Optimization of electronic clinical decision support equipment based on hospital demographics

摘要

The electronic medical decision support (CDS) device (10) utilizes a trained CDS algorithm (30) that is applied to the covariate values of the covariate set and outputs a medical state prediction. The CDS algorithm is trained on a training data set (22). The CDS device includes a computer (12) that is programmed to provide a user interface (62) for completing a clinical research question using a display and one or more user input devices. A marginal probability distribution of covariates in the covariate set is generated from completed clinical survey questions. The trained CDS algorithm is adjusted for covariate shifts using marginal probability distributions. A trained CDS algorithm adjusted for the covariate shift (50) that operates on the values for the covariate medical objects in the covariate set is used to generate a medical state prediction of the medical object.