Electronic Frailty Index: A risk stratification tool for survival and health care outcomes in veterans with colorectal cancer.

375 Background: Frailty reflects decreased physiologic reserve and subsequent increased vulnerability to stress. Frailty has been associated with poor health outcomes in patients with cancer, however traditional frailty indices are time consuming and not widely implemented in clinical oncology practice. An electronic frailty index (EFI) derived from electronic medical records (EMR) is an objective and standardized approach to frailty that can be automated. Little is known about the association between EFI and survival and healthcare utilization outcomes in patients with colorectal cancer (CRC). Methods: We used VA administrative files from 2016 – 2020 and the VA Central Cancer Registry to identify a cohort of patients diagnosed with CRC. EFI was calculated based on the Veterans Affairs-FI (VA-FI), a validated 31-item cumulative deficit FI, to define three groups: robust (≤0.1), prefrail (0.1-0.2), and frail (> 0.2). We conducted Cox proportional hazard analyses to evaluate survival. We performed logistic regression to examine prolonged length of stay (LOS) ≥ median (11 days) during all-cause hospitalization and emergency department (ED) visits within one year after diagnosis. All models were adjusted for age, gender, race, Charlson comorbidity index, and stage. Confidence intervals (CI) were calculated as 95%. Results: Of 6,043 CRC patients (age: 69.7 ± 10.6 years), 45.6% were robust, 34.5% were prefrail, and 19.9% were frail. The cohort included 22.9% rectal cancer and 77.1% colon cancer cases, distributed across stages I (29.2%), II (25.0%), III (25.2%), and IV (20.6%). Increased risk of death was found in prefrail (adjusted hazard ratio [aHR] 1.21; CI 1.10-1.32) and frail (aHR 1.90; CI 1.71-2.11) patients when compared to robust patients. Increased odds of one-year ED visits were identified in prefrail (adjusted odds ratio [aOR] 1.18; CI 1.06 – 1.35) and frail (aOR, 1.66, CI 1.43-1.94) patients. Frailty status was associated with increased odds of one-year hospitalization of prolonged LOS among prefrail (aOR 1.38; CI 1.11-1.70) and frail patients (aOR 1.79; 95% CI 1.39-2.30). Conclusions: EFI was significantly associated with survival and healthcare utilization among patients with CRC, independent of stage. Frail patients had nearly 80% increased likelihood of prolonged hospitalization. EFI has the potential to be an automated and objective decision support tool at the point of care for risk assessment prior to CRC treatment-related stressors, such as surgery or chemotherapy. Future work is needed to develop a cancer specific EFI with dynamic variables that can be followed prospectively, as well as evaluating barriers to implementation within the EMR to guide treatment decisions and improve the quality of cancer care.