Use of standardized patients in the assessment of medical practice.

  • Robyn Tamblyn
  • Published 1998 in
    CMAJ : Canadian Medical Association journal…


Since the inception of the Canadian Task Force on the Periodic Health Examination in 1976, we have had a mechanism in Canada of identifying preventive interventions that, if instituted by Canadian physicians, are expected to improve the health of the country’s population. Yet substantial disparities appear to exist between optimal institution of preventive care recommendations and actual practice. In articles presented in this issue, Dr. Brian Hutchison and colleagues have attempted to overcome some of the limitations of previous studies by means of blinded evaluation of practice using standardized patients (pages 185 and 197). “Standardized patients” are people trained to present, accurately and reproducibly, the problems that would be evident in a real patient. Theoretically, blinded assessment of practice by means of standardized patients should provide a number of scientific advantages over more conventional methods of practice assessment. First, researchers can select the exact problem they wish to study and can prospectively establish and collect information on case-specific standards of performance. Second, the need to adjust for differences in the kinds of patients seen by different physicians is eliminated because the “same patient” is presented to all physicians. Third, the problems of incomplete documentation with chart audit and overestimation in self-reporting are addressed by having the standardized patient record the actions taken by the physician during the visit. When this information is coupled with chart documentation, as well as the referral, test and prescription recommendations generated by the visit, a more comprehensive and accurate picture of a physician’s practice should emerge. The use of standardized patients has a place in the methodological armamentarium of health care research, but like all methods it has pitfalls. The “valueadded” component only prevails if 5 conditions are met: a specific question is posed that can be assessed by a limited number of standardized patient cases, blinding is maintained, the patient’s presentation is standardized, the standardized patient accurately records relevant details of the visit, and the aspect of performance to be assessed can be evaluated during the first visit. The study of physician performance by Hutchison and colleagues is a good example of the use of standardized patients to evaluate a specific research question — the determinants of adherence to evidence-based preventive guidelines. It was feasible to train the standardized patients to assess physician performance, there were no major physical findings that needed to be standardized, and there was minimum risk that the problem presented would have led to invasive interventions during the visit or deferral of care of real patients. These latter issues must be considered in judging the suitability of standardized patient assessment and the ethical and legal implications of blinded practice assessment. If the physician believes that the standardized patient has an acute problem, one that requires immediate intervention, then not only is the standardized patient at risk, but resources may be inappropriately diverted from real patients. For example, in one case involving an elderly person with abdominal pain, the physician requested emergency abdominal ultrasonography, necessitating overtime pay to hospital technicians. In a second case, a physician wanted to give cortisone injections to a standardized patient presenting with chronic hip pain. The main pitfall in using standardized patients to answer questions about practice behaviour is the technical restriction that physicians with closed practices (in the study reported in this issue, 48% of the 480 physicians who responded to the initial survey) have to be excluded. When this limitation is coupled with low participation rates, estimates of the prevalence of preventive care as well as their determinants may be biased. Editorial


Citations per Year

131 Citations

Semantic Scholar estimates that this publication has 131 citations based on the available data.

See our FAQ for additional information.

Cite this paper

@article{Tamblyn1998UseOS, title={Use of standardized patients in the assessment of medical practice.}, author={Robyn Tamblyn}, journal={CMAJ : Canadian Medical Association journal = journal de l'Association medicale canadienne}, year={1998}, volume={158 2}, pages={205-7} }