This was a cross-sectional study conducted between March and October 2024 in both hospital inpatient and outpatient settings in Brasília, the capital of Brazil and the administrative center of the Federal District. Data collection took place at 2 sites: the inpatient geriatric care unit at the Hospital de Base do Distrito Federal (HBDF), a large public hospital integrated into Brazil’s Unified Health System (Sistema Único de Saúde [SUS]), and the Integrated Outpatient Center (Centro Integrado Ambulatorial [CIA]) of the Catholic University of Brasília (Universidade Católica de Brasília [UCB]).

The study protocol was approved by the research ethics committee through the Brazilian CEP/CONEP (Comitê de Ética em Pesquisa–Comissão Nacional de Ética em Pesquisa) system (6.567.586). All participants provided written informed consent prior to enrollment. Data were anonymized and stored in a secure cloud-based system with encrypted storage and restricted access limited to authorized researchers. Participants received no financial compensation for participation.

Participants were aged ≥60 years, in accordance with Brazilian and World Health Organization guidelines. A trained research assistant screened potentially eligible participants in both inpatient and outpatient settings and obtained written informed consent. Inpatients were identified through daily screening of vascular surgery admissions, whereas outpatients were recruited during routine geriatric clinic visits.

Cognitive eligibility was established through medical record review, clinician judgment, and the participant’s ability to independently complete the assessment and provide informed consent. Individuals with severe cognitive impairment, illiteracy, or acute clinical instability were excluded.

Participants completed the self-administered CGA on a touch-screen tablet, with research staff available to provide technical clarification if needed. After completing the CGA, participants answered a standardized usability questionnaire assessing ease of use, navigation, and clarity. Within 48 hours, a geriatrician performed the geriatrician-administered assessment using the same instruments. Data were stored in a secure cloud-based system with encrypted storage and restricted access. No major clinical interventions, including surgery, occurred between the self-administered and geriatrician-administered assessments.

A mobile CGA tool was implemented using a structured Google Forms (Alphabet Inc) interface optimized for tablet-based use. The design followed geriatric-friendly principles, incorporating simplified language and touch-responsive options to facilitate independent navigation by older adults. All instruments were based on validated versions adapted for the Brazilian context. Data were stored in an encrypted cloud-based system with restricted access. Prior to data collection, the tool was tested by the research team to ensure usability and appropriate navigation flow.

Selected items from the CONSORT-EHEALTH (Consolidated Standards of Reporting Trials of Electronic and Mobile Health Applications and Online Telehealth) checklist were used to guide reporting of the digital tool, including its usability features and implementation workflow.

The assessment incorporated validated measures in Portuguese, aligned with the domains recommended by the Brazilian Society of Geriatrics and Gerontology for CGA. The CGA evaluated functional status (Vulnerable Elders Survey–13 [VES-13]), cognition (Cognitive Change Questionnaire–8 [CCQ-8]), emotional health (GDS-5), nutritional status (MNA), fall risk (self-reported falls in the past 6 months), social support (Gijón Scale), and frailty (G8 screening tool). Vision and hearing impairment were assessed through self-reported structured questions. All instruments were applied in their validated Portuguese versions. Participants completed the digital CGA on a tablet designed according to accessible design principles, including large fonts, high contrast, intuitive icons, and a light background. Demographic variables included sex, age, weight, height, BMI, marital status, and education level.

Feasibility outcomes included assessment completion time, participant satisfaction, and perceived ease of use of the digital tool.

Sample size was determined based on the precision required to estimate the intraclass correlation coefficient (ICC). Assuming an expected ICC of 0.80, a sample of 80 paired assessments was estimated to provide approximately 80% assurance that the lower bound of the 1-sided 95% CI would remain >0.70, a threshold considered indicative of good agreement. This approach followed the precision and assurance framework described by Zou [34].

Statistical assumptions were initially checked, including the presence of outliers, missing data, and normality. Because the data were not normally distributed, nonparametric tests were applied, following the recommendations of Tabachnick and Fidell [35]. Spearman rank correlation was used to examine agreement between self-administered and geriatrician-administered assessments. Wilcoxon signed-rank tests were used to compare paired scores overall and by subgroup (inpatients vs outpatients). For categorical variables such as history of falls, hearing impairment, and vision impairment, Cohen κ coefficient was used to assess item-level agreement, as it provides a more robust estimate than simple concordance percentages by adjusting for chance agreement [36,37]. Between-group comparisons by sex and care setting used the Mann-Whitney U test. Spearman rank correlations also examined associations between education and assessment or usability variables, including completion time and satisfaction. All analyses were 2-tailed, and P<.05 was considered statistically significant. Data were processed using SPSS (version 27; IBM Corp).

This study demonstrates that a mobile, self-administered geriatric screening tool based on the domains of the CGA is a feasible, acceptable, and reliable method for screening older adults in both inpatient and outpatient settings in Brazil. It is important to emphasize that this mobile tool should be interpreted as a screening and triage instrument rather than a replacement for the full CGA. Its main value lies in helping to save specialist time, broaden access, and identify older adults who are most likely to benefit from a complete CGA. Our findings support the utility of this mobile tool as a screening gateway to the CGA, complementing but not substituting for the clinical comprehensiveness of professional assessments.

High concordance was observed between self-administered and geriatrician-administered assessments across multiple objective domains, including hearing, vision, and history of falls. The use of Cohen κ strengthened the reliability analysis in these categorical domains, as this measure adjusts for agreement expected by chance, reinforcing the robustness of concordance in falls, hearing, and vision assessments.

Our findings are consistent with prior studies demonstrating the feasibility and reliability of self-administered geriatric assessment tools, particularly those delivered through digital platforms and supported by user-centered design principles. Previous research has shown that mobile or electronic CGA tools can produce results comparable to clinician-administered assessments, even among older adults with complex health conditions or those from vulnerable populations [23,24,27,38]. Prior studies have similarly shown that mobile platforms, culturally adapted instruments, and user-centered design approaches can enhance accessibility, reliability, and clinical utility.

Despite these promising results, discrepancies emerged in more subjective domains, particularly nutritional status (MNA) and social vulnerability (Gijón Scale). In both inpatient and outpatient settings, participants rated their nutritional status as poorer than geriatricians, which may reflect self-perceived vulnerability influenced by appetite changes, undocumented weight loss, and the cumulative impact of chronic conditions—factors that may not always be captured in professional evaluations. In the social domain, inpatients reported higher levels of social risk than were perceived by professionals. This likely reflects feelings of isolation, loss of autonomy, and heightened vulnerability associated with hospitalization, experiences that are not always evident in an objective clinical assessment. Such findings suggest that self-reports can capture subjective perceptions and psychosocial stressors that may remain invisible in geriatrician-administered assessments [36,39].

Educational level also influenced a few outcomes. Geriatricians’ social risk ratings increased with lower educational attainment, whereas older adults’ self-assessed social risk was not associated with education. Similarly, lower educational attainment correlated with poorer cognitive scores on the CCQ-8 in geriatrician-administered assessments, with only marginal significance in self-administered results. As expected, participants with fewer years of education required more time to complete the self-assessment and reported lower ease of use, yet satisfaction remained high across all educational groups. These findings highlight the importance of inclusive and user-centered interface design, including large fonts, high-contrast layouts, simplified language, and intuitive navigation, to reduce digital barriers for older adults with lower literacy or limited technology experience. Additional accessibility features, such as audio narration, multimedia guidance, and optional user support, may further enhance usability, particularly among individuals with very low educational attainment or limited digital familiarity.

From a systems perspective, integrating mobile CGAs into electronic health records (EHRs) and care coordination platforms could enable automated risk stratification, population health management, and timely referrals to interdisciplinary teams. In low- and middle-income countries, this approach may improve care continuity while reducing demands on scarce geriatric specialists. In our study, the average completion time of <20 minutes, combined with high satisfaction ratings, supports the feasibility of incorporating the self-administered CGA into routine workflows. This is particularly relevant within Brazil’s Unified Health System (SUS), where specialist availability is limited. Previous research has shown that digital CGAs can streamline clinical encounters without compromising care quality [24] and that structured electronic tools can enhance diagnostic accuracy and geriatrician acceptance [40].

This study has several limitations. It focused primarily on urban, literate older adults, which may limit generalizability to rural populations, individuals with low literacy, or those with moderate to severe cognitive impairment. Future adaptations should explore simplified interfaces, audio guidance, caregiver-assisted administration, and other accessibility strategies to improve the inclusion of socially, educationally, and cognitively vulnerable older adults.

A prospective screening log documenting all potentially eligible individuals, refusals, and reasons for nonparticipation was not systematically maintained. Therefore, participation rates and the potential magnitude of selection bias could not be fully assessed.

Because the self-administered assessment was completed before the geriatrician-administered evaluation, an order effect may have artificially increased agreement between methods. However, geriatricians performing the professional assessments did not have access to participants’ self-administered responses, reducing the likelihood of direct observer bias. Although assessments were conducted within 48 hours, minor clinical fluctuations may still have influenced some responses.

The cross-sectional design allowed the evaluation of agreement and usability but not predictive validity. Longitudinal studies are needed to determine whether self-administered CGA scores are associated with clinically meaningful outcomes such as hospital readmissions, functional decline, institutionalization, and mortality.

Finally, because multiple comparisons were performed across CGA domains and care settings, the possibility of type I error should be considered. Given the exploratory nature of the study, no formal correction for multiple comparisons was applied; therefore, domain-specific findings should be interpreted with caution.

The mobile, self-administered CGA demonstrates strong potential to expand access to geriatric screening and triage, particularly in settings with limited specialist availability. As populations age and health care systems face increasing demands, such tools may support early detection, care coordination, and more equitable outcomes, especially in low- and middle-income countries.

By integrating patient-reported data into digital health systems, these tools enhance person-centered, data-informed care. High concordance in objective domains supports the reliability of the self-administered tool, while the link between perceived vulnerability and adverse outcomes highlights the value of including subjective measures when they are culturally adapted.

Importantly, this tool is not a substitute for clinical evaluation but rather a complement to existing workflows and may ultimately enhance patient engagement in care.

In summary, the self-administered mobile tool should be understood as a screening and triage instrument that complements, rather than replaces, the full CGA. By integrating objective measures with patient-reported perceptions, it has the potential to expand access to geriatric screening, reduce the burden on specialists, and support clinical decision-making and public health strategies aimed at healthy aging, particularly in resource-limited health systems.