Background Subtle impairments in complex functions cruicial to independence, such as instrumental activities of daily living (IADLs), can predict cognitive decline. Mild cognitive impairment (MCI) is a condition defined by changes in cognitive function and minor impairment in functional abilities. Digital technologies, including ambient sensors/wearables, can capture changes in IADL-related behaviours in an objective and continuous way, which may indicate disease progression. It is unclear which digital metrics and methods have been used to measure IADL-related behaviours in the literature, and how these digital endpoints may differ between people with MCI and normal ageing. A systematic review was conducted to understand how IADL-related behaviours have been digitally assessed in people with MCI.

Methods A systematic search for articles published between 2004–2022 that used continuous, remote, digital technology to measure IADL-related behaviours in people with MCI was conducted in eight databases (CINAHL/Embase/Medline/ProQuest/PsycInfo/PubMed/Web of Science/Scopus) using appropriate key terms. Research conducted in aged-care facilities, that used scripted tasks or that included populations with diagnoses of other conditions that impacted cognition or function were excluded. Screening and quality assessment were undertaken by three independent reviewers. This review followed the PRISMA guidance and was registered on PROSPERO (CRD42022326861).

Results A total of 16,099 titles and 995 abstracts were screened, with 15 articles included for narrative synthesis. IADL-related behaviours were grouped into relevant domains, five of which were assessed using 11 different technologies; ambient motion sensors were the most commonly used. Seventy-nine discrete metrics were reported. In the domain Activities Outside of the Home, two studies reported significant decline in IADL-related behaviours for people with MCI compared to normal ageing. Significant decline in the domain Everyday Technology Use was reported by two studies, and a significant decline in IADLs related to the domain Orientation was reported by one study in people with MCI compared to normal ageing. No digital technology was used to measure IADL-related behaviours in the domains Communication and Socialisation, or Culturally Specific Tasks. User acceptability of the devices was measured by two studies, both reporting >80% positive feedback. Longitudinal findings were limited; however, increasing variability in the performance of some IADL-related behaviours was reported, including computer usage and medication management in populations with MCI compared to normal ageing.

Conclusion These review findings will inform the selection of digital tools capable of assessing IADL-related behaviours in populations with MCI. We recommend increased validation of digital technologies alongside increased longitudinal research to monitor key IADL markers of decline.