Every family living with dementia eventually invents its own instrument. A spiral notebook by the kettle. A note on a phone at 4:40 in the afternoon, when the restlessness starts again. A running tally, half memory and half guesswork, carried into a fifteen-minute clinic appointment where it has to be compressed into a sentence. The instrument is homemade, the data are real, and almost nobody has ever asked, systematically, what shape those records take once they move onto a screen.
A systematic review by Felipe Soto-Perez and Camila Ruy Castilla at the Universidad de Salamanca, with Madalin M. Deliu at the Universidad Catolica de Avila, has now done the inventory. Published in PsyCh Journal on 5 April 2026 (volume 15, issue 2, article e70092), it gathered 24 studies published between 2019 and 2025, found through a deliberately broad search of abstracts for dementia and caregiver combined with mHealth, app or mobile application. The authors were interested in one component in particular: monitoring, the act of a caregiver recording something and then looking at what they recorded. Their synthesis is narrative. They did not pool results, and they say why, citing heterogeneity in intervention designs, outcome measures and methods, along with generally small samples and what they call "considerable preliminary or non-controlled designs".
What emerges is a taxonomy, and for anyone thinking about the interface between a family's records and a clinician's notes, the taxonomy is the story. The review sorts monitoring into distinct kinds: personal diaries and chats, which combine individual logging with a direct channel to professionals; ecological momentary assessment, meaning systems built to catch an emotional state in real time; periodic health self-assessments, the instruments that measure caregiver burden and general health on a schedule; monitoring of evolving care needs; cognitive pattern management, aimed at identifying negative mental schemas; competency self-monitoring platforms that test dementia knowledge and skills, including virtual reflection sessions; and geolocation, generating safe zones as a form of indirect supervision. Named apps sit under those headings. Dementia Talk carried behavior tracking. Brain Care Notes offered care tips generated from weekly assessments. Ethica prompted caregivers to self-report emotional states daily. The MIT App sent automated daily reminders toward mindfulness practice. WECARE gave caregivers self-assessment tools whose results they could bring into group discussion.
Read as a map of raw material, this is abundant. Read as a map of clinical plumbing, it is thinner than it looks. Almost all of this monitoring was designed to point back at the caregiver: to help them notice a stressor, judge whether a strategy is working, and adjust. Only one of the review's categories, personal diaries and chats, is described as including a direct line to a professional at all. The authors' own recommendation makes the gap explicit: they call for standardizing monitoring metrics and for larger, more diverse evaluations before long-term impact can be claimed. Standardized metrics are precisely what a physician would need to read a caregiver's month at a glance rather than reconstruct it by interview.
The review is also candid about the thing every family already suspects. Engagement declines. App use falls off over time, and the authors treat that decline as a design problem rather than a caregiver failing, arguing for tools that change as needs change. Digital literacy varied widely across the studies, which means the people most isolated from support may be the least served by a phone-based version of it. And the outcomes under study were caregiver burden and wellbeing. None of this was about detecting or dating disease in the person being cared for, and none of it was measured against the number that would matter most to a clinic: whether a caregiver's log ever changed a decision.
So the honest position, on the evidence as it stands in September 2026, is that the raw material for a physician-ready summary is being generated in tens of thousands of homes, in at least seven different formats, on no agreed scale, with no demonstrated route from the phone to the consulting room.