In the context of primary care, cognitive screenings are brief, non-diagnostic tests that clinicians can administer in order to provide appropriate referrals to neuropsychologists. Annual cognitive screening for adults over age 65 (“older adults”) can help monitor cognitive functioning over time and ensure more patients with cognitive impairments receive neuropsychological assessment and care earlier. Unfortunately, time constraints and lack of training present major barriers to cognitive screening in primary care, and less than half of cognitive impairment cases are identified in these settings. A remote cognitive screening mobile app has the potential to save primary care clinics time, particularly for the majority of older adults who are cognitively healthy. Moreover, a screening app well-validated for remote clinical use can replace the inadequate or nonexistent screening practices currently employed by many primary care clinics. In order to achieve their potential, remote smartphone-enabled cognitive screening paradigms must be acceptable and feasible for both patients and clinical end users. With this goal in mind, we describe the collaborative, human-centered design process and proposed implementation of MyCog Mobile (MCM), a self-administered cognitive screening app based on well-validated NIH Toolbox measures.

We conducted foundational interviews with primary care clinicians (N=5) and clinic administrators (N=3) and created user journey maps of their existing and proposed cognitive screening workflows. We then conducted individual semi-structured interviews with healthy older adults (N=5) as well as participated in a community stakeholder panel of older adults and caregivers (N=11). Based on the data collected, we developed high-fidelity prototypes of the MCM app which we iteratively tested and refined with the older adult interview participants. Older adults rated the usability of the prototypes on the Simplified System Usability Scale (S-SUS) and After Scenario Questionnaire (ASQ).

Clinicians and administrators were eager to use a well-validated remote screening app if it saved them time in their workflows and were fully integrated into their EHR. Clinicians prioritized easily interpretable score reports tied to automated best practice guidelines. Findings from interviews and user journey mapping further informed the details of the proposed implementation and core functionality of MCM. Older adult participants were motivated to complete a remote cognitive screener to ensure they were cognitively healthy, save time during their in-person visit, and for privacy and comfort reasons. Older adults also identified several challenges to remote smartphone screening which informed the user experience design of the MCM app. The average rating across prototype versions was 91 (SD 5.18) on the SSUS and 6.13 (SD 8.40), indicating above average usability.

Through our iterative, humancentered design process, we were able to develop a viable remote cognitive screening app and proposed implementation for primary care settings optimized for multiple stakeholders. Next steps include validating MCM in clinical and healthy populations, collaboratively developing best practice alerts for primary care EHRs with neuropsychologists, and piloting the finalized app in a community clinic. We hope the finalized MCM app will promote broader screening practices within primary care and improve early assessment and diagnosis of cognitive impairment for older adults.