Methodological strategies in child clinical trials: advancing the efficacy and effectiveness of psychosocial treatments

Journal of Abnormal Child Psychology,  Feb, 1998  by Bruce F. Chorpita,  David H. Barlow,  Anne Marie Albano,  Eric L. Daleiden

• E-mail
• Print
• Link

As difficult as it is to implement a successful intervention in the treatment of a childhood psychological disorder, it is still more difficult to document and evaluate the effects of that intervention in a controlled and meaningful manner. The task of evaluating, refining, and disseminating psychological interventions is an ambitious one, and among childhood disorders it represents a frontier of challenges. Clinical child research has long tolerated methodological difficulties above and beyond those encountered in adult research (Kazdin, 1995; Ollendick, 1986; Peterson & Bell-Dolan, 1995; Werry & Wollersheim, 1989), and there is a consensus that improving and developing clinical child research is an immediate priority (Members of the Committee for the Study of Research on Child and Adolescent Mental Disorders, 1995).

We argue that of great concern is the balance between internal and external validity(6) (e.g., Kruesi & Astrachan, 1995), an already challenging task whose difficulty is compounded by the enormous number of parameters that are free to vary in a child's environment (Kazdin, 1995). Although a considerable amount of attention has recently been paid to issues in child treatment research (see Eyberg, 1995; Hoagwood & Hibbs, 1995), the present paper reviews some previously established as well as some new ideas in terms of the basic issues of treatment efficacy (internal validity) and effectiveness or clinical utility (external validity), emphasizing the emerging need for strategies in clinical trials that balance these two dimensions. Examples from our own research programs as well as the extant child literature are used to point to gaps in existing evidence and to highlight issues and approaches warranting further development and attention.

A MODEL FOR INTERVENTION GUIDELINES

The APA Task Force on Psychological Intervention Guidelines (1995) outlined a template for the development of guidelines for interventions with psychological disorders, which serves as a useful starting point in a discussion on treatment evaluation. The purpose of this template was to influence and contribute to guidelines for determining interventions of choice for particular psychological problems. The first major dimension of the template involves the evaluation of treatment efficacy, that is, how well a treatment is known to bring about change in the target problem. Essentially, confidence in the efficacy of a treatment should be a function of both (1) the observed efficacy (i.e., effect size) of the treatment and (2) the quality of the empirical support for that treatment. With respect to the latter, the greater the internal validity of that support, the greater the confidence in that intervention. Along this dimension, research approaches are ranked in a hierarchy of internal validity (see Table I) to aid in the determination of the confidence about a particular intervention. Those studies at the top of the table would represent the greatest confidence and support for an intervention.

The second major dimension of the template involves the clinical utility of the interventions, noted to be an issue of external validity. Treatments evaluated along this dimension are ranked based on their feasibility, generalizability, and cost/benefit ratio (see Table I for the constituent features of these three categories). To the extent that these dimensions of an intervention are empirically evaluated, the greater the confidence in the clinical utility of that intervention. Hence, clinical trials, or at least applied clinical research, need ultimately to strike a balance of demonstrating both high efficacy and high clinical utility for interventions to be supported. Quite reasonably, this balance is almost impossible to achieve, even within a series of clinical trials. Obstacles exist in areas of both treatment and assessment, and are an inevitable part of clinical research.

TREATMENT

Comparative Evaluation

As the template suggests, the elements of a sound clinical trial should ideally include randomization to conditions, and comparative evaluations against existing approaches. However, the ability to incorporate alternative therapies as controls demands that such approaches are available, and, if so, that they can be implemented with sufficient integrity by the researcher. Unfortunately, the number of empirically supported treatments for childhood disorders is still somewhat small and does not always allow multiple choices for a particular problem. For instance, in the 1996 update on empirically validated treatments, only two of the twenty-two "well-established treatments" were for problems of childhood (Chambless et al., 1996). Likewise, only two of twenty-three "probably efficacious treatments" were for childhood problems. Despite these limits, successful comparative evaluation of manualized treatments has recently been conducted in the context of individual versus group approaches to cognitive behavior therapy for childhood anxiety (Barrett, Dadds, & Rapee, 1996). However, such approaches have rarely been extended to comparing qualitatively different psychosocial treatments, due to the overall limits of accepted and standardized treatments for any one problem.