patient-stratification

This paper proposes a Machine-Learned Comorbidity Index (MLCI) that uses diagnosis codes and nonlinear learning to improve risk adjustment across multiple clinical outcomes, outperforming traditional mortality-centric indices.