How physicians integrate advances into clinical practices

HOW PHYSICIANS INTEGRATE ADVANCES INTO CLINICAL PRACTICES Jocelyn M.Lockyer, M.H.A. John T. Parboosingh, M.B., F.R.C.S.(C) Gerald M. McDougall, M.D., F.R.C.S.(C) Urmil Chugh, M.A.

Family physicians and specialists were asked to identifjl the sources of information they used in theprocess of making changes in their clinical practices. Wile physicians identified journals, courses, pharmaceutical representatives, colleagues and consultationsfrequently as initial sources of information, the anticipated benefit to patient care was most Pequently identified as the final motivatingfactor: Socio-demographic factors including specialty status (us. family physician], age, community size, and base hospital were found to influence the information sources used by physicians to integrate advances into their clinicalpractices. An average of 3.08 sources of information were utilized for each change and over j@?ypercent of the changes were complete in less than one year The authors discuss the implications of their findings for providers of continuing medical education. elays frequently occur between the time D new scientific information is disseminated and the time is incorporated into clinical pracit

tice. Chalmers (1) found that as many as ten or fifteen years following the conduct of well controlled trials, physicians may b e prescribing drugs which are contraindicated, or giving unsubstantiated advice. Stross ( 2 ) surveyed family physicians and internists to determine their knowledge of photocoagulation in diabetic retinopathy two years after results of a well constructed study were published. Only 28 percent of the family physicians and 46 percent of the internists were aware of the study results and Q 1985 by The Regents of the University o f California

for MOBIUS Vol. 5, No. 2, April 1985

only 21 percent of the family physicians and 42 percent of internists correctly identified the appropriate management of two patient problems in which photocoagulation was indicated. Stross ( 2 ) asked the physicians who indicated a knowledge of photocoagulation in diabetic retinopathy to identify where they had learned of the study results. More than two-thirds indicated an ophthalmologist or a colleague and 25 percent mentioned journals. To date, only preliminary work has been done to delineate the sources of information physicians use in the adoption of medical innovations. In an early study by Coleman et al. ( 3 ) , the researchers traced the adoption of a new pharmaceutical compound by 216 physicians in four communities over a 16 month period. By the end of the study period, 85 percent of the physicians were prescribing the drug. They determined that 57 percent of the physicians first learned about it from a detail man and 33 percent brom journals or other reading material. Ninety percent of their population had received information about the drug from at least one other source before adopting it and 62 percent indicated that they had heard of the drug &om three or more sources. The final source of information was a colleague 28 percent of the time, a professional journal 21 percent of the time, and drug house literature 21 percent of the time. More recently Geertsma et al. ( 4 ) interviewed 66 physicians from five specialties and asked each physician to identify changes they had instituted in their clinical practices. They found

6 MOBIUS that while the change process may be initiated Methodology by any one of a large number of information sources, physicians rarely changed their prac- Participant selection Seventy-four family physicians and 86 specialtices without seeking information from journals or their colleagues. In the Geertsma study, data ists were invited to participate in the study. relating to the adoption of new investigations Forty-two family physicians were randomly seand new technical procedures were not analyzed lected from a list of physicians practicing in separately. Given the potential differences in Calgary. A convenience sample o f rural family the information sources physicians may utilize physicians was obtained by interviewing 12 phybefore adopting a new lab test, such as the sicians in three rural communities in Southern hemoglobin AlC test for monitoring diabetes Alberta and a further 20 from among those control, and the adoption of a new surgical attending the 1983 University of Calgary Family procedure, the researchers felt changes in clin- Practice Review and Update Course. Similarly, a ical practice were better subdivided into the convenience sample of 30 internists, 31 surgeons adoption of new drugs, new laboratory or radio- and 17 gynecologists was obtained from among logical investigations and new technical pro- those attending the 1983 annual scientific meetcedures. Previous research by the authors ( 5 ) ings of the Royal College of Physicians and comparing the adoption of innovation between Surgeons of Canada and the Society o f Obstesurgeons and internists indicated there were tricians and Gynecologists of Canada. A further differences between the adoption of a new drug eight gynecologists practicing in Calgary were and a new technical procedure. In particular, interviewed. The costs of travel to interview differences were related to the length of time family physicians in rural communities and sperequired for adoption and the use of different cialists resident in Western Canada precluded information sources. While the common sources the possibility of a random sample. Relatively of information which introduced specialists to few (less than 5%) of the physicians invited innovations were medical journals, CME courses to participate in the study refused to be interand discussions with colleagues, these were viewed, mainly because of their busy schedules. rarely cited as the factors which induced the All of the physicians were practicing in Canada. change in practice. The anticipated benefit from Full-time faculty members and physicians in the new management and the availability of the practice for less than two years were excluded new drug or lab test were perceived as funda- from the study. mental in the final decision to implement the innovation. Over 50 percent of the changes were The interview The structured interview technique was used. completed in less than twelve months and each involved an average of three different sources At each interview the participants were requested to identify two or three changes in the manageof information. The present study was designed to examine ment of specific clinical problems introduced family physicians’ perceptions of the sources of into their practice within the last two years. information and the time frame involved in their On identifying a change in practice, physicians adoption of an innovation and to compare the were asked whether they could name the initial results with those obtained from a group of source of information that made them aware of the potential change, what other sources of specialists. information (if any) they researched before introducing the change in practice, and whether they could identify a precipitating factor which

LOCKYER. et al. 7

convinced them to introduce the change. Last- sources before making a change in their clinical ly, they were asked to identify the time inter- practice. The number of sources cited for each val between first hearing of the change and change ranged from six (four of the 342 changes) implementing it. A self-administered question- with a mean of 3.08 for all changes. The family naire, completed at the conclusion of the inter- physicians averaged 2.9 sources per change and view, provided socio-demographic data o n the the specialists 3.3. These differences do not achieve statistical significance. Canadian and participants. United Kingdom graduates used fewer informaData analysis tion sources for technical changes than gradThe physicians’ responses in 79 interviews uates of “other” (European and Asian) schools. were audiotaped and the data were indepen- Whereas United Kingdom graduates completed dently analyzed by two investigators. Inter-rater only 8.4 percent of their changes using four or more sources of information, Canadian gradreliability was found to be 96 percent. For the purpose of analysis, the physicians uates completed 34.5 percent and the graduates were allocated to either the family physician of “other” schools completed 70 percent o f or specialist group. The physicians’ responses their changes by utilizing four or more sources to the interview questions have been categor- (Canada and UK vs “other,” p