Determining Whether Further Testing Is Required
Thus, what is in question is whether modifications to the delivery of an intervention result in the need to retest the intervention in an efficacy trial. Whereas moving from Phase I to Phase IV is sometimes referred to as a “forward translational process,” moving backward from effectiveness to efficacy is sometimes referred to as a “backward translational effort.”
Although there are no clear metrics for determining how much change to an intervention is too much, several strategies can be employed. One strategy involves identifying a priori the core components and theoretically based principles (see Chapter 4) of an intervention that should be considered immutable or must remain intact. Similarly, it is helpful to identify a priori the aspects of an intervention that can be modified or delivered differently (Gitlin et al., 2000). Another strategy is to use analytical techniques such as mediation analyses, dose and response analyses (see Chapters 16), and fidelity assessments (e.g., analyses of the actual dose and intensity that was implemented) to inform decisions as to the aspects of the intervention that can and cannot be modified.
For example, in the randomized trial of the Get Busy Get Better intervention (Gitlin et al., 2013), an average of 8 of the 10 intended treatment sessions were found completed, which suggested that the number of visits could be reduced in future replication efforts. Importantly, using mediational analyses, it was also found that all of its five treatment components (care management, referral and linkage, instruction in stress reduction techniques, education about depression and symptom recognition, and behavioral activation) worked in concert and contributed similarly to reducing depressive symptoms (Gitlin, Huang, & Roth, 2014; Gitlin, Szanton, Huang & Roth, 2014). Thus, delivering all five components appears to be essential such that in its replication it would not be possible to eliminate one component or implement only select components. As these components can be delivered within 8 versus 10 home sessions, the intervention could be modified in this way, at least for most program participants.
National Institutes of Health Designation of Research Phases
This traditional pipeline has also been conceptualized in the National Institutes of Health (NIH) of the United States as involving three broad translational phases: research designed to bridge basic to human research, or T1 (translational) research; research to evaluate the efficacy and effectiveness of human research and moving it to the community, or T2 research; and research (dissemination, implementation, quality improvement) moving from community-based research to practice, referred to as T3 research (Kleinman & Mold, 2009). NIH funding favors T1 research with little monies offered for T2 or T3 (discussed in Chapter 23).