- Design
- Randomised controlled trial following a qualitative intervention-development phase, using the MRC complex interventions framework
- Population
- 86 informal caregivers of people with dementia randomised, 75 analysed (37 app plus usual care, 38 usual care)
- Primary outcome
- Caregiver burden (Zarit), quality of life (WHOQOL-BREF), neuropsychiatric symptoms and related caregiver distress at 8 weeks
- Effect
- Significant group-by-time interaction for overall and psychological quality of life (both p<0.001); no difference in burden, symptom severity or distress
Most digital interventions for dementia caregivers are built around what developers think carers need. This one was built the other way: qualitative interviews with informal caregivers across the dementia trajectory first, then an application designed to their stated needs following the Medical Research Council framework for complex interventions, then a randomised trial.
The app delivered dementia education, communication and daily-care guidance organised by stage, guidance for behavioural and psychological symptoms, self-care and stress-management resources, information on support services, and customisable reminders. 86 caregivers were randomised and 75 analysed - 37 with the app plus usual care for 8 weeks, 38 usual care alone.
Overall quality of life improved in the app group, as did the psychological domain, both with significant group-by-time interactions (p<0.001). Caregiver burden on the Zarit Burden Interview did not change. Neither did neuropsychiatric symptom severity in the person with dementia, or the caregiver's distress about those symptoms. Usability at eight weeks was good.
That pattern is worth reading precisely rather than as a mixed result. The app made carers feel better without reducing what they were doing or how difficult the person they care for was to look after - which is a real benefit and a limited one. Eight weeks is short, 75 people is small, and quality of life is the softest of the outcomes measured, in an unblinded trial where the intervention group knew they had been given something.
Where it lands practically: signposting a carer to a well-designed resource is worth doing and is not a substitute for respite, home care or treating the neuropsychiatric symptoms. If the burden score has not moved, the burden has not moved.
- Offer a structured caregiver resource, but do not count it as an intervention for caregiver burden
- Assess neuropsychiatric symptoms separately - the app did not change them and they are treatable
- Ask the carer directly about their own mood; the benefit here was psychological
- Check the resource is stage-specific; needs at diagnosis and at end of life are different
- Eight weeks and 75 participants: treat this as encouraging, not as established
The statistics, in plain English
A significant group-by-time interaction means the two groups changed differently over the eight weeks - it does not tell you how large the difference was, and the paper reports the p value rather than an effect size for these outcomes, which is a real limitation when judging whether a carer would notice. Null results for burden and neuropsychiatric symptoms in 75 people should be read as 'not shown' rather than 'no effect': a trial this small can miss a modest but useful difference. Nobody was blinded, and quality-of-life questionnaires are the outcome most sensitive to knowing you received the intervention.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for geriatrics, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free