Objective: To examine the relationship between resident physicians' perceptions of their preventive cardiology practices and a chart audit assessment of their documented services.
Design: A criterion standard comparison of two methods used to assess resident physicians' practices: self-report and chart audit.
Setting: Physician ambulatory care in a residency program.
Patients and other participants: Coronary artery disease (CAD) risk factor assessment was evaluated by self-report for 72 resident physicians and by chart audit of randomly selected records of 544 of their patients who did not have CAD or a debilitating chronic disease during a one-year period.
Intervention: Measurements of the residents' perceived CAD risk factor assessment practice by self-report, and chart audit assessments of their recorded care.
Main outcome: The relationship between self-reported and chart audit assessments of CAD risk factors.
Results: Chart audit assessment of CAD risk factor management was highly significantly (p < 0.01) lower than self-reported behaviors for evaluation of cigarette smoking, diet, physical activity, stress, plasma cholesterol, blood pressure, and body weight/obesity.
Conclusions: Three different interpretations of these findings are apparent. 1) Physician self-report is a poor tool for the measurement of clinical behavior, and therefore research of physician behavior should not rely solely on self-reported data; 2) physicians' chart recording of their clinical practice is insufficient to reflect actual care; or 3) neither is an accurate measure of actual practice.