Frailty – a comparison of diagnostic criteria in older people undergoing chemotherapy

Abstract

medical record. GA captures functional status: activities of daily living (ADLs) and instrumental activities of daily living (IADLs), cognition (Mini-Cog Test), fall history, medication list, nutritional status (weight loss, albumin) and comorbidities (Charlson comorbidity index, CCI). Pts are medically optimized and risk-stratified for surgery. After their surgery, pts were evaluated in the hospital by the Geriatrics Service and evaluated for delirium using the Confusion Assessment Method (CAM). Utilizing Fisher-Exact tests/Wilcoxon-Rank Sum tests, the associations between GA measures and post-operative outcomes were evaluated. Univariate logistic regressions are also performed to determine the prediction value of the preoperative GA for the outcome. Results: Between 9/1/2010 and 12/31/2011, 1188 pts underwent pre-operative evaluation by the MSKCC Geriatric Service. 416 pts (35% of total), median age 80 (75-98) underwent a major cancer surgery and were followed postoperatively by the Geriatrics Service. Most of the patients had hepato-pancreatobiliary (20%), colorectal (18%), head and neck (13.6%) and urological (12.5%) tumors. Delirium (diagnosed by a positive CAM) occurred in 79 pts (19%). CCI score (p = 0.013), falls (p= 0.012), dependence in I-ADLs (p = 0.001) and abnormal Mini-Cog test (p=0.046) were associated with post-operative delirium. ASA score (range 2-4) and operative time (range 8-1005 minutes) were not associated with postoperative delirium. Patients who developed delirium did not show a significant increase in their 30 days readmission rate (11.3% vs. 13.4%) or 6month mortality (7.1% vs. 13.9%). Post-operative delirium resulted in significantly longer length of hospital stay (median 6 vs. 8 days) (Pb0.001) and had greater risk of discharge to a rehabilitation facility (8.2 vs. 26.9%) (Pb0.001). Conclusion: Pre-operative GA may promote a better understanding of older patients’ perioperative risks. It can detect high-risk features for postoperative delirium including high comorbidities, fall history and dependence in I-ADLs. Further prospective studies with preoperative GA and interventions are ongoing.