Using Charlson Comorbidity Index to Predict Short-Term Clinical Outcomes in Hospitalized Older Adults

This study examined the prognostic value of the Charlson Comorbidity Index (CCI) in predicting short-term clinical outcomes in hospitalized older adults. We conducted a retrospective cohort study of patients, older than 75 years, admitted to the medicine service at a large tertiary hospital (New York). We used the Enhanced International Classification of Disease, 9th Revision, Clinical Modification adaptation to abstract the CCI from electronic medical records. The CCI scores were compared, using the standard Deyo version and the Schneeweiss version. Outcome measures included in-hospital mortality, length of stay (LOS), and 30-day readmissions. When comparing Charlson/Deyo and Charlson/Deyo/Schneeweiss with and without age, we found similar significant association with regard to in-hospital mortality, with a moderate predictive ability (area under the curve [AUC]: 0.5906-0.6433). However, for 30-day readmissions and LOS, the predictive ability was poor (AUC: 0.5598-0.6106 and rho: 0.11-0.12, respectively). The CCI is, at most, a moderate predictor of in-hospital mortality and a poor predictor of other important healthcare outcomes relevant to administrative healthcare practices.