Postoperative Cardiac Troponin I Thresholds Associated With 1-Year Cardiac Mortality After Adult Cardiac Surgery: An Attempt to Link Risk Stratification With Management Stratification in an Observational Study

Abstract

Objectives: Cardiac troponin (cTn) concentrations are measured routinely in some centers after cardiac surgery as part of risk stratification, but there are no data on how increased cTn concentrations could change patients' management. The aim of this study was to estimate relevant cTnI thresholds and identify potential interventions (additional monitoring/therapeutic interventions) that could be part of management changes of patients with cTnI greater than relevant thresholds.

Design: Retrospective, single-center, observational study.

Setting: Bichat-Claude Bernard Hospital, Paris, France, between January 1, 2009, and December 31, 2012.

Participants: Consecutive adult patients undergoing cardiac surgery.

Measurements and main results: cTnI was measured on the 20th postoperative hour. Causes of death and possible interventions were determined by analysis of individual medical records. cTnI thresholds for 1-year cardiac mortality with a specificity >80% were calculated. For this study, 3,228 procedures were analyzed; 129 deaths occurred (4%), 83 of which (2.6%) were cardiac deaths. Threshold cTnI values were 4.2 µg/L for coronary artery bypass grafting (95% confidence interval [CI] 3.9-4.5) and 10.7 µg/L for non-coronary artery bypass grafting (95% CI 10.0-11.3). In multivariable analysis, the EuroSCORE II (odds ratio 1.1 [95% CI 1.06-1.13]; p < 0.001) and cTnI concentrations greater than the thresholds (odds ratio 5.62 [95% CI 3.37-9.37]; p < 0.001) were associated with significantly increased risk of death. The additive and absolute Net Reclassification Index were 0.288% and 14.1%, respectively, for a logistic model including cTnI and EuroSCORE II (area under the curve C-index 0.82 [95% CI 0.77-0.87]) compared with a model including only EuroSCORE II (area under the curve C-index 0.80 [95% CI 0.75-0.84]). Fifty-three of the 83 patients who experienced cardiac death (64%) had a cTnI concentration greater than the threshold, and an intervention was deemed possible in 47 of those 53 (89%) (mostly patients with mild postoperative cardiac dysfunction). For noncardiac deaths, 28% of patients had a cTnI concentration greater than the threshold and no interventions were deemed possible.

Conclusions: In an attempt to evolve from risk to management stratification, this study's results identified a subgroup of patients with mild cardiac dysfunction and a cTnI concentration greater than the threshold who could be the target for interventions in future validation studies concerning changes in patient management.

Keywords: Cardiac Surgery; Cardiac Troponin I; Mortality; Risk Stratification.