Risk Adjustment in Outcome Assessment: the Charlson Comorbidity Index

TLDR

The study adapted the Charlson comorbidity index to ICD‑9 codes in the MED‑ECHO database to quantify comorbidity burden. The adapted index was applied to 62,456 patients with ischemic heart disease, heart failure, stroke, or pneumonia, and logistic regression—including gender, age, principal diagnosis, and various score transformations—was used to evaluate its predictive accuracy for inpatient death. The Charlson index was strongly associated with death, achieving an AUC of 0.83, demonstrating its usefulness for risk adjustment in administrative database research.

Abstract

To measure the burden of comorbid diseases using the MED-ECHO database (Quebec), the so-called Charlson index was adapted to International Classification of Disease (ICD-9) codes. The resulting comorbidity index was applied to the study of inpatient death in a group of 62,456 patients having one of the following conditions: ischemic heart disease, congestive heart failure, stroke, or bacterial pneumonia. Multiple logistic regression was used to relate inpatient death to its predictors, including gender, principal diagnosis, age, and the comorbidity index. Various transformations of the comorbidity score were performed, and their effect on predictive accuracy was assessed. The comorbidity index was constantly and strongly associated with death. When gender, age, comorbidity and the principal diagnoses were taken into account, the area under the receiver-operating curve was 0.83. Therefore, the Charlson Index is a useful approach to risk adjustment in outcomes research from administrative databases.