Co-Producing Dementia Support With Families: Roles, Boundaries and Safe Decision-Making

February 22, 2026

Families are often the most consistent source of knowledge, advocacy, and day-to-day support in dementia care. Yet “family involvement” can become informal, inconsistent and unsafe if services do not set clear roles, boundaries and review points. This article sets out an operational co-production approach that protects the person’s rights, improves outcomes, and reduces avoidable conflict. It should be used alongside person-centred planning in dementia services and embedded within dementia service models and care pathways so co-production is part of routine delivery, not an ad-hoc conversation after a problem occurs.

Why co-production breaks down in dementia services

Most conflict is predictable. It usually arises from one of four gaps:

Unclear decision rights: who decides what, and when the person’s choices can be followed safely.
Inconsistent communication: families receiving mixed messages from different staff, or updates only after incidents.
Undefined risk tolerance: families and staff holding different views on “acceptable risk” without a structured way to agree plans.
Poor recording: verbal agreements not captured, leading to disagreement about what was said or decided.

Co-production is not the same as “the family decides.” It is a structured method to involve families meaningfully while keeping the person at the centre, applying least restrictive practice, and meeting safeguarding duties.

Set the rules early: a simple co-production framework

1) Establish the “care partnership agreement”

Within the first weeks (or on transfer), create a short, plain-English agreement that covers:

How updates happen: frequency, format, and who is contacted for what (routine vs urgent).
What “good” looks like: outcomes that matter to the person and family (comfort, routines, community contact, reduced distress).
Decision-making boundaries: what the person decides day to day, what needs best-interest decision-making, and how disagreements are handled.
Escalation routes: named roles (keyworker/lead, manager, clinical input if relevant) and response time expectations.

This is not bureaucracy. It prevents repeated “re-set” conversations and makes reviews faster and safer.

2) Make capacity and consent practical (not theoretical)

Capacity is decision-specific and can fluctuate. Operationally, teams should:

Record how the person is supported to understand choices (timing, environment, communication aids).
Capture what the person can decide reliably (clothes, meals, routines, visitors) and what needs structured review.
Re-check capacity at key points (health changes, hospital discharge, significant incidents, safeguarding concerns).

Families should understand that the person remains central even when memory is impaired. Co-production must reinforce this, not replace it.

3) Turn disagreement into a managed process

Disagreement is not automatically safeguarding, but unmanaged conflict increases risk. Use a consistent approach:

Define the issue: what decision is being disputed (e.g., medication timing, going out alone, moving rooms).
Clarify evidence: what incidents, outcomes, and observations exist (not opinions).
Agree options: least restrictive option first, with safeguards and clear review points.
Set time-limited trials: “Try for two weeks, measure X, review on date Y.”
Record decisions: what was agreed, who agreed, what the person expressed, and what triggers escalation.

This keeps the focus on the person’s welfare and rights while protecting staff from “moving target” expectations.

Operational examples

Example 1: Family requests constant observation after a fall

Context: A person had two falls in one week. The family demanded 1:1 observation and restricted movement. Staff were concerned this would reduce mobility and independence and could increase agitation.

Support approach: The service held a structured co-production review within 48 hours. They shared the incident pattern (times, locations, footwear, toileting), and gathered the person’s views about walking and routines.

Day-to-day delivery detail: A time-limited plan was agreed: increased supervision at peak risk times (toileting and late evening), environmental changes (lighting, decluttering, chair height), proactive prompts for hydration and toileting, and a short daily mobility routine. Staff recorded early fatigue cues and agreed a “rest before risk” approach rather than blanket restriction.

How effectiveness is evidenced: Falls reduced, the person maintained walking ability, and the family saw clear data in weekly updates. The review record showed least restrictive options were trialled first, with measurable checks and a defined escalation threshold.

Example 2: Disagreement about medication and “chemical restraint” concerns

Context: The family believed prescribed medication was over-sedating the person. Staff reported distress and poor sleep when doses were missed, leading to late-night incidents.

Support approach: The manager set up a co-produced monitoring plan: a shared log of sleep, distress cues, engagement, and side effects, with clear times and staff guidance on consistent observations.

Day-to-day delivery detail: Over two weeks, staff recorded daytime alertness, participation in activities, appetite, and night waking. The family contributed observations during visits. The service agreed that any PRN use required a documented trigger, de-escalation attempts first, and a review note explaining why it was proportionate at that time.

How effectiveness is evidenced: The combined log showed patterns, reduced assumptions, and enabled a balanced discussion at review. The plan included clear review dates and demonstrated that restrictive interventions were not used as a default response.

Example 3: Community access and “acceptable risk” for a person who wants independence

Context: A person wanted to continue walking to a local shop. The family feared wandering and demanded a ban on leaving the service unaccompanied. Staff were concerned this would increase distress and reduce quality of life.

Support approach: A co-produced risk enablement plan was agreed that prioritised least restrictive practice. The person’s preferences were documented clearly, and the family’s concerns were converted into measurable safeguards rather than blanket restrictions.

Day-to-day delivery detail: The service mapped a “safe route,” agreed time windows, and introduced graduated support: initially accompanied walks, then shadowing at a distance, then independent trips with check-in points. Staff checked footwear and hydration, noted triggers for confusion (noise, crowds), and agreed a rapid response plan if the person did not return within an agreed time. The review schedule was explicit: weekly for four weeks, then monthly.

How effectiveness is evidenced: The person maintained independence with reduced distress, and the family saw structured safeguards. Records demonstrated a clear rationale, a least restrictive approach, and ongoing review rather than one-off permission.

Commissioner expectation

Commissioner expectation: Services can evidence meaningful involvement of families and carers without undermining the person’s rights. Commissioners typically expect clear communication routines, consistent review cycles, and demonstrable impact (reduced complaints, fewer avoidable incidents, improved engagement, better continuity). They will look for audit trails showing how disagreement is managed, how decisions are reviewed, and how learning feeds into delivery.

Regulator / Inspector expectation (CQC)

CQC expectation: Inspectors will look for care that is personalised and safe, with people (and those important to them) involved appropriately. They will test whether staff understand decision-making, capacity, and least restrictive practice, and whether concerns are managed through governance rather than informal promises. They will also review how complaints, incidents, and safeguarding concerns are investigated and how improvements are embedded.

Governance and assurance: keep co-production consistent across the service

To avoid co-production depending on one good keyworker, put simple controls in place:

Standard meeting rhythm: planned reviews (e.g., 6–12 weekly) plus rapid reviews after significant incidents.
Templates that force clarity: decision, options considered, least restrictive rationale, time-limited trial, review date.
Complaint learning loop: themes analysed monthly, actions tracked, and feedback given to families.
Supervision prompts: how staff handled disagreement, how the person’s voice was captured, and whether records match practice.

When these elements are routine, families feel informed rather than excluded, staff feel protected rather than pressured, and the person’s outcomes and rights remain the constant reference point.