Push, pull or co-produce?

There is increasing interest in bringing researchers, service providers and policymakers together in partnerships that seek to improve patient outcomes through the conduct and application of applied health research.[1] In England, this has been promoted by the national funder for health and care research in the form of collaborations that seek to facilitate the use of research knowledge by health organizations and their participation in its production.[2] But understanding how this can be done, and learning from the experience of those who have tried to make an impact in this area, is often given little attention: ‘… evidence is lacking … on the impact … particularly in relation to the knowledge mobilisation processes and practices adopted’.[3]

Why is this so? ‘Research’ is often described as ‘evidence’, but ‘evidence’ is itself a contested term. Within health care, effectiveness research (does X work better than Y?) and its associated evidence hierarchy continues to dominate but is also increasingly challenged. When the ‘intervention’ is complex and interacts with the context in which it is intended to operate, ‘evidence-based medicine’ may be less applicable, although this also depends on the paradigm of those who are considering evidence. The influence of professional training, especially for clinicians, can lead to challenges in accepting alternative views of evidence. Viewing context as ‘a process rather than a place’[4] is a new concept if experience of research has been in controlling out context in order to test effectiveness.

Framing research evidence as being about what you do (X or Y?) and how you do is helpful in considering what is meant by evidence, along with the increasing emphasis on process evaluations alongside intervention studies to help understand the role of context and other variables.[5] Yet, evidence about ‘how’, typically drawing on qualitative research, appears to remain less visible, viewed by researchers as an add on or perceived to lack the same opportunities for peer-reviewed publication available to effectiveness research.[6] It is also questionable whether research about ‘how’ actually gets used in practice or whether it is instead generating academic research that is itself difficult to apply.

Academics are increasingly attempting to ‘push’ research results into practice through the development of (supposedly) innovative dissemination methods such as toolkits, video, etc.[7] Focus on research impact places increasing emphasis on this aspect of research, although this may be contributing to research waste.[8] Viewing non-academics as ‘evidence users’ appears common but may not be helpful, as it reinforces the ‘knowing/doing’ gap.

Implementation research is subject to similar ‘push’ approaches, based on the assumption that it will ‘increase the rate at which research findings are implemented into practice’.[2] Much implementation research is descriptive, however, with models criticized as ‘rudimentary and implicit forms of theory, often reducing complex relationships to prescriptive checklists or stages’.[9] Increased emphasis on the use of theory in implementation science may well increase its rigour, but may not make it more applicable in practice. Research funding and academic infrastructure are not supportive of the long-term development of such research, leading to calls for the ‘research enterprise’ in implementation science to be ‘redesigned’.[10] Despite ‘push’ being the predominant approach among the research community, it is not leading to ‘evidence’ being used in practice.

Few practitioners or organizations successfully ‘pull’ evidence from those who develop it (academics): ‘most health and care organisations aim to base decisions on the best available evidence, but accessing and interpreting the right evidence at the right time is hard’.[11] Even if researchers were to make the evidence available in a timely manner, and in an appropriate format, formal research evidence is only one type of information used in decision-making. Managers also ‘value examples and experience of others, as well as local information and intelligence’.[11] Despite attempts by research funders to be more responsive to health care and service priorities, the timescale to get research funded and then carried out frustrates this aim: ‘having good enough evidence at the right time trumps perfect research which arrives too late for decision makers to use’.[11] Those funding research may need to encourage interim findings which are still robust before study end, although this will challenge existing methods and approaches and involve working in the research ‘middle ground’.12 Another perspective on ‘pull’ is provided by the developing literature on the absorptive capacity of organizations which calls for improved ‘coordination capacity’ if evidence from research is to be used in practice,[13] although this remains largely an academic approach rather than something that can be enacted in practice.

It is argued that co-producing research would be helpful in producing timely evidence. Co-production with decisionmakers is more likely to inform subsequent decisions. There is also a human rights rationale for co-production with the public and service users,[14] but there remain structural challenges in implementing this and, importantly, ‘… the experiential knowledge of service users is rarely afforded equal value to that of scientific/expert knowledge’.[15]

So what can be done despite the structural and funding challenges? I propose some practical steps that can be taken, recognizing, however, that messy reality[15] means these cannot be ‘solutions’:

Have more conversations and interactions with a range of stakeholders outside academia.[15] Academics need to ‘get out more’, and there is great value in shadowing, informal (i.e. non-research) observation and building links. Better understanding of, for example, where and how the data researchers are analysing is generated can be transformative, as they can see first hand the priorities of those generating them.

Have more conversations with other academic disciplines and get out of ‘disciplinary silos’. Funders and researchers rarely draw on learning from different fields, nor is learning shared between disciplines and professions. Reviews of knowledge mobilization approaches in health care have concluded that there is much to learn from other disciplines, specifically management and organization studies.[13] There is a need for support for early career researchers ‘through diverse, cross-disciplinary career pathways’,[12] currently lacking at the institutional level.

Do something together. The most effective collaboration comes when people from different backgrounds work together on something with a shared objective, although this will inevitably involve some compromise on both sides.

Make the most of the research funding that we do have. We have a moral obligation to ensure that research funds invested are not wasted, even if we believe that the current system requires reform. We should do more to ensure that funded research meets practice priorities and challenges, considers implementability from the start and is, as far as possible, co-produced with those who will use it. Through peer review and membership of funding bodies, even individuals can make a difference here.

Stop wasting resources on more sophisticated ways to ‘push’ research findings into practice. Basic good practice is often omitted; asking those who might use evidence how they access information is a simple (and usually ignored) approach, as is using existing professional networks. We have a lot to learn from marketing and communications approaches and can be slow to recognize the value of working with communications professionals. Tailored approaches are more likely to be effective; ‘… researchers need to go to where their audience is, using many platforms’.[11]

We should be cautious about recommending more research on whether such actions make any difference. We need more understanding of what has worked, more learning from others and a more critical approach to the way we generate, select, apply and communicate evidence. We need to get what we already know into practice.