Objectives: The adult comorbidity evaluation (ACE-27) is a medical record-based comorbidity index that predicts survival among various types of cancer patients. The purpose of this study was to compare the medical record-based ACE-27 instrument to a newly developed administrative claim-based ACE-27 measure.
Study design and setting: Cross-sectional study of 4,300 breast and prostate cancer patients from the Centers for Disease Control and Prevention Patterns of Care Study.
Results: Comorbidities with the highest concordance were diabetes (sensitivity=84.6%, κ=0.58 for breast cancer patients; sensitivity=0.764, κ=0.54 for prostate cancer patients), and hypertension (sensitivity=78.5%, κ=0.32 for breast cancer patients; sensitivity=69.6%, κ=0.28 for prostate cancer patients). Diseases with fair or moderate agreement in one or both cancer sites include congestive heart failure, arrhythmia, hypertension, respiratory diseases, hepatic disease, renal disease, dementia, and neuromuscular disease. For overall indices, agreement was fair but with high sensitivities in the collapsed indices, and the highest sensitivities in the lowest level of decompensation.
Conclusions: The ACE-27 comorbidity score derived from administrative claims data provides a tool to examine the relationship between comorbidity, cancer diagnosis, and outcomes in future epidemiologic research, particularly when medical record review is logistically impossible. The classification of most comorbidities into 2 or 3 levels of severity within a claim-based measure is a major development. Future research should be directed toward refining the measure with a longer review period or different paradigms for diagnosis identification, and testing the predictive ability of the measure in terms of survival, complications, or other outcomes of care.