Comorbidity Index Score Increases Due to Coding Artifacts
Comorbidity indices calculated from electronic health records are commonly used in research studies for various reasons, including study inclusion criteria or to adjust for overall health status. However, changes in diagnosis codes and/or billing practices may alter index values and lead to inaccurate inclusions/exclusions or covariate adjustment. While extracting data for two retrospective cohort studies at Kaiser Permanente Washington (KPWA) in two very different populations, we identified unexpected increases in comorbidity over time. We want to inform other groups who are using comorbidity indices that include chronic kidney disease and renal failure (CKD/RF) as a comorbid condition (Klabunde and Charlson-Deyo-Romano indices and their respective variants), to evaluate indices over time to ensure accurate adjustment and comparisons in longitudinal studies.