A: There are three primary reasons. The first is that therapists find it challenging to fit the manualized treatments into their clinic routines and workloads. Second, the referred clients they treat are often quite different from the volunteers treated in many university trials; comorbidity is rampant in the referred cases but usually excluded from university trials. Finally, the parents don't look much like the middle class volunteers who tend to sign up for laboratory trials. Financial and social stresses often limit the clinic parents' participation in their child's treatment: appointments are frequently missed, and termination can occur without notice. For these reasons, I'm questioning the value of the Biomedical Intervention Testing (BIT) Model that has guided psychosocial treatment development research for decades.

A: In the BIT Model, which originated in drug treatment research, investigators move from tightly controlled laboratory trials to dissemination trials in real-world practice settings (to test "public health impact"). Placing research in practice settings at the tail end of the treatment testing process may arguably make sense for medications, but it may not work so well for psychosocial treatments. This is true, in my view, because the very real-world factors that experimentalists might view as a nuisance (e.g., child comorbidity, parent pathology, life stresses that produce no-shows and dropouts, therapists with heavy caseloads) and thus attempt to avoid (via exclusion criteria) or control, are in fact, precisely what we need to understand and address to make psychosocial treatment protocols work in real clinical practice. Treatments that cannot cope with these factors may not fare well at all in clinical practice no matter how efficacious they appear to be in laboratory clinical trials.

A: I argue for the Clinic-Based Treatment Development (CBTD) Model, which moves the development and testing process into real-world contexts as soon as feasible. Under this model there are six steps of treatment development:
Step 1: Theoretically and clinically-guided construction, refinement, and manualizing of the treatment protocol.
Step 2: Initial efficacy trial under controlled conditions to establish evidence of benefit.
Step 3: Single-case applications in practice settings with progressive adaptations to the protocol.
Step 4: Initial cost-conscious effectiveness test.
Step 5: Full test of effectiveness and dissemination.
Step 6: Effectiveness of treatment variations; moderator and mediator tests.

A: Step 2 is an efficacy trial under controlled laboratory conditions. Study volunteers are used here rather than clinic-referred cases for ethical reasons in order to avoid any risks from using an untested intervention with severely disordered individuals who were seeking clinic treatment.

A: The goal here is to insure faithfulness to the core principles and the model of change that guides the treatment protocol but to improve the appropriateness and fit of the treatment for the clinical setting and the community. Modifications are made to the protocol to improve the fit while always adhering to the theory.

A: I define "success" in two ways: both effectiveness and disseminability. Of course, effectiveness means whether clients treated with the new intervention show greater gains than those receiving treatment as usual. But disseminability is also important to study. We want to know whether the trained staff practitioners actually adhere to the manual in their treatment sessions.

A: Some might argue that lab studies are the most appropriate context for these types of studies but I am not so sure. My concern is that what we learn through dismantling studies and through moderator and mediator tests in laboratory clinical trials may not apply so well to the treatment when it is used in real world service settings. Because the clients, therapists, and conditions of treatment in those settings differ in so many ways from those of the lab tests, the degree to which separate components of the treatment package can produce effects may be quite different from what was found in the lab. Similarly, the impact of various moderators, and the processes that mediate treatment effects, may be different for real world clinical cases and settings.