For Prospective Collaborators

#Start with the problem and purpose

A potential healthcare collaboration should begin with the need it would address, not with a preferred product or technology. Describe who experiences the problem, where it occurs and what happens under current practice. Separate observed difficulties from assumptions that still need to be checked with affected people.

Clarify whether the proposed work is intended to understand a problem, test feasibility, evaluate an outcome or support routine use. These purposes require different evidence and safeguards. This page offers general evaluation questions; it is not an invitation to a particular programme or a description of an available service.

#Examine evidence and practical fit

Ask what evidence supports the proposed approach and whether it relates to the intended population, setting and task. A promising technical result does not establish clinical benefit or reliable performance in everyday work. Record important gaps, including uncertainty about potential harms, unequal effects and performance under changing conditions.

Consider how the work would fit into existing care processes. Identify the people who would contribute time, the systems that would exchange information and the approvals that may be needed. Include the effort required for training, supervision and evaluation, rather than treating these activities as costs that can be addressed later.

#Agree boundaries before commitments

A useful early discussion makes responsibilities visible. Who can authorise data access, accept operational risk or decide whether the work should stop? Establish how disagreements, unexpected findings and safety concerns would be handled. Do not assume that shared interest means shared authority or agreement about acceptable risk.

Document the proposed scope, exclusions, resources and unanswered questions in a short brief. Agree what evidence would justify the next step and what would prevent it. Early evaluation can reasonably conclude that more preparation is needed, that another approach fits better or that the work should not proceed.

#Common misunderstandings

A shared interest in improving healthcare does not necessarily mean that collaborators share the same priorities. One group may want to test an idea, while another expects a service ready for routine use. Clarifying that difference can prevent a promising conversation from becoming an unclear commitment.

Clinical involvement should not be treated as automatic endorsement. A clinician might advise on a narrow question without supporting every feature, claim or proposed use. Similarly, encouraging feedback is not the same as evidence that an approach improves care. Agreement to explore an idea also does not establish permission to use someone’s name in publicity.

Another misunderstanding is that a successful pilot settles every question. A small test may reveal useful lessons while leaving uncertainty about workload, accessibility or performance in other settings. Collaboration should make those limits visible. Revising, pausing or ending a project can be a responsible outcome, rather than a failure of the relationship.

#Questions worth asking a clinician

• What specific healthcare need would this collaboration address, for whom, and how would we define its intended purpose before choosing a solution?
• What evidence supports the proposed activity, and what further evidence would we need before involving patients or changing care?
• What safeguards would this activity require for patient safety, consent and data privacy, and who would check that they work?
• Who would have authority over clinical and research decisions, and what staff time, funding and expertise would each partner commit?
• What findings or safety concerns would trigger a pause or stop, who could make that decision, and how would affected patients be supported?