Evidence alone doesn’t change systems. People do

Too often, research produces interventions that are difficult to scale. This is frustrating for researchers, funders, and critically, those poised to benefit from these interventions. So, how can we bridge the implementation gap?  

The Lancet Psychiatry Commission on Transforming Mental Health Implementation Research challenged mental health research and implementation communities to close the gap, and to support this sector discussion we partnered with the team on a series of symposia. Our aim was to bring the issues and opportunities to life, stimulate discussion and reflection, and explore where there is resonance for action.  

Over the course of the past year, we looked at how systems thinking can transform implementation, why the evidence we generate isn’t always the evidence we need – and how all of this (and more) comes together in decision-making.  

Time and time again, the message was clear: people are at the heart of the system. From a person-centred perspective, it seems like the natural state of the research and implementation system should be to understand problems together, have access to a varied toolkit for evidence generation to probe problems and generate meaningful solutions, and make recommendations to funders and investors about where best to place their limited resources to help create healthier and happier people and communities. When this does happen – and it does – we start to close the implementation gap.  

What are the forces that get in the way? 

Firstly, we hold strong assumptions about what makes for ‘good’ evidence – to the detriment of other forms of evidence and ways of knowing. Speakers discussed that evidence too often means ‘quantitative data generated by experts’, invalidating important sources (such as from people with lived experience) and types of evidence (such as systematic qualitative approaches). Incentives, rewards and funding can too often follow this thinking – and don’t always create space for integrating values, preferences and context alongside data. If the evidence generation system is skewed from the get-go, we miss out on a lot. The panel concluded that a hierarchy of evidence is not helpful when it pits one form of evidence against another – it’s only useful in terms of describing the type of evidence that is needed to answer a particular research question. It’s good to be rigorous, but not to be myopic.  In an ideal world, we should start with the research question and then select the right combination of methods. This is easy to say, but harder to do! 

Second, the system isn’t set up to generate evidence in an integrated way overall – it generally supports siloed thinking. Where integration works well (such as Youth Wellness Hubs Ontario), proactive efforts have been made to embed a collaborative, cross-disciplinary learning culture – thinking about values, influential leadership and supportive governance. We need to create spaces that break down silos and build networks between researchers, implementers, contexts and regions. 

Finally, we don’t always understand the needs of people who are making decisions about evidence and how to communicate with them. It can be too easy to take an approach that says ‘generate the evidence and they will come!’ – but in reality, decision-makers have competing demands and pressures. And they are people too! Through the Friendship Bench project, we heard from how finding the right champion can really move the dial on decision-making – and how stories and connecting with impact can open the door, so the numbers can keep it open. Stories help decision-makers understand what numbers mean for people’s lives and reduce stigma by inspiring hope and possibility – and they have the greatest impact when told in people’s own words, when those people are present in the rooms where policy is shaped. We heard that implementation and system reform is never a one-off event; it’s built through years of shared effort and shared strategy. Our panel discussed that progress depends on embracing the messiness of scale-up as part of learning, not failure. To do this, implementation and research need to meet in the middle – rigour and reality learning together. 

What could funders do?  

Speakers had many ideas. For example: 

• Direct resources to solutions and strategies that are culturally sensitive and contextually relevant and developed by stakeholders who have firsthand knowledge of system challenges.   

• Question what counts as ‘good evidence’ by carefully comparing different ways of doing research, rather than defaulting to the usual methods. Develop funding tools that can respond to innovative approaches to evidence generation and assess whether the methods proposed are the best approach – as well as supporting learning between different research communities.  

• Keep evolving how we give credit to evidence generators – away from traditional incentives & towards the team science behaviours that we know drive impact. 

• Targeted support for transdisciplinary training from early career stages, and funding approaches (e.g. using a mission driven framing, that calls on different skills and experience to rise to the challenge) that requires, inspires and provides support for collaboration. 

• Invest in qualitative research to combine and elevate lived stories, which could add a powerful layer to how we build storytelling as evidence. 

• Build on policy translation mechanisms, storytelling and relationship building – how can implementation and uptake be better embedded from the start? Continued collaboration between government and partners, including making the case for local investment, will be key to lasting impact. 

It has been inspiring to look at the challenges through the lens of examples that buck the trend – examples that show us what is possible and give us ground to build on.  

Turning learnings into impact

This series has helped shape how we at IAMHRF think about the impact of mental health research, and the conditions that are needed to create it. This has already guided us to use a systems-informed approach to designing our recent in-person meeting in Cape Town. 

Creating impact through mental health research and implementation feels like a jazz band uniting rhythm and skill. This arises from players’ individual skill and flexibility, knowing each other, and knowing when to lead and when to follow – so coherence can emerge from true partnership. Our ongoing work focuses on areas where there is a real need to build such coherence – whether that is in funder practice on lived experience, helping connect understanding and functionality across the measurement system, or identifying those sticky areas where sector discussions are needed – will be informed by this principle. 

We encourage you to watch the full series, and continue to let us know your thoughts. Thank you to our incredible speakers, chairs and partners! 

*This series was run in collaboration with The Lancet Psychiatry and two of our members: Grand Challenges Canada and Wellcome.  

Learn more

Symposium 1: Systems matter

• Watch the recording

• Detailed reflections on LinkedIn 

Symposium 2: What do we mean by evidence?

• Watch the recording

• Detailed reflections on LinkedIn 

Symposium 3: Hearts and Minds

• Watch the recording

• Detailed reflections on LinkedIn