balanced perspectives from key health care providers which could
in turn contribute to the generalization of the views obtained.
6

Focus groups brought together emergency and primary care physicians from the same region in a
collaborative and non-confrontational environment for the purpose of encouraging a partnership
for further phases of the study. These sessions explored current difficulties encountered in the
transfer of pertinent clinical information between the ED and physician offices in the primary
care setting. These focus groups were also set out to define the key data elements necessary for
PCPs and emergency physicians (EPs) to take informed clinical decisions. Two animators
moderated the meetings, one representing the ED and the other representing the PCPs. The
animators presented 2 case scenarios that were used as a springboard to focus discussions
relating to specific questions pertaining to issues of information transfer (annex A). The focus
groups provided information on the key elements (content and format) that would facilitate the
development of the ideal communication tool. The research team also took this opportunity to
introduce details associated with the second and third phases of the SCS research project with the
intent of fostering interest in the project’s development.
Throughout the course of the focus group session, the research team recorded elements that were
felt by the participants to be an important component of the SCS application. The focus group
concluded with the distribution of a survey that measured the PCPs rating of key elements
(annex A). Ratings on the potential impact that the SCS tool might have on resource utilization,
continuity of care, physicians’ satisfaction (PCP & ED), patient satisfaction and patient
management was also collected (annex A). The focus group sessions were tape-recorded and a
transcript was produced (available upon request).
Data on the importance of key elements and the potential impact of the tool were compiled and a
report was produced (annex A). A document summarizing the elements present in the SCS tool
was mailed out to all physicians that took part in the focus group process (annex B).
Phase 2 : Development of SCS
Designing, testing, and piloting electronic communication modalities which would best be suited
for the receipt and delivery of the informational components identified during Phase I.
7

Call for tenders
A document incorporating all the needs of the three hospitals was created (annex C) and
distributed to four companies specialized in health informatics technology; Gerfo, Oracle,
Medisolution and MBS. Subsequent to presentations made by these four firms and after
consulting with the Information Technology Department of the Jewish General Hospital, it was
decided that Oracle was best suited to design and program the SCS tool. Oracle was selected
because they presented the best proposal and possessed the requisite expertise in the medical
field.
Design of SCS
Medical information identified to be important from the first phase of the study required
integration into the SCS application. A steering committee was formed with representatives
from Oracle, the hospital Information Technology Department, the ED research division, ED
data management systems and the SCS study coordinator. Personnel from Oracle began the tool
development by first analyzing the ED data environment in terms of data collection and it’s
management. A prototype of the tool was then produced. Close collaboration between Oracle
and the steering committee continued throughout the development of the application (format,
functionality, performance, etc.). After weeks of programming and format adjustment, SCS took
its final form (annex D). Extensive data validation was made to assure quality and integrity of
the information being delivered. Project Progress Reports (annex E) were made by Oracle as
part of their commitment to both the project timeline and the budget. Development, testing and
validation extended over a period of 5 months.
In its final form, the SCS application could best be described as a secure, rapid web-based
application accessible from any computer with Internet access. The SCS application was
developed to maximize user friendliness, reliability and security through a 128-bit level of
encryption.
8

Content and format
SCS application was meant to be fully automated, allowing instant retrieval of laboratory and
medical imaging information from a central location i.e. the data warehouse databases (annex D)
of the Jewish General Hospital (JGH). SCS was developed in parallel with a data warehouse
(centralized electronic location of all hospital data) project at the JGH also led by Oracle. There
were, however, important delays in the realization of the data warehouse project. Pressured by
the SCS timeline, patient recruitment began before data from radiology, hematology,
biochemistry and microbiology laboratories were integrated into the application. Therefore, all
data that could not be automatically integrated, such as laboratory testing and radiology, were
scanned and entered by research assistants (RAs) in the SCS database as a transferable image
file.
Medical information related to each visit was presented in 6 sections labeled: About the patient,
About the visit, About the tests, About the consultations, About departure and About the
discharge (annex D). The SCS tool provides visit specific information, therefore multiple visits
for the same patient would be signalled as different visits.
Notification to family physician
For every authorized intervention visit (patient consented to the electronic transfer of
information) the SCS application would send out at 6:00am an email inviting the identified
family physician to log on to the application and seek information on the latest visits. No
numerical data (RAMQ number, hospital card number, etc) was present in the email for reasons
related to security. A direct link to the web site was present in this email (annex D) to facilitate
the use of the application. The physician then logged on with their username and password
before accessing any medical information. Regardless of which group the FPs were in, twentyone
days after the ED visit, they received an email asking them to answer an electronic
questionnaire (annex F) used to assess the outcomes of the study.
Phase 3: Evaluation of the impact of the SCS
Evaluation of the impact of the SCS on various outcomes pertaining to health care delivery.
9

Implementation of the tool
Sir Mortimer B. Davis – Jewish General Hospital, Hôtel Dieu de Lévis, and Hôpital Charles
LeMoyne were ED research sites that had expressed a firm commitment to this project. These
sites were approached because of their interest, their informatic infrastructure and their
geographical location. However, despite the interest in participation from the major stakeholders
(emergency physicians, primary care givers and directors of EDs), the information technology
department of Charles LeMoyne and Hôtel Dieu de Lévis could not assure the manpower nor to
support the SCS application. The decision to cease collaboration with the 2 centres necessitated
a review of the study methodology.
Study design
The following elements influenced the choice in study design. First, a randomized, controlled
though unblinded study design was necessary to best appreciate the effects of the SCS
intervention. Secondly, family physicians rather than patients were chosen as the unit of
randomization to create two similar cohorts. This decision was mandated by the importance of
preventing contamination within a PCPs practice at any one time. Thirdly, to counter the effect
of any cluster phenomena, it was necessary for all physicians participating in this study to
experience the SCS. Therefore, a triple cross-over design was employed. The advantage of this
design also allowed for adjustment of any lag effect that might occur as a result of the learning
phase encountered through the cross-over between the control and intervention arms. The triple
cross-over allowed for each physician to be exposed to 2 intervention periods and 2 control
periods. Finally, to ensure balance in PCP practices, stratified randomization was used to ensure
that PCP were allocated to each cohort according to the age and size of their clientele presenting
to the ED.
Period I
Period II
Period III Period IV
2001/06/15 2001/08/25 2001/11/03 2002/01/18
to 2001/08/24 to 2001/11/02 to 2002/01/17 to 2002/03/22
Cohort #1 n=11 Intervention Control Intervention Control
Cohort #2 n=12 Control Intervention Control Intervention
10

The PCPs in the control arm were provided with information about their patients’ ED visits
through a mailed carbonated copy of ED notes which is standard procedure at the ED of SMBD-
JGH. The PCPs in the intervention arm, were notified (through SCS) about their patients’ ED
visits and could get medical information specific to those visits. One week prior to every crossover,
family physicians received a letter advising them when the cross-over would occur. Status
(intervention or control) of PCPs was changed (administrator access level) on the cross-over
date; blinding (control) or permitting (intervention) them to know about their patients’ ED visits.
Sample size determinations
Sample size estimations were based on the expected change in the incidence of return visits to
the ED and ED length of stay (LOS) of patients recruited in the control and intervention arms.
The incidence of return visits within 14 days was estimated at 14% and that of LOS at 9 hours
from JGH statistics. Using a power of 80% and an alpha of 5%, the study would be sensitive to a
minimum absolute difference of 5% in return visits to ED and a 1.8 hours decrease in LOS with
a sample size of 1000 visits per group.
Family physician sampling
Family physicians (FP) whose patients visited the emergency department most frequently were
approached for participation in the study. In all three centres, 70 family physicians were invited
to participate. Although the study could only be done at the JGH, family physicians from the
other two centres had expressed an interest in participating.
The FPs were made aware that their participation would involve the following: a one year
commitment to the study, answering a general questionnaire before and after the study (annex F),
checking their email on a daily basis, having or not access to the SCS intermittently and
responding to an electronic questionnaire for each patient visit to the ED. FPs were equipped
with computers, printers, appropriate software as well as internet hook-up. Access to the web
was made available for them through a telephone line. All family physicians were given a “step
by step” manual (annex G) and had a 2 hour training session by the study coordinator.
11

Questionnaire development
Questionnaires were developed to evaluate the impact of SCS on our primary and secondary
outcomes. In all, five questionnaires (annex F) were developed. Table 2 indicates who
completed the questionnaires and when.
Table 2 : Study questionnaires.
Questionnaire Person that completed Time of completion
Electronic Family physician 21 days after patient ED visit
Patient satisfaction day 1 Patient Visit to ED
Patient satisfaction day 21 Patient By telephone 21 days after ED visit
FP Pre Study Family physician Through pre focus group
FP Post study Family physician Through post focus group
The electronic questionnaire was developed to assess resource utilization out of the ED, any
duplication of test ordering, continuity of care (knowledge of patient health status, follow-up of
ED care) and PCP satisfaction (knowledge, patient management, etc). The electronic
questionnaire was sent through an e-mail link and answered via the internet (annex F). The SCS
tool sent a link to the PCP (email) allowing him/her to respond to a maximum of 21 questions.
Internet questionnaire was chosen to make it less time consuming for PCP. The PCPs stated that
it took about 10 minutes to answer and submit a questionnaire.
The patient satisfaction day 1 questionnaire was completed when the patient was recruited in the
ED, the day 21 questionnaire was done by telephone. These 2 questionnaires assessed patient
satisfaction with respect to the communication between FP and EP, perception of FP’s
knowledge about ED visit, perception of family physician, perception of pain and severity of
illness, use of health care system, access to health care system and socio demographic
information. The questionnaires were design to be filled out by the patients directly or with the
help of the RA. The ED questionnaire took approximately 25 minutes to complete.
Pre and post study questionnaires were aimed at collecting PCP socio demographic information
and assessing their appreciation of the outcomes which are also measured in the electronic
12

questionnaire (i.e. perception of resource utilization, duplication of test, continuity of care and
satisfaction with knowledge and management of patient).
Patient recruitment in the ED
The Research and Ethics Committee of the SMBD-JGH allowed for patient recruitment to
commence after approving the patient consent and questionnaire forms. Patient enrolment
occurred on weekdays from 8:00am to 10:00pm except on statutory holidays. Two research
assistants conducted patient recruitment between 8:00am and 4:00pm; one research assistant
enrolled patients while the other entered their medical information into the SCS database
(consultation reports, emergency physician comments, follow up plan, discharge medications,
blood test results, maintained the log, telephoned patients). Between 4:00pm and 10:00pm
patient inflow was slower, therefore only one RA was needed for both tasks. After surveying at
what time PCP offices closed, decision was taken not to recruit at night because of the incapacity
of bi-directionality of information transfer.
Every 30 minutes, the application sorted out any new ED visits of patients from the 23 FP
participating in the study and indicated them on a SCS patient list. Research assistants would
subsequently take that list and approach all patients appearing on it. RAs approached patients
only after they had been triaged. The RA would explain the study to the patient and obtain
informed consent. It was also the RA’s responsibility to make sure the application had correctly
identified the patient’s PCP and to verify patient inclusion/exclusion criteria. A patient was
eligible if he/she was ≥ 18 years, spoke either French or English or had a translator and was an
active patient of the family physician identified (had seen their FP at least once within the last 2
years). A Short Portable Mental Status questionnaire (annex H) was also administered on every
candidate approached. If the patient made 5 errors or more, the patient was excluded according
to instructions from the ethics committee of the JGH. Informed consent was required for every
visit to the ED. Patient consent meant agreeing to the electronic transfer and also allowing their
family physician access to their medical information. Patients could also agree to answer a
questionnaire assessing patient satisfaction. The consent forms were designed to give patients
the choice to agree to one and/or both of the above options. The time duration for completion of
patient enrolment was 12 months.
13

Information transfer from ED to PCP
Modes of communication already in place between the ED and family physicians were
maintained over the study period. The standard method of communication at the JGH-ED is to
send a copy of the ED notes by mail to the family physicians associated with the JGH of patients
presenting to the ED. The copy is the first page of the patient’s ED notes that may include
information similar to the SCS. This method was used for the control group. As for the
intervention group, information transfer was through SCS as well as through the standard mode
which is the sending of a carbonated copy of the ED note by mail.
Information transfer from PCP to ED
For consented patients (intervention and control arms), medical information was collected from
the chart and emergency physicians. For the visits in the intervention arm, patient charts were
flagged (annex I) so that the treating ED physician could identify that patient is part of the study
(intervention arm). Emergency physicians could then ask the research assistants to contact the
FP and obtain the information requested. To facilitate the medical information transfer requested
by the ED, FPs offices and the ED were equipped with dedicated fax lines. As well, the FPs
secretaries were instructed to give priority to such requests. A fax form was specifically
developed (annex I) for the study. A RA would communicate the information faxed back from
the PCP office to the ED physician.
Evaluation of the impact of SCS tool
The first part of the analysis was to compare if the control and intervention arms were alike as
for socio demographic. Six variables were used for baseline comparison: age, gender, LOS,
admission, ambulance and stretcher. The primary outcomes included: resource utilization in and
out of ED (in FP office) and continuity of care. Secondary outcomes included: FP satisfaction
with SCS tool, FP practice satisfaction, FP perceived knowledge of patients, patient satisfaction
and patient’ representation of their FP. For each outcome, comparisons were done between the 2
groups (intervention and control). Due to the study design the analysis on resource utilization
and continuity of care were performed considering the cluster (FP cluster) and lag effect (crossover
adaptation curve to new group).
14

Resource utilization was defined as: LOS in the ED, number of hospital admissions, number of
tests ordered in the ED, number of specialist consults in the ED, number of tests ordered by PCP,
number of specialist consultations requested by the PCP, duplication of tests, number revisits to
the ED, and re-hospitalization rates. Data to assess this outcome came from two different
databases. Resource utilization in PCP offices was assessed through the electronic questionnaire
database (DB). The electronic questionnaire DB provided information on the different actions
undertaken when the PCP was notified of an ED visit of their patient and what tests or
consultations were requested. Resource utilization within the ED was studied through the SCS
database, incorporating both automated (disposition, diagnosis, etc.) and manual data entered by
the RA. This DB displayed all the tests and consultations done in the ED and disposition.
Continuity of care was defined by S.M. Shortell as:
“Medical care received as a coordinated and uninterrupted succession of events consistent with
medical care needs of patients”
The variables targeted to evaluate the impact of SCS on this outcome were: acknowledgement of
receipt of information, content of information transfer, PCP knowledge about ED visit, action
undertaken following notification of patient’s ED visit, ED follow-up visit to PCP and PCP
satisfaction with patient management.
PCP satisfaction was evaluated pre and post SCS study and was measured on a 5-point scale
where 1 corresponded to the least satisfaction and 5 the most satisfaction. Usefulness, precision,
completeness of information and PCP satisfaction towards clinical practice were assessed in both
study arms. The 5-point scale response was later changed to a dichotomous scale with 1
representing satisfied (levels 4, 5 in the old scale) and 0 unsatisfied (levels 1, 2, 3 in the old
scale) because of the smaller sample size.
Patient satisfaction was evaluated through day 1 and day 21 patient questionnaires. It was
measured using questions responding again to a 5-point scale. Patient responses were collected
at day 1 and day 21 respectively. Double data entry was performed and databases were
subsequently cross-matched. Differences were queried to ensure data quality. In order to
15

account for this repeated measurement effect for each patient, the difference of the two responses
was calculated and considered the outcome variable.
Working with Decision makers
The research project was well known and supported by Mr. Henri Elbaz, executive director of
the SMBD-Jewish General Hospital, and the Information Technology Department. Mr. Elbaz
was instrumental in showcasing the SCS project both within the hospital (executive committee)
and to other decision makers at the regional and provincial level. The project was presented
upon request to Mr. Boisvert (August 2002), President of the Association des Hôpitaux du
Québec and to Mr. Culver, Chairman of the McGill University Health Center (August 2001).
Furthermore, the hospital is actively pursuing new technologies to widen information transfer
processes between hospital and community health care professionals. We are proud to say that
SCS was the springboard bringing forth such endeavours.
Dissemination
The final report from this project will be transmitted through a co-principal investigator, Dr
Afilalo, to the MSSS, more specifically to the CECCNU. The latter is a committee that is
responsible for studying the overcrowding crisis present in Quebec EDs and ways of resolving it.
Two abstracts have been submitted to the Society of American Emergency Medicine for
presentation at their annual meeting for 2003, to be held in Boston, MA. Collaboration with the
Information Technology Department of SMBD-Jewish General Hospital allowed for new
insightful experience, obtained throughout the study, to profit the informatics community. The
final report will be distributed to AHQ, MUHC and other health care institutions and
organisations that could benefit from the study.
The results will also be reported in at least two peer-reviewed scientific publications (i.e., one
elaborating on the development of the instrument and the other discussing its evaluation,
utilization and implementation) and in as many communications in scientific conferences.
16

Results
Phase 1: Focus Groups
Key elements to be integrated into the SCS electronic tool and possible impact of SCS are listed
below as well as a few quotes from the primary caregivers themselves. They reflect their
attitudes and perceptions about being part of the present health care system.
“We are completely out of the loop”
“We now have proof that they have gone to an emergency room which is better than a few years
ago. Now we know we’ve totally lost them”
“Patients come to the office and say: they gave me this thing, I need a consultation, they told me
you would take care of it! Take care of what! I am acting under somebody’s orders now, without
getting any information? It’s ridiculous!”
“The most important thing is that we get a report”.
The following graphs show the importance of each element mentioned and the impact that the
SCS tool could have.
17

A post study focus group was held three weeks after patient recruitment ended. Here are a few
comments expressing the PCPs’ thoughts regarding the SCS.
“I’m just going to say, that in this day and age of technology, to have patients come in and to
your office and say:’ Did you get any information from ED? Did they send you anything about
that stroke I had?’ You sit there and ask what stroke.”
“So when I went on to the control group after 3 months, I got withdrawal symptoms and was
absolutely miserable.”
“The most amazing experience with the whole project was with a patient of mine who was dying
of cancer, who has died. He called me at let’s say 10 o’clock in the morning and said he just got
home from the emergency. He asked me: ‘ What does it mean if this certain result is elevated?’.
I saw the result right in front of me (on SCS) and I was able to reassure his fears or at least
understand what he was asking me. This was the most gratifying experience of the whole
project.”
“The whole concept is super. I’m for it 100 %. I want the SCS. I want every data available.
But listen, patients (over 100 or maybe 200 yearly) are discharge from ED and then come into
my office. Who’s going to cover the cost of time, ink, cartridges, paper to print out summaries
and results to be put in the patients’ charts.”
Phase 3: Impact of SCS
The present section describes the general results of the impact of SCS. These include patient
recruitment, resource utilization, continuity of care and PCP satisfaction. For a more detailed
discussion of the following results, please refer to Annex J: Statistical Analysis Report.
18

Patient enrolment
Patient enrolment diagram
Eligible population
n=3168
Approach
Missed
n=2651 (84%) n=517 (16%)
Excluded Recruited Refused
n=179 (7%) n=2022 (76%)
n=450 (17%)
1622 patients
Intervention
Control
n=1048 (52%) n=974 (48%)
Baseline characteristics indicated that patients are alike in both study arms.
70
60
Control(n=978)
Intervention(n=1048)
60
Control(n=978)
Intervention(n=1048)
50
50
40
Pourcentage
40
30
Pourcentage
30
20
20
10
10
0
Mean LOS (hour) Female Mean age (year)
0
Ambulance Admissions Stretcher
19

In comparison with mailed copies of the ED chart, electronic communication between ED and
PCPs did not reduce resource utilization in and out of the ED.
Resource utilization in ED
Resource utilization in ED; Revisit
90
Control
30
Control
Intervention
80
Intervention
25
70
60
20
Pourcentage
50
40
30
Pourcentage
15
10
0.3 vs 0.4 visit/pt
20
5
10
0
ED LOS Consultation Exam Admission rate
0
Revisit /patient Returning patients Admission on Revisit
Resource utilization in PCP office
Data for this outcome came from the electronic questionnaire, which had a 77% response rate.
2022 questionnaires sent
1566 questionnaires answered
Response rate 77%
Control n=739 (74%) Intervention n=827 (81%)
20

Investigation or controlling test
Resource utilization in PCP office
25
45
Control
40
Control
20
Intervention
35
Intervention
30
15
Pourcentage
10
Pourcentage
25
20
15
5
10
5
0
Ancilary serv. ECG Microbiology Radiology Hematology Biochemistry Consultations
0
Ref. back to ED Other action Related FU No action
Enhanced transfer of information between the ED and PCPs improved continuity of care,
primarily through improved PCP follow-up of ED visits, and better knowledge of the care
provided to their patients in the ED.
Good knowledge & pt management
Reception of information elements
100
90
80
Control
Intervention
90
80
Control
Intervention
70
70
Pourcentage
60
50
40
30
Pourcentage
60
50
40
30
20
20
10
10
0
FP knowledge
Sat. pt management
0
Consultation report
ECG
Radiology
Labs
Follow up plan
Disposition
Treatment
Reception of information
Diagnosis
21

Visit to PCP and reason of visit
80
Control
Intervention
70
60
50
Pourcentage
40
30
20
10
0
Visit to FP
FU from ED visit
PCP satisfaction was significantly increased following the SCS experience.
PCP satisfaction
100
90
Pre
Post
80
70
Pourcentage
60
50
40
30
20
10
0
Less error complete less time lost More useful clinical Sat
Patient satisfaction
Responders to patient satisfaction questionnaire were generally younger, of female gender,
ambulatory, arrived without ambulance, required fewer admissions and had a smaller LOS. This
implies that results could not be generalized to the whole study population. With respect to the
communication between ED and PCPs, patient satisfaction increased in the intervention group.
22

No statistical significant difference was observed between intervention and control arms with
regard to patient perception of PCP and of PCP knowledge about ED visit.
V-additional resources
The initial protocol submitted to the CHSRF can be found on the JGH web site. News letters
were produced and sent throughout the study to all collaborators, administrators, decision makers
as well as granting partners. All documents reporting on study developments can also be found
on the web site: http://www.jgh.ca/departments/emergency/res_projects_e.html. Moreover, e-
mail addresses of all principal investigators and collaborators are referenced on this site. The
SCS study was presented at a number of conferences either in a poster or oral presentation
format. The hosting organizations included the Society of American Emergency Medicine
(SAEM); the Canadian Association of Emergency Physician (CAEP) and the AMUQ
(Association des Médecins d’Urgence Du Québec). A list of the abstracts presented and of the
posters developed is included in annex B.
Further research
“The electronic transfer of information to and from the ED, the community family physician, and
other institutions is vital and perhaps the greatest priority. In fact we would suggest that without
the initial implementation of electronic information systems in EDs, the link between the
community and the hospital will not exist and the integration of hospital and community cannot
occur.” Canadian Association of Emergency Physicians (CAEP) Statement to
Commissioner Roy J. Romanow, November 15 2001
Clearly, the need to find ways to better integrate the work of health care providers is a nation
wide sentiment (CAEP). The present project was limited in scope in that the transfer of
information targeted the primary care providers. Future research should build upon this
experience and work toward investing in the technological infrastructure that would facilitate
access for all health care providers.
Although the SCS was evaluated as being useful, accurate and complete in the transfer of
information, further investigation would be required into the development of electronic transfer
23

methods that are seamless and bi-directional between the community and the hospital.
Furthermore, there is a need to design systems that allow the viewing and updating of patients
records by physicians that are jointly managing patients both within the hospital and the
community. Technical issues such as integrated laboratory and imaging tests, quickly accessing
and adding information about patients will still need to be addressed.
The critical question of financing this shift towards more advanced technology remains yet to be
answered. Standard and affordable systems for hospitals and the community settings still have to
be designed and evaluated.
Finally, ethical issues concerning the sharing of electronic patient records with regard to
confidentiality and patients privacy are of foremost importance and could be explored further by
lawyers and bioethiciens.
24

References
1. Romanow, R.J. commissioner. Building on Values: The Future of Health Care in Canada –
Final Report, Commission on the Future of Health Care in Canada, National Library of
Canada cataloguing in publication data, 2002.
2. Wass AR, Illingworth RN. What information do general practitioners want about accident and
emergency patients? Journal of Accident & Emergency Medicine 1996; 13(6):406-408.
3. Bodenheimer T, Lo B, Casalino L. Primary care physicians should be coordinators, not
gatekeepers. JAMA 1999; 281(21):2045-2049.
4. Geiger S, Hum S. Results of a survey of family physicians toward electronic patient records.
Unpublished manuscript 2000.
5. Rubinstein H, Levitt C, Duart-Franco É, Afilalo M, Rosenthal SE. L'amélioration de la
communication entre les médecins d'urgence et les médecins de soins primaires. Le Médecin
du Québec 1997; 67-72.
6. van Walraven C, Rokosh E. What is necessary for high-quality discharge summaries?
American Journal of Medical Quality 1999; 14(4):160-169.
7. Bernstein RM, Hollingworth GR, Viner GS. Something old, something new, something
borrowed. A review of standardized data collection in primary care. Proceedings - the
Annual Symposium on Computer Applications in Medical Care 1994.
8. Feied CF, Smith MS, Handler JA, Kanhouwa M. Emergency medicine can play a leadership
role in enterprise-wide clinical information systems. Annals of Emergency Medicine 2000;
35(2):162-167.
9. Branger PJ, van der Wouden JC, Schudel BR, Verboog E, Duisterhout JS, van der Lei J et al.
Electronic communication between providers of primary and secondary care. British Medical
Journal 1992; 305, 1068-1070.
10. Hjortdahl P, Borchgrevink CF. Continuity of care: Influence of general practitioners'
knowledge about their patients on use of resources in consultations. British Medical Journal
1991; 303, 1181-1184.
11. Atkinson B. Wisconsin cities win with WHIN (Wisconsin Health Information Network).
Infocare 1995; 36-40.
25