Influence of blood pressure reduction on composite cardiovascular endpoints in clinical trials.

BACKGROUND: The use of a composite cardiovascular endpoint (CCEP) is frequent in clinical trials. However, the relation between the reduction in blood pressure (BP) and the risk of CCEP is poorly known. METHODS: We conducted a meta-analysis of trials, which compared different BP-lowering agents with placebo or active treatments in patients with hypertension or composite features of high cardiovascular risk. The outcome measure was a triple (myocardial infarction, stroke and cardiovascular death) or quadruple (those mentioned above and congestive heart failure) CCEP. RESULTS: Thirty trials fulfilled the inclusion criteria, for a total of 221 024 patients. Experimental treatments reduced the risk of CCEP by 9\% (P < 0.0001). In a multivariable meta-regression analysis, for each 5-mmHg reduction in SBP, there was a 13\% less risk of CCEP (95\% confidence interval 8-19, P = 0.001) and, for each 2-mmHg reduction in DBP, there was a 12\% less risk of CCEP (95\% confidence interval 7-16, P = 0.001). Use of triple or quadruple CCEP (P = 0.150), its definition as primary or nonprimary endpoint (P = 0.305) and use of placebo or active control as comparators (P = 0.552) did not influence the estimates. A different BP reduction of at least 4.6 mmHg in SBP or at least 2.2 mmHg in DBP was required to achieve a 95\% prediction interval entirely lying below the unity. CONCLUSION: BP reduction is important to reduce the risk of CCEP in clinical trials. A significant difference between two treatment groups in the risk of CCEP may be anticipated for a SBP/DBP reduction differing by 4.6/2.2 mmHg or more.