Mark Ebell, MD, MS, has been accepted into the Robert Wood
Johnson Foundations Generalist Faculty Scholars Program. 07/01/98
— 6/30/2002 - $238,959.
MENTOR(S): David Rovner, M.D. (primary mentor); Bernard Ewigman,
M.D., M.P.H.; Margaret-Holmes Rovner, Ph.D.; Michael Hagen, M.D.;
Lorne Becker, M.D.
PROPOSAL: While physicians generate an average of two clinical
questions for every three patient encounters, they usually do not
answer them. Key barriers include a lack of convenient access to
information at the point of care, the time needed to search for
information, and unfamiliarity with evidence-based sources of
information. The proposed project will address these barriers by
first describing the information needs of generalist physicians,
then designing an intervention which increases the ability of
physicians to answer clinical questions at the point of care.
Specific aims are to 1) describe the information needs of
generalists physicians along 5 axes: the type of need; the heirarchy
of the information need; the type of information; the organ-system;
and the source of information., 2) study whether different
generalist specialties have different information needs, 3)
observe which types of evidence-based information generalist
physicians actually use in practice, and 4) see whether rapid
access to evidence-based information improves the ability of
generalist physicians to answer clinical questions. The results of
this work will help researchers and educators better address the
information needs of physicians, particularly at the point of
care.
Phase I: Survey of information needs of generalist
physicians We will begin by doing a qualitative interview with
3 pediatricians, 3 internists, and 3 family physicians.
Information from the qualitative survey will be used to develop a
quantitative instrument to evaluate the information needs of
generalist physicians along the axes described above. The
quantitative instrument will present physicians with several
vignettes describing a patient-physician interaction, and ask them
to describe their information needs and whether they would pursue
them. This instrument will be pilot-tested on 20 generalist
physicians, and the reliability, content validity, and construct
validity evaluated. The instrument will use the World Wide Web, in
order to be able to survey a large number of physicians, at their
convenience, using a rich set of branching vignettes (including
photos and other information to enrich the vignettes). Once
validated, we will survey 500 generalist physicians, recruiting
them from among the members of MSU practice-based research
networks (PCRNet, UPRNet, and MIRNet) and national primary care
listservers.
Phase II: Direct observation of the information needs of
different primary care specialties In the second study, 30
PCRNet generalist physicians who have expressed a readiness to
adopt a computer-based information tool will each be observed for
two days by a research assistant in the outpatient setting (PCRNet
is a research network of family physicians, general internists,
and pediatricians affiliated with MSU). The research assistant
will determine the number of clinical questions generated and
answered with each encounter. Recognized information needs will be
recorded and classified along the 5 axes described above,
including whether the question was pursued and satisfied.
Phase III: Impact of rapid access to evidence-based
information on physician behavior The group of 30 PCRNet
physicians studied in Phase II will be randomized to intervention
and control groups. Both groups will receive training (a Web-based
tutorial) on evidence-based medicine, and will receive individual
instruction in the use of the handheld computers. The intervention
group will receive pen-based handheld computers with
evidence-based information (350 Cochrane abstracts, 350 critical
appraisals, drug information, patient education materials, and
characteristics of diagnostic tests and history and physical exam
maneuvers) all indexed and searchable within 15 seconds using a
uniform interface. They will be asked to use the software for 6
months to answer their clinical questions. The software will track
how often it is used, which functions and resources are used, and
for how long. At the beginning and end of the evaluation period,
physician satisfaction with the system, the number of times they
used the software per clinical encounter, and their knowledge of
key practice guidelines, diagnostic test characteristics, and
recent treatment recommendations will be re-evaluated. At the end
of the 6 month period, the research assistant will again observe
the physicians for two days, to identify any differences from the
baseline measurements of question formulation and question
answering behavior.
Top