Updating and Validating the Rheumatic Disease Comorbidity Index to Incorporate ICD‐10‐CM Diagnostic Codes

Abstract

ObjectiveTo update and validate the Rheumatic Disease Comorbidity Index (RDCI) utilizing International Classification of Diseases, Tenth Revision, Clinical Modification (ICD‐10‐CM) codes.MethodsWe defined ICD‐9‐CM (n = 1,068) and ICD‐10‐CM (n = 1,425) era cohorts (n = 862 in both) spanning the ICD‐9‐CM to ICD‐10‐CM transition in a multicenter, prospective rheumatoid arthritis registry. Information regarding comorbidities was collected from linked administrative data over 2‐year assessment periods. An ICD‐10‐CM code list was generated from crosswalks and clinical expertise. ICD‐9– and ICD‐10–derived RDCI scores were compared using intraclass correlation coefficients (ICC). The predictive ability of the RDCI for functional status and death during follow‐up was assessed using multivariable regression models and goodness‐of‐fit statistics (Akaike's information criterion [AIC] and quasi information criterion [QIC]) in both cohorts.ResultsMean ± SD RDCI scores were 2.93 ± 1.72 in the ICD‐9‐CM cohort and 2.92 ± 1.74 in the ICD‐10‐CM cohort. RDCI scores had substantial agreement in individuals who were in both cohorts (ICC 0.71 [95% confidence interval 0.68–0.74]). Prevalence of comorbidities was similar between cohorts with absolute differences <6%. Higher RDCI scores were associated with a greater risk of death and poorer functional status during follow‐up in both cohorts. Similarly, in both cohorts, models including the RDCI score had the lowest QIC (functional status) and AIC (death) values, indicating better model performance.ConclusionThe newly proposed ICD‐10‐CM codes for the RDCI‐generated comparable RDCI scores to those derived from ICD‐9‐CM codes and are highly predictive of functional status and death. The proposed ICD‐10‐CM codes for the RDCI can be used in rheumatic disease outcomes research spanning the ICD‐10‐CM era.