Free triiodothyronine (fT3) and B-type natriuretic peptide (BNP) predict in-hospital mortality after valve surgery.

Abstract

Increased B-type natriuretic peptide (BNP) and decreased free triiodothyronine (fT3) are associated with increased mortality after cardiac surgery. However, previous studies have addressed mortality primarily in patients undergoing coronary artery bypass graft, and not in those undergoing valve surgery. We assessed abilities of BNP and fT3 to predict mortality after valve surgery. This retrospective study included 1050 consecutive patients who underwent valve surgery with normothermic cardiopulmonary bypass (CPB). Predictors of in-hospital mortality were identified with logistic regression analyses. Cutoff values were determined with receiver operating curve analyses. There were 23 deaths (2.2%). By univariate analyses, fT3, log-transformed BNP (log BNP), cholinesterase, estimated glomerular filtration rate, and albumin were profoundly associated with in-hospital mortality (p < 0.0001 for each). By a multivariate analysis, however, only fT3 and log BNP remained significant (p = 0.0053 and p = 0.0449, respectively). fT3 and log BNP remained significant after adjustment for CPB time, and also after adjustment for the EuroSCORE II or JapanSCORE (p < 0.05 for each). Odds ratio and 95% confidence interval from univariate binary analysis for in-hospital mortality were 7.22 (3.12-17.58) for fT3 ≤ 2.21pg/mL, 6.01 (2.54-15.77) for BNP ≥ 219pg/mL, and 9.79 (4.21-22.74) for both combined (p < 0.0001 for each). fT3 and BNP predict in-hospital mortality after valve surgery, independent of each other, independent of CPB time, and independent of established mortality risk SCOREs. Additions of fT3 and BNP may enhance predictive utilities of established mortality risk SCOREs.