Performance of existing risk scores around heart transplantation: validation study in a 4-year cohort.

Abstract

Several risk scores exist to help identify best candidate recipients for heart transplantation (HTx). This study describes the performance of five heart failure risk scores and two post-HTx mortality risk scores in a French single-centre cohort. All patients listed for HTx through a 4-year period were included. Waiting-list risk scores [Heart Failure Survival Score (HFSS), Seattle Heart Failure Model (SHFM), Meta-Analysis Global Group in Chronic Heart Failure (MAGGIC), Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure (OPTIMIZE-HF) and Get With The Guidelines-Heart Failure (GWTG-HF)] and post-HTx scores Index for Mortality Prediction After Cardiac Transplantation (IMPACT and CARRS) were computed. Main outcomes were 1-year mortality on waiting list and after HTx. Performance was assessed using receiver operator characteristic (ROC), calibration and goodness-of-fit analyses. The cohort included 414 patients. Waiting-list mortality was 14.0%, and post-HTx mortality was 16.3% at 1-year follow-up. Heart failure risk scores had adequate discrimination regarding waiting-list mortality (ROC AUC for HFSS=0.68, SHFM=0.74, OPTIMIZE-HF=0.72, MAGGIC= 0.70 and GWTG=0.77; all P-values <0.05). On the contrary, post-HTx risk scores did not discriminate post-HTx mortality (AUC for IMPACT=0.58, and CARRS=0.48, both P-values >0.50). Subgroup analysis on patients undergoing HTx after ventricular assistance device (VAD) implantation (i.e. bridge-to-transplantation) (n=36) showed an IMPACT AUC=0.72 (P<0.001). In this single-centre cohort, existing heart failure risk scores were adequate to predict waiting-list mortality. Post-HTx mortality risk scores were not, except in the VAD subgroup.