There is no published step-by-step protocol for delivering doll therapy as a group activity, because the evidence base treats it as a one-to-one intervention. The 2024 scoping review in Nursing Reports synthesised the documented studies and found doll therapy delivered individually, resident by resident, rather than as a scheduled group session.
So "doll therapy for memory care groups" is best understood as a unit-wide programme built from individual sessions, not as a circle-time activity. Doll therapy means offering a person living with dementia a lifelike infant doll as an optional comfort object, and supporting whatever caregiving or attachment behaviour follows. The steps below come from the two trials that published an actual sequence, plus the ethical rules that govern who is offered a doll and when it is taken away.
Medical information disclaimer: This article is for general educational purposes only and does not provide medical advice, diagnosis, or treatment. Always consult a physician or other qualified health professional about symptoms, medications, tests, or treatment decisions.
Table of Contents
- What the evidence says the therapy does
- The two documented session structures
- A practical sequence for a memory care unit
- Who should be offered a doll, and who decides
- The limits worth knowing before you start
- Frequently Asked Questions
What the evidence says the therapy does
The 2024 scoping review from Queen's University Belfast (Henderson, McConnell and Mitchell) pulled together 12 primary studies published between 2013 and 2023 across CINAHL Plus, Medline, PsycINFO and PubMed. Their synthesis found three consistent themes: reduced behavioural and psychological symptoms of dementia, increased communication, and reduced caregiver distress. A 2023 systematic review and meta-analysis of nursing-home residents, with databases searched to 25 July 2023, put a number on the main effect. Published in Geriatric Nursing, it reported a significant reduction in agitation (SMD = −0.94, p < 0.001), alongside improvements in apathy, irritability, wandering, pleasure, anxiety and depression.
Those are meaningful effects, but they are not universal. An Italian randomised trial of 52 nursing-home residents — 26 assigned to doll therapy, 26 to standard care, measured at baseline, 45 days and 90 days — found reduced agitation, aggressiveness and caregiver distress but no improvement in eating behaviour. Doll therapy addresses distress and engagement. It is not a nutrition intervention.
The two documented session structures
Cantarella and colleagues ran a single-blind randomised trial in Canton Ticino nursing homes that is the closest thing to a published step-by-step protocol. Residents were randomised 1:1 to a doll or a sham non-anthropomorphic cube, and each received 30 daily sessions of up to one hour, led by a trained nurse. Note the shape: one nurse, one resident, daily, for a month. The second sequence is the staff-mediated package trialled in Zhengzhou, China, and it is the one a facility can actually build a programme around.
As described in Frontiers in Public Health, it runs: staff training first, then a structured introduction of eight planned sessions over about 15 days, then continued access to a personal doll during routine care, relational engagement by staff, and ongoing monitoring of assent, dissent and distress. The difference matters for staffing. The Swiss protocol demands a trained nurse for up to an hour a day per resident for 30 days — rarely affordable across a whole unit. The Chinese package front-loads eight structured sessions, then folds the doll into care that is happening anyway, which is how it scales.
A practical sequence for a memory care unit
Drawing directly on the Zhengzhou package, a unit-level programme has a workable order of operations: The group element is organisational, not therapeutic. You schedule and track it across the unit; you deliver it person by person.
- Train staff before any doll appears on the unit, so the first offer is handled consistently.
- Offer the doll to one resident at a time, in a calm moment, and watch the first reaction rather than explaining or persuading.
- Run roughly eight structured introduction sessions across about two weeks for residents who respond well.
- Give responding residents continued access to their own personal doll, including during washing, dressing and mealtimes.
- Have staff engage relationally with the doll as the resident treats it, without correcting or contradicting.
Who should be offered a doll, and who decides
Suitability is not read off a diagnosis or set by a blanket facility policy. Because the doll is offered as an optional comfort object with assent monitored throughout, the Frontiers in Public Health work indicates staff judge from the resident's own first reaction. A resident who ignores the doll, pushes it away, or becomes agitated has answered the question.
This is also the practical answer to the ethical objection. Writing in Nursing Ethics, Mitchell and Templeton addressed the charge of infantilisation — treating an adult childishly and reinforcing dementia stigma — and argued for a rights-based approach grounded in the UN Convention on the Rights of Persons with Disabilities, in which the doll is offered rather than imposed. Two operational rules follow. Never place a doll in someone's arms as a fait accompli, and never remove one without the resident's agreement, including for laundering or during a family visit.
The limits worth knowing before you start
The Cantarella trial's published results found residents showed more interest in the doll than the cube, directed caregiving and exploratory behaviours toward it, and accepted separation from the nurse more readily. But the same paper reported no difference between the doll and cube groups in stress biomarkers. Observable engagement improved; measured physiological stress did not separate. Outcome measurement is still maturing.
The Zhengzhou trial (NCT06506487) cluster-randomised six nursing homes 1:1, assessed residents at baseline, 6 weeks and 12 weeks, and — per its ClinicalTrials.gov registration — used class-specific standardised daily psychotropic doses as its registered primary outcome, with NPI-12 as the main clinical outcome. It ran from 31 July 2024 to 31 December 2025. For a family or a care manager, the honest summary is this: the agitation evidence is reasonably strong, the mechanism evidence is thinner, and nothing in the record supports running doll therapy as a group session. If a facility describes a scheduled group doll activity, ask how each resident's assent is tracked within it.
Frequently Asked Questions
How long before we see a change?
The Italian trial measured residents at baseline, 45 days and 90 days, and the Zhengzhou trial at baseline, 6 weeks and 12 weeks. Both structures assume weeks, not a single session.
Should each resident have their own doll or should dolls be shared?
The Zhengzhou package gives responding residents continued access to a personal doll during routine care. A shared pool undercuts the attachment the therapy depends on and creates removal conflicts.
What if a resident's family objects?
The ethical case rests on the doll being offered, not imposed, under a rights-based reading of the UN Convention on the Rights of Persons with Disabilities. That same logic makes the resident's own reaction — not a relative's discomfort — the deciding signal.
