A Comparison of Two Preoperative Frailty Models in Predicting Postoperative Outcomes in Geriatric General Surgical Patients

Abstract

Frailty in a surgical geriatric population may identify patients at increased risk of complications. However, the optimal method to diagnose it remains to be identified. This study aims to compare two common frailty models and assess their association with postoperative adverse outcomes in elderly patients undergoing general surgical procedures. Prospective study including 298 patients age 65years or older undergoing elective general surgical operations in a tertiary hospital. Frailty phenotype (FP) was classified using a validated scale which included weight loss, weakness, exhaustion, slowed walking speed and low physical activity. A preoperative comprehensive geriatric assessment (CGA) was performed including managing daily activities (ADL), instrumental ADL, cognitive status, comorbidities, polypharmacy and nutritional status. Main outcomes measures were postoperative complications and length of stay. There were 135 (46%), 114 (38%) and 46 (15%) minor/intermediate, major and major + procedures, respectively. The agreement between the FP and CGA was moderate (kappa index: 0.45). FP was significantly associated with postoperative complications with an odds ratio (OR) of 2.3, (95% confidence interval 1.4-3.8, p < 0.01). The association of CGA with postoperative complications did not reach statistical significance (p = 0.07). Postoperative hospital stay was significantly longer in both CGA frailty (p < 0.001) and FP (p = 0.001) groups compared to the fit population. In the multivariate analysis adjusted for ASA and POSSUM category, FP retained its significance as a predictor of postoperative complications (OR: 1.9, 95% CI 1.03-3.3, p = 0.038). FP was associated more consistently than CGA with adverse postoperative outcomes in elderly patients undergoing general surgical procedures.