Study objective: We compared three approaches for improving compliance with the practice guidelines of the National Cholesterol Education Program (NCEP).
Design: A randomized controlled trial.
Setting: Academic group practices of a major urban teaching hospital.
Participants: Study physicians were three equivalent groups of PG-2 and PG-3 residents (N = 33) seeing patients in equivalent outpatient clinics. Continuity patients of these residents were included (N = 240) if they were younger than 66 years, saw their primary physician during the intervention period, were not pregnant, and had no serious life-shortening noncardiac illnesses.
Interventions: Three interventions were implemented over a 5-week period. Control group physicians (group 1) were offered only a standard lecture provided through the Physician Cholesterol Education Program (PCEP). Group 2 physicians were offered the PCEP lecture and also received generic chart reminders of the NCEP guidelines on each eligible patient's chart. Group 3 physicians were offered the PCEP lecture and also received timely patient-specific feedback, including acknowledgement of recent lipid values and management, and explicit recommendations for further action. Knowledge of lipid disorders was tested before and after the PCEP lecture, and physicians' attitudes were surveyed following the intervention period.
Measurements and main results: The three groups were similar in baseline (preintervention) compliance with NCEP recommendations (average, 39%) and physicians' knowledge. Patients were similar across groups in number of coronary artery disease risk factors and cholesterol values. Significant within-group improvements in compliance were noted for groups 2 and 3 (7.6% and 10.6%, respectively), but not for group 1 (4.5%). Importantly, there were no differences observed in improvements across groups. In exploratory analyses, however, there was a significant correlation between improved compliance and the number of patients seen by each provider in group 3 that was not observed in groups 1 or 2. Notably, changes in compliance were unrelated to PCEP lecture attendance (8.6% vs 8.1% for attenders vs nonattenders, respectively), level of postgraduate training, baseline or later tests of knowledge, or patient factors. The postintervention survey revealed marked overestimation by physicians of their personal compliance with NCEP guidelines, although there was strong support for clinic efforts that would screen patients for lipid disorders independent of physician initiative.
Conclusions: This study raises questions about the effectiveness of education alone for improving compliance with NCEP guidelines. The effectiveness and efficiency of timely, individualized feedback should be explored in studies over a longer period. Innovative alternative approaches are suggested by the responses to our survey and other research in preventive practices.