Profiles of the trial and the sample size calculation

The principle outcome measures of this study were the current prescription of b blockers and lipid lowering drugs, and the measurement of serum cholesterol after discharge. On the basis of data from local audits in 1995, we calculated that to have an 80% chance of detecting a 20% absolute increase in the prescription rate of b blockers in patients without contraindications (from 40% to 60% of all patients discharged after an acute coronary event), significant at the 5% level, we needed a sample size of 220 patients (110 in each group). This sample size would allow us to detect clinically important changes in the two other principle outcomes. Allowance for intracluster correlation, anticipated contraindications to b blockers, death less than 6 months after discharge, or loss of patients to follow up gave us a recruitment target of 350 patients.