Abstract
Introduction: In meta-analyses, odds ratios (OR) are considered more meaningful to summarize data of multiple trials while absolute risk reduction (ARR) is more relevant to specific clinical questions. We developed a novel bivariate random effects meta-analysis method for simultaneous evaluation of ORs and ARRs. Methods: Using this method, we analyzed the efficacy of different classes of antihypertensive therapy in preventing heart failure (HF) in 54 randomized clinical trials not requiring diabetes as an inclusion criterion (331,728 patients, 12,969 events), including SPRINT and HOPE-3. Results: The expected relationship between OR and ARR for a given baseline risk was observed. Diuretics were superior to all other medications, both with respect to ARR and OR (ARR 14.6 per thousand, 95% CI 7.3-22.0; OR 0.52, CI 0.38-0.70). Intensive therapy (ARR 5.1, CI -1.9-12.1; OR 0.72, CI 0.46-1.11) was the second best, followed by ACEIs (ARR 4.9, CI 0.1-9.6; OR 0.81, CI 0.69-0.95), ARBs (ARR 4.4, CI -0.3-8.5; OR 0.90, CI 0.81-0.99), and DHPs (ARR 1.0, CI -3.4-5.4; OR 0.89, CI 0.86-1.16). Verapamil was associated with a trend toward increased absolute risk (ARR 2.3, CI -0.1-4.8; OR 1.17, CI 0.99-1.37). Multiple metaregression identified the magnitude of the difference between the blood pressure lowering from baseline in the active and control groups as the only significant association (p<0.04) of the benefit in preventing HF. The effects on all-cause and cardiovascular mortality were consonant with the above with the exception of verapamil, where the adverse effect was attenuated when cardiovascular and all-cause death were considered. Conclusions: This new bivariate meta-analysis method combining ARR and OR is useful in clinical decision making, using ARR for specific clinical questions, OR to summarize studies. Based on information from all available studies, diuretics are the superior medication for the prevention of HF in hypertensive patients from both points of view.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have