Comparing the performance of comorbidity indices in predicting functional status, health-related quality of life, and total health care use in older adults with back pain

• OBJECTIVE: To determine how well the functional comorbidity index (FCI) predicts outcomes in older adults with back pain compared to Quan's modification of the Charlson comorbidity index (Quan-Charlson comorbidity index) and the Elixhauser comorbidity index. • DESIGN: Secondary analysis of a prospective cohort study. • METHODS: We included 5155 adults 65 years of age or older with new primary care visits for back pain. Comorbidity was measured using diagnosis codes 12 months prior to the new visit. Outcomes of functional limitation (Roland-Morris Disability Questionnaire), health-related quality of life (European Quality of Life-5 Dimensions [EQ-5D]), and total health care use (sum of relative value units) were measured 12 months after the new visit. We compared multivariable models containing preselected prognostic factors. • RESULTS: Spearman correlation coefficients among the indices were 0.70 or greater. Multivariable models for the Roland-Morris Disability Questionnaire had similar R² and root-mean-square error (RMSE) of prediction when using the FCI (R² = 0.190; RMSE, 6.19), Quan-Charlson comorbidity index (R² = 0.185; RMSE, 6.20), or Elixhauser comorbidity index (R² = 0.189; RMSE, 6.19). Multivariable models for the EQ-5D score showed small differences in R² and RMSE when using the FCI (R² = 0.157; RMSE, 0.163), Quan-Charlson comorbidity index (R² = 0.148; RMSE, 0.164), or Elixhauser comorbidity index (R² = 0.154; RMSE, 0.163). Multivariable models for health care use had similar Akaike information criterion (AIC) values when using the FCI (AIC = 10.04), Quan-Charlson comorbidity index (AIC = 10.04), or Elixhauser comorbidity index (AIC = 10.01). • CONCLUSION: All indices performed similarly in predicting outcomes. There does not seem to be an advantage to using one index over another for older adults with back pain. There is still a need to develop better function-based risk-adjustment models that improve prediction of functional outcomes versus standard comorbidity indices.