A novel risk score that incorporates recipient and donor variables to predict 1-year mortality in the current era of lung transplantation

Background In this study we sought to construct a novel scoring system to pre-operatively stratify a patient's risk of 1-year mortality after lung transplantation (LTx) based on recipient- and donor-specific characteristics. Methods The UNOS database was queried for adult (≥18 years) patients undergoing LTx between May 1, 2005 and December 31, 2012. The population was randomly divided in a 4:1 fashion into derivation and validation cohorts. A multivariable logistic regression model for 1-year mortality was constructed within the derivation cohort. Points were then assigned to independent predictors (p < 0.05) based on relative odds ratios. Risk groups were established based on score ranges. Results During the study period, 9,185 patients underwent LTx and the 1-year mortality was 18.0% (n = 1,654). There was a similar distribution of variables between the derivation (n = 7,336) and validation (n = 1,849) cohorts. Of the 14 covariates included in the final model, 9 were ultimately allotted point values (maximum score = 70). The model exhibited good predictive strength (c = 0.65) in the derivation cohort and demonstrated a strong correlation between the observed and expected rates of 1-year mortality in the validation cohort (r = 0.87). The low-risk (score 0 to 11), intermediate-risk (score 12 to 21) and high-risk (score ≥22) groups had a 10.8%, 17.1% and 32.0% risk of mortality (p < 0.001), respectively. Conclusions This is the first scoring system that incorporates both recipient- and donor-related factors to predict 1-year mortality after LTx. Its use could assist providers in the identification of patients at highest risk for poor post-transplant outcomes.