Machine learning prediction models and nomogram to predict the risk of in-hospital death for severe DKA: A clinical study based on MIMIC-IV, eICU databases, and a college hospital ICU.

•The database used the largest existing ICU patient database with precise and reliable information.•Both traditional regression nomograms and machine learning （decision tree models） are used.•The model is validated internally and externally, showing excellent results.•The model has been used in clinical practice, the patients have been preliminarily evaluated, and the model effect is excellent.