Abstract
Objectives:
There exists a translation gap between academic research and clinical practice in health care systems. One policy-driven initiative to address the translation gap in England are the Collaborations for Leadership in Applied Health Research and Care (CLAHRCs), funded by the National Institute of Health Research (NIHR). These aim to bring together NHS organizations and universities to accelerate the translation of evidence-based innovation into clinical practice. Our aim was to draw out lessons for policy-makers regarding the mobilization of such initiatives.
Methods:
Qualitative semi-structured interviews with 174 participants across nine CLAHRCs plus in-depth case studies across four CLAHRCs. Those interviewed were staff who were central to the CLAHRCs including senior managers and directors, junior and senior academics, and health care practitioners.
Results:
Social positions of the CLAHRC leaders, conceived as institutional entrepreneurs, together with the antecedent conditions for CLAHRC bids, had an impact on the vision for a CLAHRC. The process of envisioning encompassed diagnostic and prognostic framing. Within the envisioning process, the utilization of existing activities and established relationships in the CLAHRC bid influenced early mobilization. However, in some cases, it led to a translational ‘lock in’ towards established models regarding applied research.
Conclusions:
The CLAHRC experiment in England holds important lessons for policy-makers regarding how to address the translation gap. First, policy makers need to consider whether they set out a defined template for translational initiatives or whether variation is encouraged. We might expect a degree of learning from pilot activities within a CLAHRC that allows for greater clarity in the design of subsequent translational initiatives. Second, policy makers and practitioners need to understand the importance of both antecedent conditions and the social position of senior members of a CLAHRC (institutional entrepreneurs) leading the development of a bid. Whilst established and well-known clinical academics are likely to be trusted to lead CLAHRCs, and the presence of pre-existing organizational relationships are important for mobilization, privileging these aspects may constrain more radical change.
Introduction
The gap between academic research and routine practice in health care provision has been recognized as a problem for several decades. 1 This has been characterized as a ‘translation gap’ consisting of two dimensions: ‘the translation of basic and clinical research into ideas and products’, the so-called T1 gap; and concern for ‘introducing those ideas and products into clinical practice’, the so-called T2 gap. 2 There have been a plethora of initiatives to bridge the gap: in the United States, the development of Veterans’ Health Administration’s Integrated Health and Research System 3 ; in Canada, Clinical Translational Science Centres 4 and the Canadian Health Services Research Foundation. 5 England has followed suit with investment in Biomedical Research Centres (BRCs), Biomedical Research Units (BRUs), Academic Health Sciences Centres (AHSCs) and Academic Health Science Networks (AHSNs), which bring together universities and health care organizations. 6 Our focus within England lies with Collaborations for Leadership in Applied Health Research and Care (CLAHRCs), nine of which were established in 2008 with £100million funding from the National Institute of Health Research (NIHR) for five years (with a similar amount of matched funding from participating higher education institutions [HEIs] and health care organizations).
CLAHRCs have three key interlinked functions: conducting high-quality applied health research; implementing the findings from research in clinical practice; and increasing the capacity of NHS organizations to engage with and apply research. CLAHRCs are regional with their agendas determined by the partnering organizations and tailored to health care needs in their areas. CLAHRCs were regarded by the NIHR as experimental with considerable variation allowed for their structures and processes. Social sciences were variably integrated into CLAHRC plans, with some involving business school input. Academic research and clinical practice were blended in different ways. There were differences in the disease emphasis of CLAHRCs and the clinical disciplines involved (medical and non-medical), although all CLAHRCs focused on translational research in long-term conditions.
Our aim was to draw out lessons for policy-makers, regarding mobilization of initiatives designed to address the translation gap. This is particularly timely as CLAHRCs reach the end of their first tranche of funding, with a new commissioning round in 2013. Our intention was to identify possibilities and limitations of different translational models of evidence-based innovation.
Literature review
Pettigrew et al. 7 highlighted the importance of context and receptivity for the type of strategic change in health care systems envisaged by policy-makers in funding CLAHRCs. An important point they raised is that much research on strategic change is ahistorical. To understand possibilities for future strategic change, we need to understand past patterns of change and stability including consideration of the local as well as the national context. In analysing strategic change within comparative cases of NHS organizations, Pettigrew et al. identified receptive conditions for change. Linked to this, they showed how change is shaped by the antecedent conditions that frame the way partner organizations and individual leaders move forward as they conceive of, and progress, change. So, first, we need to examine antecedent conditions, or ‘local field characteristics’ for CLAHRC bid development amongst the key partner organizations and leaders. 8
Second, within analysis of strategic change in health care by Pettigrew et al., 7 context is presented as somewhat malleable in terms of the way that change recipients perceive and act on it. Linked to this, the importance of lead actors and their disposition towards change is emphasized, as they make sense of context for others to engender the necessary receptivity for strategic change. We extend insights about the importance of key leaders in strategic change revealed within comparative cases studied by Pettigrew et al. through drawing upon the concept of ‘institutional entrepreneurship’. 8 Specifically, we cast senior CLAHRC leaders as institutional entrepreneurs (IEs) and examine how their social position in the structure of translational interventions has an impact on early strategic decisions in bid development. Although organizational fields are in some sense structured by institutional forces, struggles still occur between different stakeholders in relation to resources and social action, and these have the capacity to recreate, even change, institutionalized practices. Institutional entrepreneurship focuses on the nature of these struggles and how actors seek to influence existing and emerging institutional configurations.9–12 We define IEs as those who can envision new ways of working as a means of advancing ‘interests’ that may be suppressed within prevailing or emerging organizational practices. 9 In other words, they are people who seek to initiate and enact institutional change. However, we highlight that translational initiatives in health care are complex, resulting in important differences in opinions, interests and power across different individuals and groups operating at the sub-organizational level, where IEs work to engender different forms of institutional change that promote interests they value highly, and which are derived from their current social positions in the structural context.8,13
The founding conditions, in terms of antecedents to CLAHRC bids and the social position of CLAHRC leaders are important. Social position represents someone’s position in the structure of social networks. 14 Each person’s social position determines the set of people to whom they are directly linked, which, in turn, affects their perception of their organizational field. Over time, actors’ social positions become internalized, promoting the development of shared subjectivities and cultures. 15 Together with antecedent conditions, actors’ shared subjectivities and culture derived from their social position provide the ‘cognitive context’ 16 through which they make sense of the world in responding to the call for CLAHRCs.
Envisioning has been highlighted as the important first stage in any change process in which actors form an ‘embryonic’ vision, based on the interplay between themselves and the context in which they are situated. 17 Individuals frame problems and propose changes in routines which are grounded in their perspectives as members of distinct and different professional communities. 16 Frameworks, or ‘schemata of interpretation’, allow the individual to derive meaning from what would otherwise be meaningless, to locate, perceive, identify and label events they experience. Envisioning encompasses two main activities: diagnostic framing and prognostic framing. Diagnostic framing is the process through which actors identify and diagnose the source of the problem affecting current practice; and prognostic framing is the potential solutions that are generated to solve or minimize the.18,19 There tends to be a correspondence between an issue’s diagnostic and prognostic framings, meaning that the identification of specific problems and causes tends to constrain the range of possible reasonable solutions and strategies advocated. In this way, we suggest that there is a generative rather than deterministic relationship between diagnostic and prognostic framing. 20 We highlight in connection with this that founding conditions (social position and antecedent context) shape an actor’s propensity to engage in institutional entrepreneurship for change and note that some actors may be pre-disposed to perpetuate the status quo from the beginning. 21
In summary, our proposition is that the social position of IEs in their structural context, alongside antecedent conditions for change and stability, engender ‘lock-in’ to the translational initiative towards distinctive ways of conceiving and mobilizing CLAHRCs, evident in the diagnostic and prognostic framing within the envisioning process. This shapes eventual outcomes of CLAHRC in terms of academic research, implementation of academic research, and applied research capacity building. As one respondent, who managed a clinical theme within the CLAHRC commented, ‘In developing the bid, the CLAHRC was pretty much set in stone, and that was difficult to change later on’.
Methods
We adopted a longitudinal strategy across comparative cases, which encompassed interview data, archival data and field notes from observations from 2009 to 2013. We undertook 174 qualitative interviews during the mobilization phase of CLAHRCs and reviewed bid documents developed prior to the launch of CLAHRCS in 2008. In the first phase, we undertook 104 interviews across all nine CLAHRCs. In a second phase, we collected data across in-depth comparative cases (CLAHRCs B, C, D, I), within which a further 70 interviews were carried out. These second phase in-depth studies were chosen along the following criteria: variation in antecedent conditions (e.g. site based on existing translational activity or ‘greenfield’ site), and the social position of the CLAHRC Director (NHS or HEI as main employer; clinical or social scientist).
Staff central to the initiatives were interviewed, including: CLAHRC senior managers (both HEI and NHS employees, clinical and social scientists); theme leads (clinical scientists and social scientists); other NHS staff involved in CLAHRC, such as senior NHS managers represented on the CLAHRC Boards and clinicians part bought out to work into CLAHRC studies. The number of interviews at each CLAHRC differed: A (6); B (37); C (28); D (32); E (18); F (10); G (28); H (12); I (3).
The focus of the interviews was broad encompassing lead actors’ backgrounds, disposition towards change, and vision for CLAHRCs during bid development. Interviews were semi-structured, openly recorded and lasted between 45 min and 2 h each. All interviews were fully transcribed. Interviewing stopped when we reached a point of theoretical saturation. 22
We also carried out observational work, involving: attendance at key meetings; workshops; presentations; other educational events. At educational events, for example, one or two of the research team would present feedback and facilitate discussion, with one other of the research team not participating, but observing and taking notes, focused upon CLAHRC vision and the background of participants. The research team attended three CLAHRC senior managers’ meetings of around half day in length, facilitated two cross-CLAHRC educational events, and gave individualized feedback to four of the CLAHRCs on at least two occasions each. During all meetings and observations, detailed notes were taken.
Finally, documents were collected, such as initial CLAHRC bids; annual reports; study protocols; corporate publicity material; minutes of operational and CLAHRC Board meetings, to develop ‘a substantial archival residue’ from the different published sources. 23 All material was collated into a case study database for each CLAHRC.
We began with a fine-grained reading of each CLAHRC case, 22 initially coding the data on a within case basis. 24 After inductively creating a list of first-order codes from the case evidence, we consolidated all of our codes across the cases through axial coding, structuring the data into second-order concepts and more general aggregate dimensions.22,25 In doing so, we engaged in deductive reasoning, whereby we linked our inductive codes with existing concepts and frameworks. 26 Analysis was conducted with the assistance of N-Vivo 8 with material inputted into a database and coded according to each high level category.
Results
We identified three inter-related themes: social position of IEs; antecedent conditions for CLAHRC bids; and the envisioning of the CLAHRCs. In presenting primary data, all respondents were significantly involved in CLAHRC bid development and early mobilization. We distinguish respondents by identifying their social position in the field in terms of: academic discipline where their primary employer is a HEI (clinical science or social science); CLAHRC administrative managers; NHS senior managers (commonly on the CLAHRC board); and other NHS employees such as clinical academics and practitioners. We have avoided identification of individuals or CLAHRCs. Before considering the social position of IEs and antecedent conditions for CLAHRC bids, we suggest that the shape of CLAHRC bids during the envisioning processes was likely to prove somewhat malleable on the basis that most of our respondents commented on the lack of clarity in their CLAHRC’s mission in the initial tender document:
All CLAHRCs have interpreted the brief in very different ways. If I was being generous, I might see that as a deliberate strategy to test different models. However, it may be that there was just not a well and clearly-defined brief for what CLAHRCs were meant to do (CLAHRC administrative manager 1, CLAHRC I).
Social position of institutional entrepreneurs
In general terms, actors whose dominant orientation was towards the NHS, were more practice facing while those in HEIs, more research facing. The latter dominated CLAHRC senior management positions. Although they may hold dual appointments, the requirements of producing world class publications and raising research income orients their behaviours accordingly. For such actors, there was a clear incentive to get involved:
I always had a post that straddled the university sector and the NHS sector. That was my job here and I ran the research here at a teaching hospital and quite simply the call came out for academic health science centres of the future as it was called […] and I just printed it out and read it and I knew without a doubt that it was for [this local region]. (Clinical science academic 1, CLAHRC F)
For clinical academics, the CLAHRC presented a clear opportunity for research funding that was aligned to their current activities. Consequently, there was a significant degree of opportunism for some clinical academics, as CLAHRCs were a suitable vehicle to fund existing work. CLAHRC I was led by a clinical academic whose research interests were mirrored in the way the CLAHRC was set up and the research carried out:
We were briefed about that and obviously it is a large grant so it was very important. I was already involved in a number of NIHR initiatives, such as the diabetes research network and we also had a programme grant with the NIHR, so I was aware of the NIHR funding stream and I was aware of the spectrum looking for applied research. (Clinical science academic 1, CLAHRC I)
CLAHRC directors that were more NHS facing engaged with the CLAHRC initiative from a different social position. For example, one with a track record of managing and delivering change management programmes in the NHS, articulated his/her reason for getting involved as follows:
Why did I get involved? Well I am a clinician by background. I have done a lot of ‘change’ in health care work, change at a variety of levels of the years including two national change programmes, and they both suffered from the fact that evaluation wasn’t built into the programmes from the beginning despite everyone’s best efforts because actually there was no funding allocated to do it. (Clinical science academic 1, NHS employed, CLAHRC D)
The focus was on using the CLAHRC as a means of engaging in change management in the NHS but doing so with a mandate to evaluate what was occurring and learn from the experience. The CLAHRC lead viewed research as important but research was a means to an end:
There is no RCT to say that we should be doing acute medicine, there is no RCT to say we should have coronary care or emergency medicine and we have them, so medicine doesn’t always run according to RCTs. In my line of work we use RCTs when we want them, you know, to stop something happening or to start something happening. RCTs we definitely need, but I just don’t think an RCT is the only way, it can’t be the only solution and it isn’t the way change happens in all of the rest of the real world, you know Microsoft, IBM, you name it, they don’t do an RCT and then decide that that is the way forward. They may use evidence but then they decide what to implement and they monitor the implementation. (Clinical science academic 1, NHS employed, CLAHRC D)
Finally, in parallel with many of the more HEI facing actors, CLAHRCs presented an interesting opportunity to align existing interests with funding:
Before the CLAHRC bids came along I was already working with XX to try and set up a centre for health care improvement and delivery … then the bid came up so we said, ‘well that is what we wanted to do; how do we make what we wanted to do meet the needs of this bid?’ (Clinical science academic 1, NHS employed, CLAHRC D)
This highlights the importance of the antecedent conditions that framed the way bids developed.
Antecedent conditions for a CLAHRC bid
The local field conditions helped shape the way in which CLAHRCs were founded. Pre-existing relationships between NHS organizations, HEIs and other organizations meant that those who were involved in putting the bid together could use these links and any existing projects as part of the bid:
Like any CLAHRC, I don’t know how much of it is about pragmatism in the sense that we had existing research expertise. […] So there is a combination of that which is pragmatic, it fits, it is a worthy candidate for this resource, but the resource isn’t likely to be wasted because it is building on a heritage and foundation of good quality science and that is important. It is also a strategic view as well on the part of [director] and a senior group of influential pro-vice chancellors, the sort of people involved in looking at what the areas are that are likely to be a good showcase for our academic expertise and the clinical networks that are out there, what are the areas that are most likely to benefit from the injection of capital funds. The CLAHRC is a well-resourced beast and it is a good opportunity and I think we all knew it is probably one of the last opportunities we are going to get for a few years on that scale (Social science academic 1, CLAHRC H)
In a number of regions, there were strong pre-existing relationships that spanned the HEIs, encompassing different departments, and local NHS organizations. This enabled actors to quickly mobilize other individuals and to coordinate existing activity, behind their bid, thereby enabling the CLAHRC to ‘hit the ground running’:
In terms of research, what we did was, first, we found that there were people with similar research interests that ran across the NHS and the university that weren’t represented by a department. Second, we form an institute of health sciences, a research network. Then this network meant that when the bid came along to form a CLAHRC, the opportunity was there because we had a readymade network and we were already working in partnership with the NHS and we already had clinical themes that extended right across [our region] … all I needed to do was scream ‘give it to us’ and that is exactly what I did. (Clinical science academic 1, CLAHRC C)
In one CLAHRC, such antecedents included strong service user networks. This might be unusual compared to other CLAHRCs. Although all CLAHRCs claim strong user networks in their bids, they did not commonly privilege this in their interviews:
When we set up our original research unit pre-CLAHRC, it included people interested in involving patients in research design. We also had a very longstanding patient organization that was well-funded. The net result is that patient public involvement is not as new in our CLAHRC as others (Clinical science academic 1, CLAHRC G)
Generally, pre-existing relationships were considered to be helpful in the formation of CLAHRCS, particularly where they spanned HEI and NHS organizations. Specifically, the thematic focus of the different CLAHRCs was shaped, in many cases, by current activity. CLAHRC funding sometimes was viewed as enabling failed research bids to be funded and the continuation of existing research:
We started with what we have got on the shelf, which might have been previously rejected by funders. Then we looked at what are we good at. So we were looking for well-developed projects. And [one clinical academic] was working in this area, had the programme grant that had not been funded but had received very good feedback […] So that is where we started with the projects and how we ‘chose’ what got into the CLAHRC projects. (Clinical science academic 1, CLAHRC B)
Whilst perceived as necessary to mobilize engagement for the CLAHRC bid in a ‘short space of time’, the presence of existing activity and networks potentially acted as ‘cognitive constraint’ on some actors in terms of how they conceived the CLAHRC initiative. Specifically, basing the CLAHRC on existing activity ran the risk of actors interpreting the CLAHRC as simply a large research grant:
So the CLAHRC partly feels a bit like a programme grant in that we’re allowed to sort of do that development work, they’re all these sorts of straight RCTs. So I don’t think that was alien for us. (Clinical science academic 2, CLAHRC C)
Where the CLAHRC was less focused on existing activity and relationships, there was an interesting effect on how it was conceived. There was less consideration about how the different elements of existing activity could be integrated into a coherent package, rather the question was framed more in terms of, ‘what do we need to do?’:
The CLAHRC came along as an opportunity and we all tossed around a fair number of ideas at that time about what we would do, so we are almost going back to where we were before CLAHRC and saying ‘well what do we now need to do?’ (Clinical science academic 1, NHS employed, CLAHRC D)
The lack of established relationships and activity presented the senior actors in this region with a blank sheet of paper in envisioning their model of a CLAHRC. However, this CLAHRC also utilized and built on existing networks and relationships. Once funded, the model was one of trying to embed into existing activity in the region. The important difference here is that this is something that was undertaken after funding was secure:
So when the money came in, there wasn’t actually an application process for round one, it was more a case of being opportunistic and almost latching ourselves on to pre-existing research projects and trying to give them a quality improvement slant. So it would have been based a lot on [the director’s] connections for example and you know his expertise […] so we said right, okay, well let’s run a number of projects and let’s try and connect them up (CLAHRC administrative manager 1, CLAHRC D).
Envisioning
We found significant variation in the way in which the different actors envisioned their CLAHRCs both in terms of diagnostic and prognostic framing. There were a number of stakeholders to whom the CLAHRC might be accountable and, dependent upon their social position in the structural context, CLAHRC leaders privileged one or the other. So, for example, those CLAHRC leaders located in the NHS were less likely to be driven by the demands of clinical academics regarding the scoping of research and more likely to consult NHS managers and practising clinicians:
Well I think that one of the things we discussed at the time out is that absolutely we are accountable to the NIHR, you know, there is a lot of money and you have got to have accountability and you have got to report to them, but it is almost equally important to our local partners because at the end of the day, if we tick every box for the NIHR, but our local NHS partners aren’t happy, then we have failed completely because the CLAHRC is set up to leave that legacy and that infrastructure. (CLAHRC administrative manager 1, NHS employed, CLAHRC H)
Diagnostic framing
A first dimension of diagnostic framing was the nature of the problem itself. Rather than starting with conducting more research that could be better translated into practice, it was important to place a greater emphasis on using existing evidence and adopting a problem solving approach to clinical questions. The problem of a lack of receptivity of research was voiced by clinical academics, particularly those with a non-medical background such as a therapist:
We know from lots of literature reviews and surveys that therapists don’t read anything other than their own professional journal. … they don’t like looking anywhere else for any evidence. I am always banging on about reviewing the literature outside your own profession to inform your practice, but that doesn’t happen and it still doesn’t happen. So there is a delayed research culture in the therapy professions, but it is getting better. (Clinical science academic 1, CLAHRC E)
Linked to this, some CLAHRC leaders were keen to emphasize the problem was translational in nature:
I think CLAHRCs are for is not necessarily what all the other senior academics in this CLAHRC or other CLAHRCs that I have spoken to think they are for. Some senior academics I know that are involved with CLAHRC seem to think they are rather like programme grants and they get down to their theme and they think ‘I am just going to do it because it is applied health research and it is good for the NHS’. I think, myself, my reading of the CLAHRC is that themes ought to all be replaceable because they are vehicles for doing something else, the research is a vehicle for doing something else. (Clinical science academic 1, CLAHRC H)
However, some people interpreted the CLAHRC mission in a different fashion:
CLAHRCs are a product of backroom decision making where the flavour of the day was the ‘translation of evidence into practice’ is the problem. It may, or may not be. However, the major problem facing our health care system is variation; the way people access and receive health care. So it may not be an issue of translation. (Social science academic 1, CLAHRC E)
Reinforcing our argument that an actor’s social position in the structural context influenced their diagnosis of the problem, CLAHRC was set up to solve, another NHS clinical science academic commented:
We, as a CLAHRC, sought to address huge health inequalities and issues around ethnicity … we were looking at how we might impact commissioning practice. (Clinical science academic 2, NHS employed, CLAHRC H).
A second dimension of variation was around who defines the problem. In essence, the problem can be defined by a number of different actors ranging from the end user of the service, to the person who is delivering the service, the commissioner of the service and others. To some extent, not only the organizational affiliation but clinical discipline had an impact on envisioning the CLAHRC mission. For example, an academic with a therapy background viewed the CLAHRC mission as follows:
I think the services we have are very caring and would do anything, but I think we just don’t have that culture and I guess what I see myself as doing is trying to sow that seed of having that kind of culture in practice so that therapists actually understand about putting this evidence in practice and understand that we need to include so many people and we need to take the evidence and implement it in our work and getting it working and accepting that there will always be limitations. (Clinical science academic 1, CLAHRC E)
Many CLAHRC actors highlighted the issue of problem definition as key to being able to more easily translate the findings of research into practice through the development of a more engaged and receptive audience for the knowledge. Also, by engaging practitioners and users in problem definition, researchers will be better able to work research knowledge into the agenda of practice, enhancing the relevance of their work to practice. In this regard, the issue of who defines the problem neatly aligns with the more translational diagnostic frame highlighted above. However, inclusion of service users in particular was variable in statements about diagnostic framing.
Patients and carers should be more involved in determining the research agenda and conducting the research the academics carry out in collaboration with their partners so it is more relevant, and so there is greater uptake and engagement at the end. (Clinical science academic 2, CLAHRC G)
For others, the issue of who defines the problem challenged some of the assumptions of the knowledge translation model. A number of actors argued that there is a degree of commonality of what is being done across the research and practice divide, but that what is missing is an agreement of the fundamental problem that is the focus of our attention. One CLAHRC director summed this up when he argued that:
Knowledge translation is a really bad idea. In my view the formulation of the problem is the problem. Start with the problem and work backwards. […] Telling them (clinicians) about the research is a pretty kind of thankless task. I don’t like the phrase knowledge translation, because what is needed is not knowledge translation, it’s about starting with the person who is going to make the decision and then working with them to understand how they can narrow the bounds of their uncertainties by using what is out there. (Clinical science academic 1, CLAHRC G)
A similar position was evident in another CLAHRC, where the term knowledge translation was also disliked. Rather than focus on translating knowledge, they argued that it is important to focus upon generating user relevant knowledge though the process of implementation.
I would start by disagreeing with the concept of knowledge implementation/translation, I guess the whole of our CLAHRC approach challenges that. It’s about creating ‘knowledge through doing’. We can only learn what works in the real world through doing things. And that doesn’t mean you can’t take knowledge from elsewhere, but you need to assess how valid is that evidence or how real is that evidence (CLAHRC administrative manager 1, CLAHRC D)
However, as was evident in the empirical section about social position of IEs, the dominant actors in defining the problem were those clinical academics located in HEIs, who conceived a need for more applied research in the more traditional academic paradigm. Inclusion of NHS practitioners and service users was much less emphasized when defining the problem that CLAHRCs were meant to solve.
The third dimension of diagnostic framing was that of whether or not the problem lies at the level of the individual or at a more organizational or systems level. In diagnostic framing, dimensions of change related to leadership and capacity building seemed relatively absent.
The commissioning brief talked about making NHS organizations more conducive to evidence-based innovation but this was vague and confusing for some CLAHRCs. CLAHRCs were essentially led by clinical academics. They reverted to type. They write and run projects for academic outputs. (CLAHRC administrative manager, CLAHRC I)
In contrast, for a number of actors who were social scientists, changing individual behaviours was important but they framed the problem as one of sustaining behavioural change through changes to the organization and system:
The problem that CLAHRCs are there to solve is a system level problem. The danger is that we could spend all of our time convincing academics to move closer to practice, and for practitioners to engage more in research, but that will all be for nothing if we cannot embed the changes at a system level. Without system level change we may able to ‘turn a few heads’ and get people to ‘act in a CLAHRC type way’, while we are resourcing CLAHRCs. But what happens when the money stops? (Social science academic 1, CLAHRC C)
The latter view was shared by one of the NHS CEOs, chairing a CLAHRC:
This is not about academics just getting their research grants. It’s about leadership and CLAHRC reaching the parts that we don’t normally reach (NHS senior manager, CLAHRC E).
However, despite the second letter of the acronym CLAHRC highlighting ‘Leadership’, there was little in the accounts of IEs about the CLAHRC’s mission to develop leadership around evidence-based innovation.
Prognostic framing
The prognosis from some CLAHRC lead actors was that research was matched to local needs:
When the call for CLAHRCs was announced my vision was to create a platform for world class health services research that had its application in the [local] health economy (Clinical science academic 2, NHS employed, CLAHRC H).
All of the CLAHRC senior actors were aware of the importance of tailoring their activity to the needs of local NHS organizations and their staff, users and carers:
It is trying to achieve the closing of the second translation gap based on the hypothesis that … you need to engage the clinicians delivering health services, and patients and carers receiving the health services, in determining the research agenda. If so, the research the academics carry out in collaboration with their partners is seen as more relevant, and clinicians and patients have more ownership of the research findings and so are more likely to engage with them at the end of the process. It’s greater engagement throughout leading to greater uptake and engagement at the end. (Clinical science academic 2, CLAHRC G)
In some CLAHRCs, the local need was not only defined by local NHS organizations but also the end user/carer:
We have a very strong ethos, our principles are around capacity development and they are around public patient engagement, so we have a strategy that crosses over, so they are very cross cutting. (Social sciences academic 1, CLAHRC F)
However, the manner in which this came through their prognostic frames was shaped by the local context and lead actors’ social positions in their structural context. In many CLAHRCs, the focus was largely determined by the research strengths (including engagement and implementation) of the HEI partner:
Basically if I am very honest, it was going to be an extension of existing research before we run out of funding. We had a clear chance really that had arisen from our previous research that then constituted our research theme. (Clinical science academic 3, CLAHRC C)
This was for a number of reasons. First, that research would translate into practice best where there was a mass of high calibre researchers. Second, that for the purposes of getting past the commissioners of the CLAHRCs, there would be a need to ‘go in with the research superstars’ in order to have legitimacy with NIHR:
In many ways we arrived at a marriage of convenience, through which we co-produced a research agenda for the CLAHRC in around 3 months, from start to finish. Other CLAHRCs were far more fortunate in that they were less time constrained and so could spend longer in aligning the interests of research and practice. Although we were working to a very tight time schedule, at least we could ‘hit the ground running’ because we already had the research in place (Social science academic 1, CLAHRC E).
In contrast, two CLAHRCs started with a blank sheet of paper:
From that perspective it actually gave us a lot of freedom to separate from the pull of ‘oh we need more research money or we need to make up for the loss of the clear funding’, which was a big challenge at the time. So we were completely free from that some extent to be creative and to really focus on what we saw as the main call for the programme around how do you implement research into everyday practice. Rather than thinking we’ve got lots of research teams that we need to bring more funding in for, we can start with practice defined problems. (CLAHRC administrative manager 1, CLAHRC D)
Similar to marrying research to local needs, some actors focused on tailoring their efforts on examining efforts for improving services in their local NHS organizations. There were two main methods of doing so. One CLAHRC aligned its activity with practice through the use of a quality improvement methodology, seeking to improve the quality of service first and foremost, and to look how research could be employed in the process of doing so:
Well I suppose the guiding philosophy for the whole CLAHRC is that we actually want to make a difference. This isn’t just research, so in terms of what we are about it should be about making it better for patients and then making it better for staff. I know that sounds a tad grandiose, but the reality is that is kind of the premise, so we always try and bring it back to what are we trying to do and we are actually trying to make it better for patients. (Clinical science academic, NHS employed, CLAHRC D)
The prognostic frame was based on the assumption that the research should not be divorced from the ‘doing’:
An implementation project which is using the evidence, but recognizing that what we don’t know is the best way of implementing it in a specific setting … which is the whole point is second translational gap … Therefore, we aim to generate new evidence on how you implement things and how you do things in the real world. We’re not generating pure new evidence for this translational gap. (CLAHRC administrative manager 2, CLAHRC D)
The idea was to generate new evidence about how to implement, for which a number of different forms of evidence needed to be generated:
For evidence-based implementation, we have a concept that there are three types of evidence that you need to be looking at. First, there’s evidence of what to implement, which often comes from previous controlled trials. Second, there’s evidence of how to implement, which we know a lot about and leadership once it’s working, ownership, measurement for improvement and those types of things. Third, there’s evidence of what happens when you implement things. … We start essentially using the three steps of the model for improvement; what is it that you’re trying to achieve? How will you know that a change is an improvement? And, what changes can you make or test to see if they make you a desired improvement? And so it’s recognizing it’s not about solutions, it’s about ideas. Because of the complexity of the world, you can never have a solution until, you can have ideas about what might work, they might have a lot of evidence behind them but until you’ve gone and tested them in the real world you don’t know. (CLAHRC administrative manager 1, CLAHRC D)
The management of this approach involved asking NHS actors in the locality to engage with the CLAHRC to help support innovation initiatives, to support them in doing so and to research the process on an on-going basis.
The second approach differed in that it did not employ a strict quality improvement methodology. Rather than seeking new service innovation projects, it sought to contribute to existing initiatives through a process of establishing what the important service innovation projects were in the local region, selecting the larger more organizationally focussed projects, and performing an on-going formative and summative evaluation of the project:
It took me a long time to understand what was going on in the XXX CLAHRC, and the other day the penny finally dropped. He’s a genius. Rather than scrabbling around for bits of matched funding here and there, he’s taken a bigger picture perspective. He’s said to the local NHS organizations, ‘what are the big service innovation projects that you have at the moment? I’ll get the CLAHRC to work into them to provide a real time evaluation.’ You can’t get any more embedded than that can you? And it deals with the issue of matched funding all in one. Genius!’ (Social science academic 1, CLAHRC E)
The advantage of this approach was that it enabled the researchers to be embedded into the local NHS organizations’ activities from day one. It was more than ‘marrying activity’ but about ‘working into’ activity.
Prognostic framing involved decisions around the allocation of funding between research and implementation activities, and the relationship between the two. The initial guidelines which set out what CLAHRCs were to achieve were interpreted by some as being very flexible, to the point of being rather unclear:
I think it was partly challenging because the CLAHRC bidding process set out a framework but there was huge room for manoeuvre within it, there was huge scope for flexibility and thinking about where the research sits and where the innovation sits and where is the balance between those two things, also in looking at what the areas of public health need were and so on and so forth, all that sort of stuff. (Clinical science academic 2, NHS employed, CLAHRC H)
The bid formation process differed across CLAHRCs and although the NIHR allocated 80% of the funding to research and 20% to implementation, CLAHRC directors took different approaches in how they allocated the money and to which projects. Many emphasized research more than implementation because of the need for scientific credibility in the eyes of the NIHR, who would make the funding decision. As outlined earlier, this was driven by the founding conditions for a number of the CLAHRCs, which meant that much of the funding was already allocated to clinical academics in attracting them to engage with the CLAHRCs. In addition to the founding conditions, a number of more practice facing actors identified that the actual form had a bias towards research commenting that:
It was very interesting completing the application form. We found a tension between the research part and the implementation part. In the implementation sections, there was nowhere to include references, whereas the research section there was. The implicit assumption was that we know how to do implementation. (CLAHRC administrative manager 1, NHS employed, CLAHRC D)
The lack of a space for references was seen as an indication that the NIHR’s assumption was that we know how to do implementation and that the challenge was to do more clinical research.
For some actors, the solution was to integrate the research and implementation into the different research themes. In contrast, for others, the solution was to run the two activities in parallel. Framing the CLAHRC as research and implementation, running in parallel, created tension regarding time horizons:
I think there was probably a bit of a misunderstanding about what applied clinical research is and what the translational gap is. It took quite a long time for us to be physically clear about how we were going to link in to implementation and what that meant about how you transfer knowledge and how you change culture in terms of how people look at research in terms of how it changes practice. [….] A lot of my clinical academic colleagues who probably aren’t particularly clear about what this sort of research is, and possibly in the past they have seen it as soft and perhaps not as important research. (Clinical science academic 2, CLAHRC I)
In terms of prognostic frames, there was also real confusion as to what the remit of CLAHRCs should be. One clinical academic succinctly made the point:
What the CLAHRC did was to give us a narrative, or a conversation. To begin with I think it completely puzzled people. All the agencies that had signed up to the creation of this thing then looked at their baby and thought: ‘What the heck is that?’ We have found over the first couple of years that as the narrative has been rehearsed in different audiences, they have come back and explained to us what that story actually has to be about. I would say that now going out and explaining to people what a CLAHRC is, has become possible. At the beginning it was impossible. (Clinical science academic 3, CLAHRC G)
Discussion
At bid development stage, different stakeholders conceived the translational problem in different ways, which shaped how the CLAHRCs were operationalized. Differences in bid development encompassed: the balance of activity between research and implementation; whether research should be clinically or implementation focused; appropriate metrics for CLAHRCs; whether CLAHRCs should orientate towards their academic or NHS partners; and whether CLAHRCs should focus upon individual behavioural or organizational/system level change. This gave rise to variation between CLAHRCs and different outcomes in terms of conducting high-quality applied health research, implementing the findings from research in clinical practice and increasing the capacity of NHS organizations to engage with and apply research. This variation in CLAHRCs is, in part, derived from the social position in the structural context of those IEs leading their development and was promoted by the lack of clarity in the CLAHRC tender document.
Given the malleability of the CLAHRC mission presented in the original tender document, there was variation in CLAHRC envisioning. Lead CLAHRC actors, conceived as IEs, framed the opportunity derived from CLAHRC funding at different points on the research-implementation spectrum. The dominant orientation, given the majority of CLAHRC lead actors were primarily employed within HEIs, was towards research publications and research income opportunities arising from CLAHRCs. In some cases, the research approach dominant in clinical science, that of the RCT, underpinned research plans. To a lesser extent, and particularly encompassing those primarily employed by the NHS, some CLAHRC lead actors were more orientated towards service improvement.
Who defines the problems and solutions related to translational activity was a key dimension of the innovation ‘lock in’, and whether the bid development started with a capability view of research expertise in the HEI or a review of service problems. These represent two ends of a spectrum across which CLAHRCs were positioned. Generally, lock in was framed by a clinical academic/HEI perspective upon the envisioning of CLAHRCs. However, in mitigation, those developing CLAHRC bids felt constrained by the short time from announcement of potential funding and submission of the bid. For the bid to prove successful, they also felt a need to set out detailed plans so they ‘hit the ground running’. This mitigated the opportunity to engage the NHS in co-producing a research agenda closer to service needs.
The third aim of CLAHRCs around capacity development was little reported upon. Rather than the development of evidence savvy clinicians, or implementation savvy clinical academics, research or service improvement outcomes were privileged. Nor was major emphasis put upon development of leaders (the ‘L’ of the CLAHRC acronym) for evidence-based innovation. Arguably, the development of capacity is implicated in service improvement. However, generally, capacity development was not a major theme in the envisioning of CLAHRCs. Initial conceptions of CLAHRCs appear higher on opportunism for immediate research, and to some extent service improvements, than a long-term view of the sustainability of evidence-based innovation. There were, however, exceptions with a NHS senior manager, supported by some social scientists, emphasizing organizational and system level development. Also, clinical academics exhibited variation in their envisioning of CLAHRCs dependent upon their disciplinary background, as in the case of the therapist who viewed CLAHRC as a vehicle for ‘other things’. However, this suggestion is tentative, as it was not examined in detail.
In their interplay with the social position of IEs leading CLAHRC bids, antecedent conditions within HEIs in particular, shaped conceptions of and activities encompassed within, CLAHRCs. Indeed, in some cases, CLAHRC was seen as a vehicle for progressing research bids that had failed to be funded. In others, CLAHRC was seen as a way of wrapping up existing research activity akin to a large scale programme grant. CLAHRC activity appeared more academic-defined than NHS-defined. In part, this model aligned with the view of those leading CLAHRC bid development that the NIHR privileged research excellence in its commissioning objectives. On the one hand, we might question some CLAHRC bids as offering little more than pre-existing activity. On the other hand, CLAHRC funding allowed pre-existing activity to be ramped up, leveraged, and potentially institutionalized, for a greater gain than was possible in the absence of support for large scale activity, and with a particular translational edge that was more service orientated.
Pre-existing research activity is likely to prove advantageous to mobilizing CLAHRCs. Two of the CLAHRCs characterized their bid as more ‘greenfield’ than ‘brownfield’ activity and claimed greater innovation as a consequence. On the one hand, such CLAHRC bids might sidestep translational lock in to established ways of doing things. On the other hand, such CLAHRCs may prove more difficult to mobilize in the face of the ‘greenfield’ context.
As a further dimension of the shape of CLAHRC bids, where bids were orientated towards service improvement, the conception of the level of change required varied in the extent to which it was individual behavioural change or wider-scale organizational and system change. Antecedent conditions and the position of lead actors played a large part in determining such matters. Where IEs were more clinical science orientated, the favoured conceptual model was one of changing individual behaviour, whereas social scientists (and at least one senior NHS manager) privileged the need for wider-scale organizational and system level change. Dependent upon configurations of CLAHRCs with regard to their director and the lead academic researchers, some CLAHRC lead actors merely added a need for individual level behaviour change to the portfolio of translational activity and seemed to lack recognition of the need for wider organizational and system level change.
Conclusion
The nine CLAHRC bids reflected antecedent conditions and the social position of CLAHRC lead actors in their structural context, with pre-existing clinical research activity privileged in their bids. In institutional terms, envisioning around CLAHRC bids reflected dominant views of the relationship between research and practice. In their early mobilization phase, rather than IEs advancing interests that are suppressed within prevailing or emerging organizational practices, they tended to reflect and buttress existing institutional arrangements in envisioning CLAHRCs.
The CLAHRC experiment in England holds important lessons for policy-makers globally regarding how to address the translation gap. First, policy-makers should consider whether they set out a more defined template for translational initiatives, such as CLAHRC, or whether they encourage variation providing a degree of learning and greater clarity about the design of translational initiatives. Second, there is a need to understand the importance of both antecedent conditions to CLAHRC bids, and the social position of senior CLAHRC actors leading bid development. Whilst established and known clinical academics are likely to be trusted to lead CLAHRCs, and the presence of pre-existing organizational relationships are important for mobilization, privileging such dimensions may constrain more radical change. Strategic change and outcomes of innovation may be locked-in, reflecting traditional clinical research, at an early stage of bid development. Strategies for change are not built independently from the structural context in which they are embedded, inclusive of consideration of the social position of the IEs leading translational innovation and antecedent conditions to CLAHRC bid development.
Regarding prescriptions for the development of implementation of knowledge translation networks, we tentatively suggest two points. First, that thought is given to the selection of leaders. More emphasis might be placed on those that recognize systems level issues and have the capability to work across organizational and professional boundaries. At the same time, we need to recognize that leaders of knowledge translation networks require legitimacy from powerful others (such as clinical academics) to drive change in research and clinical practice. IEs need to be embedded in existing power structures but willing to change such structures, if evidence-based innovation is to ensue. 27 Second, any lock in at early stages of development of knowledge translation networks might be mediated through relaxing any requirement to develop a detailed programme of activity in order to draw down funding. The co-production of a research agenda, which balances a service-centred and research-centred model for translation, might be engendered by more open funding.
These observations reflect bid development and the early mobilization of CLAHRCs. Those involved with CLAHRCs may rightly claim that CLAHRCs were emergent and that a good deal of flexibility and adaptation characterized their further development. In line with this, CLAHRC leads worked towards a greater understanding of the translational mission as research activity was mobilized. Hence, more research to track change in trajectory of CLAHRCs as they developed is needed. In particular, whether there was indeed translational lock in or whether early envisioning of CLAHRCs was mediated over time to move away from privileging clinical research. Finally, regarding the added value of CLAHRCs, the key question is the counterfactual ‘what would have happened in the localities where CLAHRC was funded, if such funding did not occur?’
Footnotes
Declaration of Conflicting Interests
None declared.
Funding
This study was supported by the National Institute for Health Research (NIHR) HS&DR Project grant 09/1809/1073 which is currently undergoing editorial review. The views expressed are those of the authors and not necessarily those of the funders.