Leaving the Ivory Tower: An Introduction to the Special Section on Doing Marriage and Family Therapy Research in Community Agencies
McCollum, Eric E., Stith, Sandra M., Journal of Marital and Family Therapy
Why do marriage and family therapy (MFr) research in community agencies? As the three papers in this special section demonstrate, the difficulties in doing community-based research are considerable. To counterbalance those difficulties, the payoff must be great. In our view, it is. The most compelling reason for MFT researchers to partner with community agencies is to develop treatment approaches that have clinical utility (Beutler & Howard, 1998). Not only must we demonstrate that our research-supported models work, we must also design them to fit the complexities of the community settings where they will be used. Notwithstanding how efficacious it is, no therapy model can be judged successful if frontline therapists are not able to use it, if it costs too much, or if it cannot be delivered within the confines of our existing health care delivery system. Questions of cost effectiveness and clinical utility are increasingly important to our field as policymakers turn to the research literature to make decisions about which types of treatment should be provided to the public. If we cannot demonstrate that we provide cost-effective and clinically useful treatment, our clients will be denied access to the breadth of our knowledge about how the problems of individuals, families, and cultures so intricately intertwine.
There has been a longstanding debate in the larger field of psychotherapy about the gap between research findings and regular clinical practice. This debate was crystallized in 1992, when Weisz, Weiss, and Donenberg reported the disturbing finding that the positive treatment effects demonstrated in a number of well-controlled, clinical outcome studies of child psychotherapy were not reliably reproduced in typical clinical practice. In other words, children were more likely to get better when treated in a research study than they were in the community settings where they are most likely to be seen. This finding called into question the very foundation of a "scientific psychotherapy" in which research findings are used to fashion day-today practice. Weisz et al. (1992) and others (see, e.g., Beutler & Howard, 1998; Clarke, 1995; Havik & VandenBos, 1996; Seligman, 1995, 1996; Shadish et al., 1997) have posited a number of ways in which research therapy differs from agency therapy. Research therapy is typically manualized, highly structured, not combined with other treatments, delivered by a therapist whose only role is to provide therapy, given to a rigorously selected and often homogeneous population, and delivered in a prescribed duration and frequency. Day-to-day practice, however, involves a therapist who must match treatment to the multiple needs of specific clients, who may have several roles with a given client (therapist, case manager, advocate), whose clients may be involved with a number of other service providers, and who can exercise considerable flexibility in the type, frequency, and duration of the therapy provided.
The discrepancy between research and clinic therapy has led to a number of proposed solutions. Weisz, Donenberg, Han, and Kauneckis (1995) suggest that clinical practice become more like research therapy. They recommend that clinical practice be based on "(a) the use of behavioral (including cognitivebehavioral) methods, (b) reliance on specific, focused therapy methods rather than mixed and eclectic approaches, and (c) provision of structure (e.g., through treatment manuals) and monitoring (e.g., through review of therapy tapes) to foster adherence to treatment plans" (Weisz et al., 1995, p. 83). A more prevalent recommendation, however, is that we test our models in settings and under circumstances more akin to the settings in which the tested treatment will ultimately be delivered (Beutler & Howard, 1998; Clarke, 1995; Henggeler, Schoenwald, & Pickrel, 1995; Howard, Moras, Brill, Martinovich, & Lutz, 1996; Neufeldt & Nelson, 1998; Shadish et al. …