Predicting in-hospital mortality of patients with febrile neutropenia using machine learning models.

•We developed machine learning models helping early identify adult febrile neutropenia inpatients with high-mortality risk.•82 non-subjective (objective) variables from administrative claims data were used for model development and evaluation.•Algorithms include linear (linear SVM, logistic regression) and non-linear (gradient boosting tree and neural network).•Unlike current scoring systems (MASCC, CISNE, qSOFA), our method does not need physician’s subjective evaluation.•Machine learning models achieved promising performances and outperform current scoring systems.