ADVANCEMENTS IN BEHAVIORAL INTERVENTION RESEARCH
Over the past 50 years, there has been a growing corpus of behavioral intervention research that has yielded well-tested programs and important advancements in the conduct of this form of inquiry. As in any research tradition, behavioral intervention research has an evolving language and specific techniques, methods, rules, and standards that are unique to this particular endeavor. Although it overlaps other forms of inquiry such as classic clinical trial methodologies, behavioral intervention research also has its own distinctive challenges and foci. In the chapters that follow, to the extent possible and where appropriate, we draw upon the most important lessons garnered from classic methodologies and approaches, but also discuss considerations specific to this type of inquiry.
Historically and broadly speaking, the initial wave of behavioral interventions had significant limitations. These included: misalignments between study samples, intervention intent, and measured outcomes; lack of theory-driven approaches and an understanding of underlying mechanism(s) of treatment effects; lack of inclusion of diverse populations; and simplistic approaches such as expecting and measuring behavioral change as an outcome from an intervention that provided only education materials and enhanced knowledge. Take, for example, initial caregiver intervention studies that sought to reduce depressive symptoms although study inclusion criteria did not specifically target depressed caregivers (Knight, Lutzky, & Macofsky-Urban, 1993). Not surprisingly, an initial wave of caregiver studies showed minimal to no treatment effects for depressive symptoms, as there was little to no room for improvement on this outcome (Callahan, Kales, Gitlin, & Lyketsos, 2013). Furthermore, many interventions were designed with little understanding of the context in which they might ultimately be implemented if they were proven to be effective. These missteps have led to a greater understanding and awareness that a translational phase is typically necessary to take an intervention and adapt it for delivery in specific service contexts (Gitlin, Marx, Stanley, & Hodgson, 2015).
Today, we have a much better understanding of best practices in the conduct of behavioral intervention research. Although there is no universal or agreed upon set of approaches, practices, designs, or strategies, the collective knowledge, experience, and empirical evidence as to what works and what does not work in conducting behavioral intervention research is being amassed. For example, we know how to align theory with intervention development, use epidemiologic findings to identify intervention targets, involve communities and stakeholders in developing and implementing interventions, evaluate who benefits the most from interventions, embed interventions in practice settings and evaluate effectiveness using sophisticated adaptive designs and analytic techniques, and monitor and measure the impact of treatment adherence on treatment outcomes. Further, we now have experience standardizing intervention protocols, developing treatment manuals and training protocols, and conducting multisite and pragmatic trial designs that can potentially accelerate knowledge generation and its transfer to broad real-world settings.