Can Physician Expertise Improve Machine Learning Identification of Delirium?
For clinicians and hospitals, this work offers a practical human-in-the-loop approach to improve the detection of a commonly missed condition, though the gains are incremental over existing methods.
The paper presents a user-centered interactive machine learning framework that integrates physician-guided feature refinement with interpretable modeling to improve delirium detection, achieving better discrimination and temporal robustness compared to automated baselines on a dataset of 3,862 admissions.
Delirium is common in hospitalized patients and is often missed in routine care. We present a user-centered interactive machine learning (UC-iML) framework for delirium detection support that combines physician-guided feature refinement with interpretable modeling. Using 3,862 labeled admissions from six Toronto hospitals in the General Medicine Inpatient Initiative (GEMINI), we integrate administrative variables, laboratory results, medications, and a radiology-derived text indicator. Physicians guide feature refinement and model evaluation, and Shapley Additive exPlanations (SHAP) are used to summarize feature attribution. We evaluate standard supervised classifiers with temporally separated holdout testing and a later-phase validation cohort. Compared with automated and baseline variants, the proposed framework shows better overall discrimination and stronger temporal robustness, while the explanations highlight clinically meaningful signals. These results support UC-iML as a practical human-in-the-loop framework for clinically relevant delirium modeling.