Effects of blood pressure-lowering on outcome incidence in hypertension: 5. Head-to-head comparisons of various classes of antihypertensive drugs - overview and meta-analyses

摘要

Background and objectives:We have recently published an overview and meta-analysis of the effects of the five major classes of blood pressure-lowering drugs on cardiovascular outcomes when compared with placebo. However, possible differences in effectiveness of the various classes can correctly be estimated only by head-to-head comparisons of different classes of agents. This has been the objective of a new survey and meta-analysis.Methods:A database search between 1966 and August 2014 identified 50 eligible randomized controlled trials for 58 two-drug comparisons (247006 patients for 1029768 patient-years). Risk ratios and their 95% confidence intervals of seven outcomes were estimated by a random-effects model.Results:The effects of all drug classes are not significantly different on most outcomes when their blood pressure effect is equivalent. However, there are also significant differences involving almost all classes of drugs. When compared to all other classes together, diuretics are superior in preventing heart failure; beta-blockers less effective in preventing stroke; calcium antagonists superior in preventing stroke and all-cause death, but inferior in preventing heart failure; angiotensin-converting enzyme inhibitors more effective in preventing coronary heart disease and less in preventing stroke; angiotensin receptor blockers inferior in preventing coronary heart disease; and renin-angiotensin system blockers more effective in preventing heart failure. When stratifying randomized controlled trials according to total cardiovascular risk, no drug class was found to change in effectiveness with the level of risk.Conclusions:The results of all available evidence from head-to-head drug class comparisons do not allow the formulation of a fixed paradigm of drug choice valuable for all hypertensive patients, but the differences found may suggest specific choices in specific conditions, or preferable combinations of drugs.