Prognostic value of troponins in sepsis: a meta-analysis

The role of biomarkers such as troponin in risk stratification of sepsis is still debated. The aim of this meta-analysis is to assess the relation between troponin elevation in sepsis and mortality. All observational studies from Embase, Medline and those manually searched up to September 2010 were included. Studies identified were those which reported on patients with a diagnosis of sepsis and if a 2 x 2 table could be constructed based on troponins and death. We pooled the relative risk (RR) and odds-ratio (OR) using the inverse variance method in studies that conducted univariate and multivariable (adjusted) analysis. Thirteen studies encompassing 1,227 patients were included. The prevalence of elevated troponin was 61 % ([95 %] CI 58-64 %). Elevated troponin was significantly associated with all-cause mortality (RR 1.91; CI 1.63-2.24), with homogeneity across studies. In adjusted analysis (four studies comprising 791 patients) according to prognostic scores, elevated troponin was associated with an increased risk of death (OR 1.92; CI 1.35-2.74). The area under the ROC curve was 0.68 (CI 0.63-0.71). Pooled sensitivity and specificity were 77 % (CI 61-88) and 47 % (CI 30-64) with heterogeneity across studies. It corresponded to positive and negative likelihood ratios of 1.50 (95 % CI: 1.20-1.90) and 0.49 (CI 0.38-0.64), respectively. Elevated troponin identifies a subset of patients with sepsis at higher risk of death. Further studies are needed to define the precise role of troponins and their optimal cut-offs.