The diuretic response of patients with congestive heart failure to establish doses of diapamide (750 mg) and furosemide (80 mg) was compared in an open, crossover study. Peak urine output occurred in the first 6 hours after administration of furosemide but somewhat later (12 to 18 hours) with diapamide. Both agents produced active diuresis and natriuresis in most patients. Comparisons of drug effect during the first days of each treatment period and analysis of the entire first treatment period indicated that urine output with furosemide was significantly greater than with diapamide. Urinary sodium excretion on the first day of treatment was not significantly greater with furosemide than with diapamide, nor were the differences significant on subsequent days. The observed differences between drugs on urinary potassium and chloride excretion were not statistically significant. The most frequently occurring adverse reaction was mild to moderate nausea, which was reported by five patients receiving diapamide and two patients receiving furosemide. Diarrhea and vomiting were also more frequent with diapamide. Diapamide would appear to serve a role between the milder thiazide diuretics and the more effective furosemide.

KEY WORDS: Metoprolol, myocardial infarction, mortality, beta-blocker. The Lopressor Intervention Trial (LIT) was a double-blind, randomized, placebo-controlled, multicentre study designed to evaluate the effect of oral metoprolol on overall mortality in patients surviving a recent acute myocardial infarction. Patient enrollment began in August 1979 and ended on 15 April 1982, with 2395 patients (1200 on placebo and 1195 on metoprolol). Hospitalized patients, 45 to 74 years of age, began therapy from 6 to 16 days after their myocardial infarction. Following a short titration period, maintenance therapy with metoprolol 100 mg b.i.d. or placebo was continued for up to 1 year. Enrollment was prematurely terminated because of a progressive and marked decline in patient accession; it was not feasible to reach the original goalof'3200 patients in a practical period of time. This target sample size was based on an anticipated 1 year placebo mortality rate of 10%, a 50% reduction in total mortality with metoprolol and premature discontinuation of study medication in no more than 15% of patients in the metoprolol group. Two primary analyses were planned: total mortality among all randomized patients at 7 and at 12 months of trial entry.After 7 months of treatment there were 54 deaths in the placebo group and 42 deaths in the metoprolol group. After 1 year there were 62 deaths in the placebo group and 65 in the metoprolol group. Thus, the 1 year placebo mortality rate of 5-2% was half that predicted at the outset. In addition, study medication was prematurely discontinued in over 30% of patients in the metoprolol group. A t termination, the power of the study was found to be too low to reliably detect the mortality reduction reported recently by other studies. Hence the results of LIT are deemed inconclusive.