Every family caring for someone with dementia keeps a diary of sorts, even if it never touches paper. The night he got up four times. The afternoon she swore at the neighbour and then forgot. The stretch of good days in early spring. And then, in the clinic, a doctor asks how the last four weeks have gone, and all of that compresses into a number on a scale, filtered through exhaustion and the human tendency to remember the worst hour of the worst day.
That compression is what a group of Italian neurologists set out to attack. In a protocol published in Frontiers in Dementia on 18 December 2023, Federico Emanuele Pozzi and colleagues at the Fondazione IRCCS San Gerardo dei Tintori in Monza, with co-authors from Rome, Chieti, Padua, Bari, Milan, Florence and Pavia, laid out a parallel-arm randomized controlled trial called BPSDiary. The plan: randomize 300 dyads (a person with dementia and their caregiver) across nine Italian centres, 60 of them in Monza and 30 at each of the others, to either usual care or a structured diary in which the caregiver records specific behavioural and psychological symptoms, including insomnia, agitation and anxiety, aggression, purposeless motor behaviour, and delusions or hallucinations, noting the time of onset, the severity, and what might have set it off. Each dyad is followed for 12 weeks. The primary outcome is not a brain scan or a biomarker; it is the change in the caregiver's own Zarit Burden Interview score at three months. Secondary outcomes include NPI scores, caregiver distress in specific symptom domains, satisfaction of both caregivers and physicians, and olanzapine equivalents, a measure of how much antipsychotic medication is being prescribed.
The reasoning behind it is stated plainly in the protocol: existing instruments such as the Neuropsychiatric Inventory rely on recall and lack direct observation. The trial's design does something unusual with the data it collects. Caregivers in the diary arm are asked to send in the first six pages for analysis, and the treating physician then acts on what those pages show, which could mean a phone call, an earlier appointment, a new prescription, or stopping one. Repeat after another six weeks. In other words the diary is not just a measuring stick; it is wired into the consultation, which is precisely why the trial matters to anyone who has been told to "keep track of things at home" without ever being told what happens to the tracking.
The registration tells the rest of the story, and it is a quieter one. ClinicalTrials.gov record NCT05977855, sponsored by the Monza hospital, was first posted on 7 August 2023 with an actual start date of 1 July 2023 and an estimated completion of 31 July 2024. The status information was last verified in September 2023, and the record's overall status now reads unknown, with recruiting as the last known state. No results have been posted there, and no results paper for BPSDiary has been published. Registered protocols outrunning their own timelines is ordinary in academic trial work, particularly for unfunded, low-tech, multi-centre studies that depend on busy clinics. It is still worth saying out loud, because a protocol is a statement of intent, and intent is not evidence.
What has appeared in the meantime is smaller and narrower, and it comes from one of the same authors. On 23 April 2026 the Journal of Alzheimer's Disease published, ahead of print, a cross-sectional study by Marco Toccaceli Blasi and colleagues at the Center for Cognitive Disorders and Dementia at Sapienza University of Rome, with Fabrizia D'Antonio, who is also a BPSDiary co-author, among them. Forty patient-caregiver dyads completed the procedures. Caregivers kept a purpose-built symptom diary for one month; a standard NPI was then administered, and two trained raters independently reconstructed NPI scores from the diary alone. The standard interview produced higher scores and took longer to administer than the diary-derived versions, the diary-based ratings agreed well between raters, and they were less influenced by how burdened the caregiver was. That is a measurement study, not an outcome study: it says something about how the two methods behave side by side in 40 dyads at a single clinic, and nothing about whether keeping a diary makes a caregiver's life easier or changes what a doctor prescribes.
For a family, the practical meaning sits in the gap between those two documents. The idea that a written daily record could be more faithful than a monthly recollection now has some empirical support in a small clinic sample. The idea that handing that record to a clinician measurably reduces the weight a caregiver carries remains, three years after nine Italian centres agreed to test it, an open question with a registry entry and no answer attached.