The added value of geriatric screening and assessment for predicting overall survival in older patients with cancer

Abstract

BACKGROUND: The aim of this study was to determine and compare the added prognostic value of screening tools, geriatric assessment (GA) components, and GA summaries to clinical information for overall survival (OS) in older patients with cancer. METHODS: A screening and a 10-item GA were systematically performed in patients ≥70 years old with cancer. Cox regression analyses were conducted to evaluate the added prognostic value for OS of screening tools, GA, and GA summaries to clinical information (age, stage, and tumor type) in 2 cohorts (A and B). Cox models were compared on the basis of the Akaike information criterion and the concordance probability estimate. The 2 cohorts for the analyses were similar but independent. RESULTS: A complete case analysis was available for 763 patients (median age, 76 years) in cohort A and for 402 patients (median age, 77 years) in cohort B. In both cohorts, most individual GA components were independent prognostic factors for OS. Nutritional status (assessed with the Mini Nutritional Assessment Short Form) and functional status (assessed with the Instrumental Activities of Daily Living) consistently displayed a strong capacity to predict OS. Less consistent results were found for screening tools. GA summaries performed the best in comparison with the screening tools and the individual GA components. CONCLUSIONS: Most individual GA components, especially nutritional status and functional status, are prognostic factors for OS in older patients with cancer. GA summaries provide more prognostic information than individual GA components but only moderately improve the prognostic baseline model with clinical information.