Public involvement is not an intervention

Perhaps it’s not surprising – given that we’re all working in the world of health research whose fundamental purpose is to support evidence-based medicine – that we can’t help but think of involvement as an ‘intervention’ and seek ‘evidence’ of its impact. The questions we ask about involvement are the same ones we ask about new treatments:

• Does it work?
• What difference does it make and can we measure that?
• What are the benefits and risks?
• Is it cost-effective?

I think we’re getting stuck with this thinking – because involvement isn’t an intervention. There’s not a precisely defined, standardised action which is public involvement – it’s so varied and complex and highly dependent on context. You can’t just add two patients to a steering committee and predict a specific outcome…

I find it more helpful to think about involvement as a conversation. It’s an exchange of ideas, values, assumptions and experiences between researchers and patients. It’s an ongoing dialogue – an interaction that evolves over time. When you think about it this way, those questions don’t seem so appropriate.

• Does a conversation work?
• What difference does a conversation make? Can we measure that!?!
• What are the risks and benefits of a conversation?
•  Is it cost-effective to have a conversation?

The answers become ‘Well, it depends’. Mostly it depends on why you want to have the conversation and what you hope will come out of it. It depends on what the conversation is about, what gets said, who takes part, where and when it takes place, how it’s done, and importantly what people knew before they started – these are all the factors that influence the impact of involvement. You can’t neatly separate out the outcomes from its purpose, context and the way it’s carried out – these are all inter-related.

So I think we should be asking different questions about involvement, something along the lines of:

• How does it work? What useful things get said?
• What kinds of learning come out of it?
• What makes it go well?
• When and where does it work best

Or, are there other questions?