Gensheimer MF, et al. - In order to train a prognostic model for patients with metastatic cancer, experts evaluated the electronic medical record dataset. They trained and tested the model using 12,588 patients treated for metastatic cancer in the Stanford Health Care system from 2008 to 2017. High predictive performance was demonstrated by the model, which will need to be validated using external data. Since the model is fully automated, it can be used to examine providers’ practice patterns and, in order to help improve the quality of care, this model could be deployed in a decision support tool.