Effects of perioperative statin use on cardiovascular complications in patients submitted to non-cardiac surgery: protocol for a systematic review, meta-analysis, and trial sequential analysis

Erica Aranha Suzumura, Rodrigo Antonini Ribeiro, Leticia Kawano-Dourado, Pedro Gabriel de Barros e Silva, Claudia Oliveira, Mabel Fernandes Figueiró, Alexandre Biasi Cavalcanti, Renato D. Lopes, Otavio Berwanger

Preliminary evidence suggests statins may reduce major perioperative vascular events. However, evidence is limited to observational studies, underpowered trials, and non-comprehensive systematic reviews. This review aims to assess the effects of perioperative statin use on cardiovascular complications in patients submitted to non-cardiac surgery. We will search MEDLINE/PubMed, EMBASE, LILACS, CENTRAL, Web of Science, and CINAHL for randomized controlled trials assessing the effects of perioperative statin use in adults undergoing non-cardiac surgery and reporting cardiovascular complications. For patients already using statins for hyperlipidemia, a preoperative loading dose of statin is required in the experimental group. We will place no language or publication restriction on our search. Teams of two reviewers will independently assess eligibility and risk of bias, and will extract data from the included trials. Our primary outcome is a combination of cardiovascular mortality or non-fatal myocardial infarction. We will also assess the following outcomes: individual components of the primary outcome, all-cause mortality, total myocardial infarction, elevated troponin in the first seven postoperative days, total stroke, total venous thromboembolism, postoperative atrial fibrillation, elevation of creatine phosphokinase or liver enzymes, and rates of myalgia or rhabdomyolysis. We will conduct meta-analyses using random-effects model and will use trial sequential analysis to establish monitoring boundaries to limit global type I error due to repetitive testing for our primary outcome. We will rate the quality of evidence using the GRADE system. The results of this systematic review may help to inform clinical practice and also the design of future large-scale randomized trials. PROSPERO CRD42016035987.