Reliability, Validity, and Feasibility of a Computer-Based Geriatric Assessment for Older Adults With Cancer

Abstract

Purpose: The goal of this study was to evaluate the feasibility, reliability, and validity of a computer-based geriatric assessment via two methods of electronic data capture ( SupportScreen and REDCap) compared with paper-and-pencil data capture among older adults with cancer. Methods: Eligible patients were ≥ 65 years old, had a cancer diagnosis, and were fluent in English. Patients were randomly assigned to one of four arms, in which they completed the geriatric assessment twice: (1) REDCap and paper and pencil in sessions 1 and 2; (2) REDCap in both sessions; (3) SupportScreen and paper and pencil in sessions 1 and 2; and (4) SupportScreen in both sessions. The feasibility, reliability, and validity of the computer-based geriatric assessment compared with paper and pencil were evaluated. Results: The median age of participants (N = 100) was 71 years (range, 65 to 91 years) and the diagnosis was solid tumor (82%) or hematologic malignancy (18%). For session 1, REDCap took significantly longer to complete than paper and pencil (median, 21 minutes [range, 11 to 44 minutes] v median, 15 minutes [range, 9 to 29 minutes], P < .01) or SupportScreen (median, 16 minutes [range, 6 to 38 minutes], P < .01). There were no significant differences in completion times between SupportScreen and paper and pencil ( P = .50). The computer-based geriatric assessment was feasible. Few participants (8%) needed help with completing the geriatric assessment (REDCap, n = 7 and SupportScreen, n = 1), 89% reported that the length was “just right,” and 67% preferred the computer-based geriatric assessment to paper and pencil. Test–retest reliability was high (Spearman correlation coefficient ≥ 0.79) for all scales except for social activity. Validity among similar scales was demonstrated. Conclusion: Delivering a computer-based geriatric assessment is feasible, reliable, and valid. SupportScreen methodology is preferred to REDCap.