A Machine Learning Model Approach to Risk-Stratify Gastrointestinal (GI) Cancer Patients for Hospitalization and Mortality Outcomes.

Abstract

PURPOSE\nGastrointestinal (GI) cancer patients frequently experience unplanned hospitalizations, but predictive tools to identify high-risk patients are lacking. We developed a machine learning model to identify high-risk patients.\n\n\nMETHODS\n1,341 consecutive patients undergoing GI (abdominal or pelvic) radiation treatment (RT) from March 2016-July 2018 (derivation) and July 2018- January 2019 (validation) were assessed for unplanned hospitalizations within 30 days of finishing RT. In the derivation cohort of 663 abdominal and 427 pelvic RT patients, a machine learning approach derived random forest, gradient boosted decision tree, and logistic regression models to predict 30-day unplanned hospitalizations. Model performance was assessed using area under the receiver operating characteristic curve (AUC) and prospectively validated in 161 abdominal and 90 pelvic RT patients using Mann-Whitney rank-sum test. Highest quintile of risk for hospitalization was defined as high-risk and the remainder low-risk . Hospitalizations for high- vs. low-risk patients were compared using Pearson s Chi-squared test and survival using Kaplan-Meier log-rank test.\n\n\nRESULTS\nOverall, 13% and 11% of patients receiving abdominal and pelvic RT experienced 30-day unplanned hospitalization. In the derivation phase, gradient boosted decision tree cross-validation yielded AUC=0.823 (abdominal patients) and random forest yielded AUC=0.776 (pelvic patients). In the validation phase, these models yielded AUC=0.749 and 0.764, respectively (P<0.001 and P=0.002). Validation models discriminated high- vs. low-risk patients: in abdominal RT patients, frequency of hospitalization was 39% vs. 9% in high- vs. low-risk groups (P<0.001) and 6-month survival was 67% vs. 92% (P=0.001). In pelvic RT patients, frequency of hospitalization was 33% vs. 8% (P=0.002) and survival was 86% vs. 92% (P=0.15) in high- vs. low-risk patients.\n\n\nCONCLUSIONS\nIn GI cancer patients undergoing RT as part of multimodality treatment, machine learning models for 30-day unplanned hospitalization discriminated high- vs. low-risk patients. Future applications will test utility of models to prompt interventions to decrease hospitalizations and adverse outcomes.