IJCS | Volume 31, Nº3, May/ June 2018

291 Martucheli et al. Cystatin C and acute coronary syndromes International Journal of Cardiovascular Sciences. 2018;31(3)290-307 Review Article studies. In light of this, this systematic review and meta- analysis aimed to assess the prognostic value of cystatin C in patients with ACS. Methods We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) recommendations. 12 Search strategy An electronic search was conducted in Medline via PubMed, Web of Science and Scielo databases. Descriptors were determined using the Medical Subject Headings (MeSH) for the search in PubMed and Web of Science, and the Health Sciences Descriptors for Scielo database. The search was conducted until 30 May, 2016. The search strategy in Pubmed and Web of Science included the term “cystatin C” and its variations, combined with all variations of the term “acute coronary syndrome”, using the connector word “AND”. The search strategy in Scielo included the term “cystatin C” combined with all variations of the term “acute coronary syndrome”, using the connector word “AND”. Eligibility criteria Articles written in English, Portuguese or Spanish that met these eligibility criteria were included: • Study design: observational cohort studies. • Studypopulation: patientswithACS - unstable angina, ST-segment elevation myocardial infarction (STEMI) and non-ST-segment elevationmyocardial infarction (NSTEMI) - with increased baseline cystatin C levels. • Exposure: increased cystatin C levels. • Clinical outcome: cardiovascular events or mortality evaluated by odds ratio/relative risk and/or differences between the proportions of patients with higher and lower levels of cystatin C. The following events were considered cardiovascular events: acute myocardial infarction, need for revascularization, stroke, recurrent angina, unstable angina, heart failure and cardiovascular death. Article selection After exclusion of duplicate articles, articles published until 30 May 2016 that met the eligibility criteria were selected. The articles were selected by two independent investigators in two steps: in the first step, analysis of the title and abstracts was performed; in the second step, the articles selected in the previous step were read in full. Data extraction from the articles The following information was extracted from each article: type of ACS, diagnostic method for ACS, number of patients, patients’ age range; time of follow-up, outcome measures, method for cystatin Cmeasurement, patients’ kidney function (normal or not), GFR or serum creatinine, patients’ classification by cystatin levels, variables included in the multivariate analysis, results (frequency of cardiovascular events, cardiovascular death or all-cause mortality and/or odds ratio). Evaluation of the methodological quality of the articles Methodological quality of the articles included in the systematic review was assessed by two reviewers. The Newcastle-Ottawa Scale (NOS) 13 for cohort studies was used, which also included the following evaluation categories - cohort selection, comparability of cohorts and outcome. A maximum of one star can be attributed to the categories selection of the cohorts and outcome, a maximumof two stars can be attributed to comparability of the cohorts, such that quality of the studies can be awarded up to nine stars. Articles awarded 5 or 6 stars were considered of good methodological quality, and those awarded 7 stars were considered of excellent methodological quality. Meta-analysis In this meta-analysis, we included only studies that analyzed similar outcomes, studies that compared the fourth quartile with the first quartile of cystatin C, and studies that performed multivariate analysis (which included, among other variables, GFR or serum creatinine). Odds ratio and 95% confidence interval adjusted by multivariate analysis and heterogeneity between studies were analyzed by the I2 test. The studies were considered homogeneous when I2 was greater than 50% and p-value was lower than 0.10. Odds ratio was calculated using the fixed or the random effect model in case of homogeneity or heterogeneity, respectively. The Comprehensive Meta-Analysis (CMA) software version 3 was used for statistical analysis.