Development and application of a frailty index derived from patient-reported outcome measures and comorbidities in oncology patients.

BACKGROUND: Emerging evidence supports the use of a frailty index as an objective measure of frailty. We constructed a frailty index from patient-derived outcome measures, as well as self-reported comorbidity data, and tested its correlation with survival. METHOD: Data from 464 patients with lung, breast and colorectal cancer were obtained from the Continuous Improvement in Care - Cancer Project database, and a 42-variable frailty index was generated by dividing the total number of deficits present by the total number assessed and expressed as a ratio between 0 and 1. Frailty was defined as a frailty index of >0.2. Its association with survival was assessed using Cox proportional hazards models. RESULTS: Frailty index scores ranged from 0.00 to 0.58 (mean = 0.16, standard deviation = 0.12) and approximated a gamma distribution. Each year of patient age was associated with a small but significant increase in frailty index scores (0.001 points, P = 0.035). Among patients with lung or colorectal cancer, women had significantly higher scores than men (mean difference = 0.06, P = 0.003). Every 0.1 increase in the frailty index significantly increased the risk of death. The probability of mortality, after adjusting for age, sex and cancer type, was higher in the frail cohort compared to the non-frail cohort (hazard ratio 1.96, P = 0.027). CONCLUSION: Our data suggest that a self-reported patient-reported outcome measure and comorbidity-derived frailty index constitute a feasible measure of frailty. This offers practical advantages especially in oncology and geriatric care, where time and resources are limited and patient-centric care is critical.