e19586 Feasibility of computer-based self-administered cancer-specific geriatric assessment (SA-CSGA) in older pts w/ gastrointestinal malignancy (GIM) Background: The CSGA (Hurria, JCO 2011) is a brief geriatric assessment consisting of validated measures primarily self-administered using paper format. We developed & tested feasibility of a computer-based SA-CSGA in pts ≥70 yrs w/ GIM. Methods: From 12/2009 - 6/2011, pts ≥70 yrs receiving treatment (rx) for GIM at Dana-Farber Cancer Institute were consented to complete SA-CSGA at baseline (T1= new or change rx) & follow-up (T2 = w/in 4 wks of completing rx). Feasibility endpts are (1) proportion of eligible pts consenting; (2) proportion completing SA-CSGA at T1 & T2; (3) time to completion of SA-CSGA; (4) proportion of MDs reporting change in clinical decision-making due to SA-CSGA. Results: Of the 49 eligible pts, 38 consented (55% female, 89% White, 76% enrolling prior to new rx). Mean age was 77yrs (range 70-89), 38% completed college, 49% married, 27% live alone, and 78% retired. 50% were diagnosed w/ colorectal cancer (ca). Mean MD-rated Karnofsky Performance Status was 87.5 at T1(range 60-100), 83.5 at T2 (range 70-100). At T1, 92% used a touch screen computer; 97% completed the SA-CSGA (51% independently). At T2, all pts used a touch screen computer; 71% completed the SA-CSGA (41% independently). Reasons for not completing SA-CSGA were withdrawal of consent (n=1 at T1 & T2), transfer of care (n=3; T2) or death (n=7; T2). The dominant reason for needing assistance was lack of computer familiarity (n=17 T1, n=14 T2). Mean time to completion was 23min at T1 (range 15-58); 20min at T2 (range 13-35). Among the 8 MDs who consented to participate, SA-CSGA added information to clinical assessment for 75% at T1 (n=27) and 65% at T2 (n=17) but did not alter immediate clinical decision-making. Conclusions: The computer-assisted SA-CSGA feasibility endpt was met for older pts w/ GIM although approximately half required assistance. While the SA-CSGA added information to clinical assessment, results did not impact clinical decision-making. Reasons for this may include relatively high-functioning patients enrolled in this study.
Computer-Based Patient Reported Outcomes in Outpatient Palliative Care: Feasibility
