Risk Prediction Model for Survival of Wait List Patients on Axial CF-LVAD: A UNOS Database Analysis

Abstract

<h3>Purpose</h3> Prior literature shows that survival of waitlist patients is improved on axial continuous flow left ventricular assist devices (CF-LVAD). This study attempts to generate a risk prediction model for survival using the UNOS database population. <h3>Methods</h3> UNOS database was queried to extract patients (≥ 18 years of age) listed for heart transplantation between 2010 and 2015 with an axial CF-LVAD (Heartmate II(HMII)) while on the waiting list. The multivariate model was used to predict the probability of survival at 3,6 and 12 months after listing. Patients were divided into derivation (80%) and validation (20%) groups. Derivation group was used to develop the multivariate model and validation group was used for model validation. The model accuracy was assessed using Receiver Operating Characteristics (ROC) curves and Area Under Curves (AUCs). Kaplan Meier survival curves were generated to show the impact of different risk factors on waiting list survival. All the analysis was performed using MATLAB software from the MathWorks, Inc. <h3>Results</h3> Significant risk factors on multivariate analyses were diabetes type1 (HR = 2.5,p=0.018), presence of inotropes (HR = 1.6,p=0.005), creatinine at listing (HR = 1.2,p=0.00016). No significant differences were observed between the derivation and validation groups for all variables. The ROC curves generated using these risk factors showed AUC at 3, 6 and 12 months on the wait list of 0.7, 0.65, 0.63 respectively in the training set and 0.71, 0.65, 0.6 respectively in the validation set (figure 1). Survival analyses showed that patients implanted with HMII before listing had a better survival than those who did after being on the wait list (p-value <0.001, HR = 0.259) <h3>Conclusion</h3> This is the first time a risk prediction model has been generated for wait list survival of HMII patients. A significant difference in survival was noted between patients who received their HMII prior to listing versus those who had it while waiting on the list.