Procurement and consent in care settings
A governance checklist for trials in residential care, hospitals, community programs, and supported living.
Products, prices, policies, and evidence change. Verify the current primary sources for the exact model and region before acting.
Define a bounded purpose
Name the population, setting, intended activity, trained facilitator, expected benefit, exclusions, and stop conditions. Do not buy a broad promise such as “reduces loneliness” without defining the outcome and how it will be measured.
Include residents, families, direct-care staff, infection control, accessibility, privacy, information security, procurement, and clinical governance before the trial begins.
Protect consent and alternatives
Consent must be understandable, revisitable, and separable from unrelated care. Record assent and distress, provide an equivalent non-robot activity, and avoid covert monitoring or remote family access.
For fluctuating capacity, establish who can authorize what, when the person's own refusal controls, and how staff will recognize nonverbal discomfort.
Can participation stop immediately?
Is a non-robot alternative available?
Who receives data or summaries?
How are incidents and complaints recorded?
Evaluate before scaling
Use a pre-registered or at least pre-specified evaluation plan. Compare engagement, staff burden, adverse events, equity, maintenance, infection control, downtime, and total cost—not only satisfaction among people who completed the trial.
Framework sources
These sources shape the questions in this guide. They do not certify any listed robot.