Community-Based Participatory Research as Worldview or Instrumental Strategy: Is It Lost in Translation(al) Research?

Community involvement in community-wide interventions is important for a variety of scientific, ethical, and pragmatic reasons.^1,2 However, the specific meaning of community involvement depends on the details of how it is enacted. Katz et al.³ outline an ambitious effort to blend the science of randomized controlled trials (RCTs) with the processes of community-based participatory research (CBPR) in translational research. RCTs provide the science, while CBPR provides the processes of tailoring and implementation. Katz et al. offer a detailed example of how research might occur through the use of community portals and community health advisors as local advocates for the delivery of interventions. Their examples are rich and raise fundamental issues regarding the importance of CBPR and the role of local participation in translational research more generally.

CBPR AS WORLDVIEW OR INSTRUMENTAL STRATEGY

Over time, CBPR has developed as a coherent worldview, and the concept includes several characteristics outlined both in the Katz et al. article and in the writing of leaders in the CBPR field^1–5:

1. the community as the unit of identity, solution, and practice^1–5;

2. community involvement in decision-making throughout the intervention process, from problem definition to planning for sustainability;

3. structural, policy, community capacity, empowerment, and individual change goals;

4. a critical realist or constuctivist philosophy of science¹; and

5. a concern that communities are able to sustain what they find useful resulting from the intervention.⁶

The use of CBPR as an instrumental strategy differs fundamentally from every aspect of this worldview. As described by Katz et al., CBPR is selectively invoked to accomplish predetermined aims or goals not collaboratively developed or locally defined. Local influence appears in translating findings to local context. This is an important area on which to focus, and local participation in this process is important to theorize and enact. However, it relegates local knowledge and influence to carrying out the science devised by others. While there is flexibility in how it is done, local involvement occurs within a framework of expert-designated fixed components. In the context of these givens, CBPR as instrumental strategy brackets rather than embraces the larger CBPR worldview in terms of when and how local influence is expressed. The RCT and CBPR paradigms are not blended, but sequenced under the assumption that, early on, science is enhanced by ruling context out (the RCT, not the CBPR assumption) but scientific findings need to be subsequently recontextualized through local participation. In this role, CBPR is in the service of the RCT, its assumptions, and its givens.

Included in these givens are the outcomes of importance. In the exemplar Diabetes Prevention Program (DPP) intervention described in the Katz et al. article, the goals are exclusively individual-level or “lifestyle” outcomes. But the goals of CBPR also address community-level concerns such as increasing capacity-building of individuals and social settings.⁶ In Katz et al., the heuristically useful concept of portals is expressly viewed as means to the end of diabetes prevention (defined in terms of individual outcomes), rather than also being explicitly included as objects of community capacity building and empowerment for future local problem solving. In fundamental ways, then, CBPR as an instrumental strategy ranges far from its worldview.

THE FORM VERSUS FUNCTION DISTINCTION

In arguing for CBPR as an instrumental strategy for translational efforts, Katz et al. employ the form versus function distinction of Hawe et al.⁷ The notion is that instead of standardizing the specific forms or components of an intervention, the functions they are designed to serve should be standardized across settings, allowing the forms to vary. For example, if the function of an intervention is to educate patients about depression, rather than standardize and distribute a patient information kit, sites would be free to devise their own ways of distributing the relevant information tailored to local “literacy, language, culture, and learning styles.”^7(p1562) As they put it, intervention integrity is defined as the “evidence of fit with the theory or principles of the hypothesized change process.”^7(p1563)

There is obvious merit to freeing the RCT from a literal definition of standardization-as-repetition of the same activities across diverse communities. However, the Hawe et al. worldview from which the form and function distinction flows is one that regards community interventions as complex rather than simple.⁷ Whereas simple interventions are conceptualized as the sum of their discrete parts or core components, complex interventions rest on systems theory perspectives that address interactive functions rather than core components and ripple effects of interventions including, but going beyond, individuals. Here, “reducing a complex system to its component parts amounts to 'irretrievable loss of what makes it a system.”'^7(p1561)

Applying this perspective to translational research raises important questions about the concept of core components. First, they are neither easy to identify nor isolate as independent contributors to outcomes.⁷ The DPP, for example, “was not designed to test the relative contributions of dietary changes, increased physical activity, and weight loss to the reduction of risk in diabetes.”^8(p398) Rather, it was planned to assess the combined effects of the intervention components addressing these different outcomes. How these components combined, or how they may combine in different contexts or with different populations, is unknown.

More fundamentally, the Hawe et al. perspective begins with theorizing principles of change, mechanisms, or mediators of intended outcomes, and translating those principles into forms that may vary across contexts. In the translational example in Katz et al., the core (“fixed”) components are conceptualized as forms rather than functions. The participatory task is to figure out how the function they presumably serve can be translated. But starting with the form rather than the function reverses the process suggested by Hawe et al. and seemingly assumes that forms or core components serve the same function across communities but need to be tailored to local context.

A focus on function at the outset of the translation process would complicate greatly the notion of core components and the science behind it. For the translation of the 16-session curriculum of the DPP, for example, a discussion of function in diverse communities may result in varying the length or content of a prescribed curriculum, or it may imply developing something other than a curriculum to fulfill the same function in different communities. In CBPR terms, the greater the emphasis on form, the less formative the local influence on translation, and the greater the emphasis on function, the stronger the local influence is, but the issue of fidelity as traditionally defined becomes more imperative. Katz et al.'s concern about local influence diluting the generalizabilty of interventions reoccurs in the translation process if Hawe et al. are to be taken seriously.

EXTERNAL VALIDITY AS SCIENCE AND RESOURCE BURDEN

The rationale behind translational research rests in part on the generalizability of scientific findings. A contextualist perspective congruent with CBPR is likely to ask for whom and under what conditions findings were generated, and how do the findings apply to the local community. The DPP screened over 158 000 individuals to achieve a final sample of 3234,⁹ with “most exclusion criteria … chosen to reduce the risk of adverse effects of the intervention.”^9(p624) The implications of such selectivity needs to be clarified for local participating organizations not only as a translation issue but as an ethical issue of not overstating what we know about for whom the intervention has shown efficacy.

However, translational research, in addition to issues of local acceptability and commitment, also involves an assessment of resources necessary to translate and sustain the intervention. The high impact arm of the DPP included an elaborate recruitment process, culturally tailored dietary and exercise components, a 16-session core curriculum delivered by trained case managers, an exercise regimen of 150 minutes per week,⁹ and additional staff to monitor adherence and retention.¹⁰ These are all part of what it took to get the intervention effect. The lack of “take” of the DPP program cited in Katz et al. may be influenced less by lack of community involvement than by the scope of needed resources. Scientific humility about the generalizability of externally derived and candid assessments with local portals on the resources necessary to conduct and sustain such an effort would seem to be a prime topic for important local input in translational research.

The CBPR worldview includes thinking systemically and contextually. It addresses sustainability through such goals as capacity building and empowerment. The form versus function distinction of Hawe et al. likewise conceptualizes interventions as complex and the intervention process as systemic. The Katz et al. article operates on a quite different paradigm that focuses on intervention as simple with fixed components, and views the intervention as the technology rather than an event in a system with multiple outcomes and interactive rather than additive aspects.¹⁰ In these areas the CBPR worldview differs fundamentally with the worldview as portrayed in Katz et al. and in the DPP example and is subservient to it. As Katz et al. point out, the adaptive requirements for investigators to secure external funding often set constraints on the degree to which local participation can meaningfully contribute to the co-creation of locally relevant interventions. We need more conversations about moving translational research beyond the image of community interventions as product development and product dissemination. The Katz et al. article serves this very important function.