The skill or fidelity with which a model—even a well-researched and demonstrably effective one—is used can have a major impact on whether the model makes it into the big leagues of widespread practice or languishes at the margins. And the success or failure of make-or-break trials of an approach may hinge on circumstances surrounding the trial that have nothing to do with the inherent worth of the model itself.
Case in point: in the last 30 years, Brief Strategic Family Therapy (BSFT) has acquired solid empirical support for its effectiveness with substance-abusing adolescents and their families, especially in the Hispanic community. But the intricacy of the model requires considerable training for therapists and an investment of agency time, so cash-strapped community agencies—the likeliest venues for treating inner-city, substance-abusing adolescents—need more dramatic proof before committing to it. Therefore, when one of its primary developers, University of Miami psychologist Jose Szapocznik, landed a major grant from the National Institute on Drug Abuse (NIDA) to measure BSFT’s effectiveness against agency treatment-as-usual with 480 families in eight community agencies in the mainland United States and Puerto Rico, he hoped the results would convince agencies to use the approach more often. However, the study didn’t live up to expectations, although several of the outcomes were slightly better than those of the usual treatments. What went wrong may reveal more about the kinds of research problems that block widespread adoption of complex treatment approaches than about the qualities of BSFT.
The researchers ran into trouble from the outset. The grant was conducted at community agencies that work with NIDA’s Clinical Trial Network. This network supplied the therapists, who, typically, were already providing clinical services at the agencies. “Normally when we go into an agency and train therapists,” Szapocznik says, “we choose people who already have experience working with adolescents or families, and who have a certain level of clinical skill.” This time, the researchers were told which therapists to train. Because of the limited pool of 79 therapists offered for the study, Szapocznik’s team could exclude only 2. It wasn’t that the therapists selected were bad, but family therapy in general, and particularly BSFT, requires an orientation and set of observational and interactional skills that not only take substantial time to acquire, but frequently run counter to the previous training of clinicians who haven’t practiced family therapy.
The study design influenced the outcome in another way. No-shows are a common occurrence in therapy with poorer urban families. “In every evidence-based family therapy model,” Szapocznik points out, “to achieve the best results, it’s important to take hold of a family and not let them go.” BSFT therapists work hard to bring all relevant family members to sessions, sometimes using their free time between sessions for calls and reminders. However, the therapists in the study were under agency pressure to fill their “empty” hours with billable tasks. The result was that 66 percent of the therapy sessions lacked at least one key family member—not because the therapists didn’t know the model required full attendance, but because circumstances militated against strict adherence to BSFT principles.
Still, Szapocznik’s analysis of his therapists’ adherence found that, on a 5-point scale, 90 percent of the BSFT therapists were in the acceptable 3- to 4-point range. Three is considered minimum adherence. What really undermined results, however, was the lack of clinical competency in the way the model was delivered. Training therapists who have no family therapy background in BSFT “is like teaching someone to ice skate,” says Szapocznik. “You can train them to skate, but you can’t expect triple axels.”
A subsequent analysis of a subset of the BSFT cases by independent researchers Michael Rohrbaugh and Varda Shoham, who are on the faculty of the University of Arizona, suggests that the fidelity of the BSFT therapists was considerably below the quality Szapocznik would have wanted. Among the most common fidelity-related omissions were therapists’ failure to engage key family members or to treat communications and interactions in terms of three people—both of which are cornerstones of family therapy.
Given the difficulties with adherence and fidelity, BSFT still outperformed treatment-as-usual. Szapocznik points out that therapists with the highest adherence ratings had the best outcomes, perhaps the most positive finding for BSFT’s effectiveness in the study. Meanwhile, Szapocznik is conducting a four- to five-year follow-up study, hoping that BSFT’s gains will prove more durable than those of the usual agency treatments. And better days may be ahead for family therapy and clinical trials. The National Institute of Mental Health has set aside approximately $1.5 million in grants this year to study and improve treatment fidelity. Hopefully, some of the studies developed from these grants will lead to suggestions for how to ensure that therapists participating in clinical trials genuinely have the skills to correctly follow the model under investigation.
Chasing Therapy “Facts”
Despite the millions of dollars spent on research, psychotherapy will always have one foot on the bedrock of science and the other in the loam of subjective truth. Therapy has always been both enmeshed with and an arbiter of the shifting landscape of cultural values and ideas.
Remember when most therapists “knew” that it’s always healthy to express and vent emotions? That notion has now been convincingly disproven, but in the buttoned-down 1950s, it may indeed have been healthier, even necessary, for people to free up and express their emotions. Did the shift in social attitudes toward more acceptance of emotional expression perhaps come about in part because legions of past therapists helped make emotional catharsis a mainstream ideal? Other widely believed “facts” among the public and therapists that are unsupported by research include the beliefs that there’s a common pattern to working through grieving, that men and women are biologically engineered to relate and communicate differently, and that extramarital sex always signals intimacy problems in marriages.
Unlike laws and empirically derived theories in the physical sciences, psychotherapeutic “truths” can change with bewildering speed. This has been especially true during the digital information age, when ideas and values mutate and propagate with unprecedented frequency.