The Effect of Physiciansʹ Training on Prescribing β-Blockers for Secondary Prevention of Myocardial Infarction in the Elderly

PURPOSE: The objective was to determine whether the year and medical school of graduation, the medical specialty and the sex of the treating physician was associated with prescribing β-blockers after hospital discharge among survivors of myocardial infarction (MI), after adjusting for patient characteristics. METHODS: We carried out a dynamic retrospective cohort study using data from a longitudinal database that contained information on Quebec residents over the age of 65 years sent home from hospital after MI between 1990 and 1993. The outcome was a β-blocker being dispensed after hospital discharge. Logistic regression was used to estimate the association between training characteristics and β-blocker dispensation and clustering of patients within physicians was accounted for using Generalized Estimating Equations. RESULTS: The cohort consisted of 14,334 MI survivors who were treated by 3209 physicians, yielding a mean of about 4.5 patients per physician [standard deviation (SD) = 8.2]. β-blockers were prescribed to approximately one-third of subjects. After adjusting for patientsʹ demographic characteristics, comorbid medical conditions, and markers of MI severity, physicians who were more likely to prescribe a β-blocker included cardiologists and the most recent graduates (graduating after 1989). Systematic differences were also observed between graduates of different medical schools. CONCLUSIONS: After adjusting for differences between patientsʹ, the sex of the physician was largely unrelated to prescribing β-blockers for secondary prevention of MI. However, prescribing differed by training characteristics such as medical specialty and year and medical school of graduation.