Agreement Between a Mobile Self-Administered Comprehensive Geriatric Assessment Screening Tool and Geriatrician-Administered Assessment: Cross-Sectional Feasibility Study

Abstract Background Comprehensive geriatric assessment (CGA) is a widely recommended, multidimensional approach for guiding clinical decision-making and management in older adults. However, its implementation remains limited by a shortage of geriatric specialists, the complexity of geriatric care, and the increasing demands of an aging population. Mobile self-administered tools offer a potential strategy to expand access to multidimensional geriatric evaluation, particularly in resource-limited settings. Objective This study aimed to evaluate the feasibility and agreement of a mobile self-administered CGA screening tool compared with geriatrician-administered assessments among inpatient and outpatient older adults in Brazil. Methods This cross-sectional study included 80 older adults recruited from inpatient and outpatient geriatric clinics in Brasília, Brazil. Participants completed a mobile CGA, which incorporated validated instruments covering functional status (Vulnerable Elders Survey–13 [VES-13]), cognition (Cognitive Change Questionnaire–8 [CCQ-8]), depressive symptoms (5-item Geriatric Depression Scale [GDS-5]), nutritional status (Mini Nutritional Assessment [MNA]), frailty (G8 screening tool), social support (Gijón Scale), falls, and vision and hearing. Within 48 hours, a geriatrician independently performed a geriatrician-administered CGA using the same instruments. Agreement between the 2 assessment methods was evaluated using Wilcoxon signed-rank tests, Spearman correlation coefficients, intraclass correlation coefficients (ICCs), and Cohen κ coefficients for categorical variables. Results Participants had a mean age of 70 (SD 7) years; 58.8% (47/80) were female, and 65% (52/80) had ≤8 years of education. The mean completion time for the self-administered CGA was 18.5 (SD 7.5) minutes. Most participants rated the tool as easy or very easy to use (65/80, 81.3%) and reported satisfaction with the assessment process (n=79, 98.8%). High concordance was observed between the self-administered and geriatrician-administered CGA versions for most domains. Cohen κ coefficients demonstrated strong agreement for falls (κ=0.878; P <.001), hearing impairment (κ=0.826), and vision impairment (κ=0.634). No significant differences were observed between the 2 assessment methods for functional status (VES-13), frailty screening (G8), depressive symptoms (GDS-5), and cognition (CCQ-8). Nutritional status and social vulnerability showed lower concordance than other CGA domains. Among inpatients, significant discrepancies were identified in both nutritional scores and Gijón social risk scores ( P ≤.05), with participants reporting greater perceived vulnerability. For outpatients, significant differences were observed for nutritional status (MNA; P=.02) and frailty screening (G8; P=.046). Conclusions A mobile self-administered CGA demonstrated good feasibility, high user satisfaction, and substantial agreement with geriatrician-administered assessments across multiple geriatric domains. The tool may support geriatric screening and triage in settings with limited specialist availability, particularly when used as a complement to comprehensive clinical assessment.