This is a quantitative risk score model for recipient risk stratification of patients on ECMO bridge to lung transplant. The calculator is based on data from the United Network for Organ Sharing (UNOS) since 2004 and comprises 669 adult patients awaiting lung transplantation since 2004.
All recipient specific pre-transplant variables with plausibility for predicting in-hospital mortality were tested using univariate regression. These variables included: Age, gender, BMI, recipient functional status, days on waitlist, transplant center size, mechanical ventilation, liver function (total bilirubin), kidney function (creatinine and need for dialysis) and major diagnosis codes (CODP/ emphysema, idiopathic pulmonary fibrosis, pulmonary hypertension). All potential explanatory variables were assessed for collinearity. Continuous variables were explored for linearity by considering as quartiles before being converted into categorical variables with 2 degrees of freedom for practical purposes. Those associated with in-hospital mortality on exploratory analysis (p ≤ 0.2) were incorporated into a multivariable logistic regression model. Final model was constructed using automated stepwise backward regression and contained factors that improved explanatory power. Linear prediction was used to construct the recipient STratification risk Analysis in Bridging patients to Lung transplant on ECMO (STABLE) score. K-fold and leave-one-out cross-validation were used to provide an unbiased estimate of out-of-sample performance of the STABLE score.
This project was approved by United Network for Organ Sharing (UNOS).
Data presented at the the 29th Annual Meeting of the International Society for Heart and Lung Transplantation, Orlando, FL.