Advancing a mission of translational intervention science: Comment on premature implementation

We appreciate the perspective shared by Freedland, Powell, Czajkowksi, and Epstein (2024) in response to our Viewpoint, entitled, “Testing psychosocial interventions in the context in which they are meant to be delivered.”

We offer that we agree about the need to move away from traditional approaches to the discovery process as it relates to intervention development and that we share concerns about the “classical treatment package approach” described in Guastaferro & Collins (2021). We share Freedland and colleagues’ enthusiasm for newer methodological approaches, including MOST (Collins, Murphy, & Strecher, 2007) and the ORBIT model (Czajkowski et al., 2015). Unfortunately, many of these innovative approaches for intervention development, refinement, and optimization still are not the norm, nor are they taught in graduate programs or always well-received by grant review committees. Furthermore, applications of these approaches do not always attend to speeding the translation process for the pressing needs and problems facing our communities now. This may be particularly true among systemically marginalized communities that have poorer access to resources and experience vastly different contextual barriers that impact intervention delivery and uptake, shaped by many years of historical and ongoing structural disadvantage and disinvestment. We do not view these approaches as mutually exclusive to our suggestions for ways to speed translation to meet community need, and with this response amplify the opportunity to include these approaches within the armamentarium of impactful translational intervention science.

Our primary assertions included the following two points. First, irrespective of the “how” (i.e., which scientific approach or study design) used in the discovery process, many psychosocial treatments are best tested in the context in which they are meant to be delivered, particularly those that build on decades of theoretically informed psychological and behavioral science demonstrating safety, efficacy, effectiveness, and evidence for mechanisms of action. Concomitant with this laser focus on real-world context and prioritization of external validity (note, having multiple centers in a trial does not necessarily result in high external validity), we propose ways for translational intervention science to be more community engaged and driven by community-identified need, thus resulting in interventions that are more likely to be adopted, successfully implemented, and sustained across a wide variety of contexts. Second, hybrid studies that include questions related to both clinical and implementation effectiveness can accelerate and enhance impact by identifying and understanding factors related to implementation earlier in the translational continuum, making it less germane to conduct highly controlled studies among “best case” recipients or implementers (Curran et al., 2022) who meet narrowly defined eligibility criteria. It is possible that our argument may have been misunderstood in the limited word count available to us. The response by Freedland and colleagues does not refute these core ideas. The approaches they raise offer alternative ways to advance psychological science, but do not supersede nor replace our assertions or our plea for scientists to consider forgoing efficacy work when possible and appropriate, and to instead co-create and test interventions in the contexts where the interventions are meant to be delivered. Balancing rigor and pragmatism is crucial, particularly when deploying interventions that build on decades of research on mechanisms of action, for pressing problems facing our communities now and to reduce longstanding health inequities.

Our different lenses and experiences may shape the “how” in terms of which scientific approaches and pathways to use to evolve the discovery process. We gave an example of a randomized clinical feasibility pilot study followed by a fully powered study that was in response to a community crisis (i.e., parental opioid and methamphetamine use within the child welfare system) where there was already robust implementation momentum and the intervention was built on decades of existing evidence-based treatment approaches. Our example was but one of many ways to move forward around this shared vision and does not preclude the opportunity to conduct explanatory mechanistic work while also advancing viable interventions in communities that are suffering and urgently need them now. We appreciate the point that pilot studies should not be used alone to make the decision to progress to a fully powered trial. In fact, in the example we offered, the intervention has continued to evolve and be refined via rigorous methods. Freedland and colleagues offer other options which we agree can be paths forward – but only if they are viable within the timelines, needs, and priorities of the participants, settings, health systems, and communities with whom we are partnering.

Our respective background and lenses likely shape our focus. Our group prioritizes impact, pragmatism, and external validity, after being in settings with patients, clinicians, organizational leaders, community members, and policy-makers too many times to count where internal validity was prioritized at the expense of producing viable interventions to deploy in contexts where care is delivered. Our lens highlights the implications of a long tradition of over-valuing, through the investment in research, work that focuses earlier in the translational pipeline that has not historically centered or amplified the voices of minoritized communities and individuals with lived experience (Baumann & Cabassa, 2020, Buckley, Murry, Gust, Ladika, & Pampel, 2022, Rachel C. Shelton, Adsul, Oh, Moise, & Griffith, 2021). This prioritization has resulted in often non-viable interventions that are not feasible to implement outside of controlled and well-resourced settings, or do not produce the same level of effective outcomes for real-world communities resulting in limited return on investment (Glasgow, 2008). In reading Freedland and colleagues’ commentary, it appears that their lens prioritizes internal validity and more explanatory and mechanistic work. While we also value these scientific goals and concur that the approaches they identify are clearly methodologically rigorous, we do not think the approaches will substantially reduce the unacceptable translation gap or address the fundamental issues of context. Our approach recognizes that there is tremendous value in co-creating solutions and interventions with patients, clinicians, and community members in the settings where we are seeking to promote health and address health inequities, and questions traditional assumptions and paradigms that scientists ‘know best,’ have effective solutions, or should hold all of the power and knowledge (Brownson, Shelton, Geng, & Glasgow, 2022, Sanchez et al., 2023, R. C. Shelton, Adsul, & Oh, 2021). We believe it is critical that we expand the pathways through which we advance the science of intervention science in a meaningful and impactful way, and with more explicit attention to issues of context, equity, engagement, and external validity. In our opinion, the narrow repertoire of methods and incremental approaches that have been historically prioritized are insufficient to speed translation or create widespread, equitable, or lasting health improvements; as such, we see a critical need to explore other paradigms such as the one that we put forth.

The path forward is a dialectic accomplished via respectful discourse and paradigm shifts. In fact, this debate has been ongoing in the past twenty years (Glasgow, 2008) as the scientific community has not yet come to consensus on the best way to rigorously and adequately address these important issues. We must strive to balance how we come together in such a way that allows us to provide impactful and scientifically meaningful work in collaboration with communities on the striking health inequities and pressing problems that are priorities now. In summary, we concur that innovative and different research methods are needed to advance a mission of translational intervention science. We strongly suggest that approaches which focus on context and external validity earlier on and that strongly engage the community and end-users throughout the research process will more rapidly and thoroughly translate into equitable and sustainable practice that results in more widespread impact.