A validated score to predict one-year and long-term mortality in patients with significant tricuspid regurgitation.

Abstract

Most patients with significant (defined as ≥ moderate) tricuspid regurgitation (TR) are treated conservatively. Individual mortality rates are markedly variable. We developed a risk score based on comprehensive clinical and echocardiographic evaluation, predicting mortality on an individual patient level. The cohort included 1701 consecutive patients with significant TR, half with isolated TR, admitted to a single hospital, treated conservatively. We derived a scoring system predicting 1-year mortality and validated it using k-fold cross-validation and with external validation on another cohort of 5141 patients. Score utility was compared with matched patients without significant TR. One-year mortality rate was 31.3%. The risk score ranged 0-17 points and included 11 parameters: age (0-3), body mass index ≤ 25 (0-1), history of liver disease (0-2), history of chronic lung disease (0-2), estimated glomerular filtration rate (0-5), haemoglobin (0-2), left-ventricular ejection fraction (0-1), right-ventricular dysfunction (0-1), right atrial pressure (0-2), stroke volume index (SVI) (0-1) and left-ventricular end-diastolic diameter (0-1). One-year mortality rates increased from 0 to 100%, as the score increased up to ≥16. Areas under the receiver operating curves were 0.78, 0.70, and 0.73, for the original, external validation, and external validation with SVI measured cohorts. The score remained valid in subpopulations of patients with quantified RV function, quantified TR and isolated TR. Significant TR compared to no TR, affected 1-year mortality stronger with higher scores, with a significantly positive interaction term. We suggest a robust risk score for inpatients with significant TR, assisting risk stratification and decision-making. Our findings underscore the burden of TR providing benchmarks for clinical trial design.