Prospective evaluation of a comprehensive geriatric assessment (CGA) in multidisciplinary bladder cancer care: Feasibility and impact on decisional conflict.

Abstract

479 Background: Commonly utilized risk stratification tools demonstrate inconsistent associations with salient clinical outcomes in bladder cancer leading to a disproportionate reliance on providers’ subjective impression of a patient’s fitness for therapy. Current guidelines advocate for use of a CGA to quantify vulnerabilities in older ( > 65 years) patients before treatment selection. Our objective was to prospectively evaluate CGA in our Bladder Cancer Multidisciplinary Clinic (BCMC). We hypothesized that CGA implementation would be feasible and that discussion of the results during shared decision-making would be associated with reduced patient-reported decisional conflict. Methods: Patients seen in BCMC were prospectively enrolled from 6/1/20 to 7/20/21. In the first 3 months, participants underwent non-standardized risk stratification (“Routine cohort”, N = 27). Between 9/1/20 and 7/20/21, participants completed a CGA incorporating validated assessments of frailty, functional status, multimorbidity, nutrition, cognition, and mental health (“CGA cohort”, N = 67). Results were shared with patients during BCMC visits. All patients and providers (three physicians per clinic from: Uro-Oncology, Medical Oncology, and Radiation Oncology) completed a follow-up survey including the Decisional Conflict Scale (DCS). Time required to complete the CGA, completion rates, and patient-reported burden were assessed. Concordance of patient- and provider-reported decisional conflict was compared between Routine and CGA cohorts. Results: Of 138 eligible patients, 94 patients were successfully enrolled (68%) with median age of 72 years, ECOG PS ≥3 in 13%, and Charlson Comorbidity Index ≥3 in 18%, of whom 18% were women. Most patients had pT2 bladder cancer (87%; cN+ and M1 in 23.4% and 9.6%, respectively). CGA component completion rates were 79-100%. Survey response rates were high (patients: 77%, providers: 86%), and most (86%) patients felt that the CGA was, at most, minimally burdensome. Vulnerabilities detected across CGA domains triggered relevant referrals. Patient-reported median (IQR) DCS scores were numerically higher (e.g. greater decisional conflict) for the CGA cohort: (27 [14-33] vs 16 [2-30] for Routine patients, p = 0.28). Provider- and provider reported DCS score was correlated in the CGA (p = 0.04), but not the Routine cohort (p = 0.07). Conclusions: We prospectively evaluated use of CGAs in bladder cancer care and found that CGAs were successfully implemented with high rates of completion and low rates of perceived burden. Notably, in this pilot cohort of 94 patients, DCS scores did not differ significantly between patients and providers with CGA use. Future work will evaluate associations between individual instruments, treatment decisions, clinical outcomes and patient-reported quality of life measures.