RESUMO
AIM: To determine the clinical risk factors predictive of the 5--year mortality in patients with low cardiac output syndrome (LCOS) after cardiac surgery. In addition, to assess the influence of inflammation and myocardial dysfunction severity, as measured by C--reactive protein (CRP) and N--terminal pro--brain natriuretic peptide (NT--proBNP) concentrations, on outcome. METHODS: We studied 30 patients who underwent cardiac surgery and developed postoperative LCOS requiring inotropic support for longer than 48 hours after intensive care unit (ICU) admission. All patients received a 24--hour infusion of levosimendan after study enrolment. We measured the following at baseline, 24 h, 48 h and 7 days: clinical data, serum NT--proBNP and serum CRP levels. Patients were followed--up at 5 years for death by any cause. A risk--adjusted Cox proportional hazards regression model was used for statistical analysis. Hazard ratios and their 95% confidence intervals (CI) are presented. RESULTS: The 5--year mortality was 36.6% (n = 11). The predictors of 5--year mortality were the presence of dilated cardiomyopathy (HR = 36.909; 95% CI: 1.901-716.747; P = 0.017), a higher central venous pressure (CVP) at 48 hours (HR = 2.686; 95% CI: 1.383-5.214; P = 0.004), and lower CRP levels on day 7 (HR = 0.963; 95% CI: 0.933-0.994; P = 0.021). NT--proBNP levels showed a trend to higher initial levels in survivors without statistical significance, but were not associated with 5--year mortality. CONCLUSIONS: The presence of dilated cardiomyopathy, elevated CVP at 48 h and reduced CRP levels on day 7 predicted 5--year mortality in patients who developed postoperative LCOS after cardiac surgery. NT--proBNP levels in the first postoperative week were not predictors of long--term outcomes.
