Conducting clinical interventions, in a research setting poses numerous challenges to clinicians, researchers, supervisors, and research participants. These issues often reflect a tension between the rigorous demands of a research protocol needed to ensure internal validity of the research and clinical flexibility that is representative of an externally valid intervention. In this paper, we explore difficulties in defining the roles of clinicians and research participants, delineating the scope of the intervention, and problems encountered at the intersection of clinical intervention delivery and such methodological techniques as randomization, blinding, and incentive payments. We provide guidance for setting frames for the intervention and clinical relationship, conducting training and supervision of clinical staff, and structuring quality-assurance procedures in order to meet these challenges.
Clinical research, including randomized clinical trials of behavioral interventions, is critical for the development and validation of effective behavior-change approaches, including psychotherapies (1). Conducting clinical research, however, is challenging in a number of ways. A primary task is to maintain a proper balance between the broader clinical components of human intervention and the specific research aims of the study. The struggle between intervention protocol and clinical flexibility is not unique to research. For example, some treatment modalities, such as cognitive-behavioral therapy, involve a degree of structure that may seem overly regimented and goal directed to some clients and/or clinicians. Nevertheless, the task of clinical research presents a number of unique situations and dynamics that create challenges not typically encountered by clinicians and clinical supervisors in nonresearch settings.
One source of tension comes from rigorous demands required in a behavioral efficacy trial of an intervention, which can be experienced as artificial in the clinical environment. Therefore, creating an optimal balance between the clinical intervention and the research methodology is crucial to the success of the study. Close attention to intervention fidelity is required in research in order to reduce variability or "drift" from the prescribed protocol that is being tested. Significant variation from the prescribed protocol (i.e., across participants in the trial, or across intervention facilitators or intervention sites) can jeopardize the integrity of the research trial and render the results (positive or negative) uninterpretable. In contrast, if there is poor attention to the real-world clinical aspects of intervention delivery, it would be very difficult to retain participants and study personnel, and again, can jeopardize the trial's success. The purpose of this paper is to present challenges likely to be confronted when conducting a behavioral-intervention research trial and to offer recommendations for meeting these challenges, providing guidance for researchers, clinicians, and supervisors involved in such research.
Adherence to an Intervention Manual
Developing a treatment manual is an important early step when setting out to test the efficacy of a clinical intervention. A standardized document facilitates consistency in training and ongoing supervision of clinicians delivering the intervention. Such standardization of intervention can serve central roles in the evaluation of an intervention and the translation of effective interventions into clinical services settings. A number of challenges associated with treatment manuals have been identified in the context of clinician training (2), clinical service delivery (3, 4) and research (5). Specifically, the tasks of monitoring and assuring fidelity to manualized protocols have been described in behavioral intervention research (6-8). Threats to internal validity, or the degree to which the delivery of the intervention successfully achieves the aims of the protocol, can exist if the intervention protocol is not followed sufficiently, by a given facilitator of the intervention over time, and across multiple facilitators. …