Geriatric Assessment as a Predictor of Delirium and Other Outcomes in Elderly Patients With Cancer

Objective: This study aimed to describe the implementation of preoperative geriatric assessment (GA) in patients undergoing major cancer surgery and to determine predictors of postoperative delirium. Background: Geriatric surgical patients have unique vulnerabilities and are at increased risk of developing postoperative delirium. Methods: Geriatricians at Memorial Sloan Kettering Cancer Center risk-stratify surgical patients with solid tumors, ages 75 years and older, using preoperative GA, which includes basic and instrumental activities of daily living (ADLs, IADLs), cognition (Mini-Cog test), history of falls, nutritional state, and comorbidities (Charlson Comorbidity Index). The Geriatrics Service evaluates patients for postoperative delirium using the confusion assessment method. A retrospective review was performed. The associations between GA and postoperative outcomes were evaluated. Univariate logistic regression analysis was performed to determine the predictive value of GA for postoperative delirium, and a multivariate model was built. Results: In total, 416 patients who received preoperative evaluation by the Geriatrics Service between September 1, 2010, and December 31, 2011, were included. Delirium occurred in 19% of patients. Patients with delirium had longer length of hospital stay (P < 0.001) and greater likelihood of discharge to a rehabilitation facility (P < 0.001). Charlson Comorbidity Index score, history of falls, dependent on IADL, and abnormal Mini-Cog test results predicted postoperative delirium on univariate analysis. Developed using a stepwise selection method, a multivariate model to predict delirium is presented including Charlson Comorbidity Index score (P = 0.032), dependence IADLs (P = 0.011), and falls history (P = 0.056). Conclusions: Preoperative GA is feasible and may achieve a better understanding of older patients' perioperative risks, including delirium.