FRailty in Australian patients admitted to Intensive care unit after eLective CANCER-related SURGery: a retrospective multicentre cohort study (FRAIL-CANCER-SURG study)

Ryan R Ling,Ryo Ueno,Muhammad Alamgeer,Krishnaswamy Sundararajan,Raghav Sundar,Michael Bailey,David Pilcher,Ashwin Subramaniam

DOI: https://doi.org/10.1016/j.bja.2024.01.020

Abstract:Background: The association between frailty and short-term and long-term outcomes in patients receiving elective surgery for cancer remains unclear, particularly in those admitted to the ICU. Methods: In this multicentre retrospective cohort study, we included adults ≥16 yr old admitted to 158 ICUs in Australia from January 1, 2018 to March 31, 2022 after elective surgery for cancer. We investigated the association between frailty and survival time up to 4 yr (primary outcome), adjusting for a prespecified set of covariates. We analysed how this association changed in specific subgroups (age categories [<65, 65-80, ≥80 yr], and those who survived hospitalisation), and over time by splitting the survival information at monthly intervals. Results: We included 35,848 patients (median follow-up: 18.1 months [inter-quartile range: 8.3-31.1 months], 19,979 [56.1%] male, median age 69.0 yr [inter-quartile range: 58.8-76.0 yr]). Some 3502 (9.8%) patients were frail (defined as clinical frailty scale ≥5). Frailty was associated with lower survival (hazard ratio: 1.72, 95% confidence interval [CI]: 1.59-1.86 compared with clinical frailty scale ≤4); this was concordant across several sensitivity analyses. Frailty was most strongly associated with mortality early on in follow-up, up to 10 months (hazard ratio: 1.39, 95% CI: 1.03-1.86), but this association plateaued, and its predictive capacity subsequently diminished with time up until 4 yr (1.96, 95% CI: 0.73-5.28). Frailty was associated with similar effects when stratified based on age, and in those who survived hospitalisation. Conclusions: Frailty was associated with poorer outcomes after an ICU admission after elective surgery for cancer, particularly in the short term. However, its predictive capacity with time diminished, suggesting a potential need for longitudinal reassessment to ensure appropriate prognostication in this population.