Abstract P358: Do Cardiac Interventions Affect Noncardiac Mortality?

Abstract

Listen

Translate article

Clinical trials of cardiovascular drugs and devices primarily assess benefit and risk in terms of cardiovascular mortality and other major adverse cardiovascular events. There is little data on how noncardiac mortality is impacted in clinical trials. In this study, we sought to determine how noncardiac mortality was affected in clinical trials that were published in the past 12 years in major general medicine journals. We calculated the standardized mortality ratio and its confidence intervals for cardiac, noncardiac and total mortality in all arms of 93 large clinical trials. Relative change and power to detect significant changes in all three components was calculated. Twenty two studies had a decrease in cardiovascular mortality of at least 20%, while only 4 studies had an increase in cardiovascular mortality more than 20%. Only 7 studies had a statistically significant decrease in cardiovascular mortality. Noncardiac mortality was at least 5% higher in the treatment arm in 31 (33%) of all studies. In 4 of these studies, noncardiac mortality was at least 20% higher in the treatment arm. Noncardiac mortality was at least 5% lower in the treatment arm in 24 studies, with 3 of then having at least a 20% lower mortality rate. In only 1 study was there a significant difference in noncardiac mortality. Only 3 studies were adequately powered to detect a 20% difference in noncardiac mortality. None were powered to detect a 10% difference in noncardiac mortality. Overall, there was no difference in noncardiac mortality in these studies. There is wide variation in differences in noncardiac mortality determined in clinical trials, with most studies showing at least a 5% difference in noncardiac mortality rates and over 1/3 of the studies demonstrating at least a 10% difference in noncardiac mortality. In primary prevention and hypertension studies, one would only need a 6-10% increase in noncardiac mortality to negate a 20% decrease in cardiac mortality. No studies were adequately powered to evaluate for this effect. With the strong push to evidence based medicine, studies need to be adequately powered to demonstrate a lack of harm as well as benefit. The increased use of multiple endpoints will exacerbate the problem of assessment of risk in cardiovascular clinical trials.