Prospective study comparing self-administered geriatric assessment to provider’s routine clinical assessment of older patients with metastatic breast cancer treated at community oncology practices.

Abstract

12029 Background: Geriatric Assessment (GA) is recommended for evaluating an older cancer patient’s (pt) fitness for treatment. We conducted a prospective study evaluating the current gaps that exist in the assessment of older pts with metastatic breast cancer (MBC) in community practices (CP). Methods: Self-administered validated GA was compared to provider assessment (PA) of MBC pts ≥ 65-years-old treated at CP in the US. Providers were blinded to the GA results until their evaluation was completed. Differences in PA vs GA detected abnormalities were assessed using McNemar’s test. The effect of patient/provider factors on the rate of abnormalities not identified was assessed using regression models, clustering by provider and adjusting for the number of prior pts seen. Results: 100 pts were enrolled across 9 CP (median age 73.9, (65-90)). GA detected a total of 356 abnormalities in 96/100 (96%) pts, of which 223 required immediate interventions. African American and widowed/single pts were more likely to have abnormalities identified by GA. On average PA did not identify abnormalities detected by validated GA in 2 of 8 domains. 73% of functional status, 86% of social support, 44% of nutritional, and 96% of cognitive abnormalities detected by GA were not identified by PA (all P < 0.0001). Providers with more years of clinical experience were more likely to identify abnormalities (compared to < 5 years (y) in practice: 5-10 y in practice, p = 0.149; 11-15 y in practice, p = 0.028; > 15 y in practice, p = 0.017). GA had the most significant impact on pts with decreased ECOG PS (p = 0.045). Pts found to have an abnormal Timed Up and Go (TUG) test were more likely to have additional abnormalities in other domains (mean 4.3 vs 2.1, Wilcoxon p < 0.001), and more abnormalities not identified by the PA (p < 0.001). Providers were “surprised” by GA results in 33% of cases, mainly with cognitive or social support findings, and reported plans for management change for 40% of pts based on GA findings. Conclusions: Including a GA in the care of older pts with MBC in CP is beneficial as validated GA has a high detection rate of abnormalities not detected by PA.