The search for social work treatments of choice: What interventions work better than others?

Abstract

Dr. William Reid spent his professional life building, critiquing, and synthesizing research on social work interventions. With his death last winter, social work lost one of its most prolific and dedicated researchers. Our readers are privileged to have in this issue one of Dr. Reid's last articles, one that is destined to have impact and be cited for years to come. Long before the terms or guidelines were vogue, or even familiar, pioneering scholars like Bill Reid worked to identify works for social work. In 1982, Dr. Reid critically reviewed a set of group experiments on social work interventions, and he published the results of this review with Dr. P. Hanrahan in (Social Work). Each of the studies reviewed for that 1982 article was based on an equivalent-group experiment in which clients were randomly assigned to service or control groups or to alternative forms of social work intervention (Reid & Hanrahan, 1982). Most of these studies were tests of various structured forms of practice, such as behavioral treatments. That important publication was a milestone in the profession's quest to identify treatments with demonstrated effectiveness. The advocacy of evidence-based practice is based on premises that treatments vary in their effectiveness. It assumes that service providers should know and should deliver treatments that are effective. In an effort to identify the comparative effectiveness of treatments, scholars have conducted systematic reviews, including meta-analyses, of experiments that compare different methods of psychotherapy. Often these reviews yield a dodo bird verdict: Different treatments yield similar results. We are left to ponder, Does the treatment method really make no difference? Probably not. Rather, the failure to find (presumably real) differences has been attributed to such methodological problems as the studies' insufficient attention to client variables that moderate effectiveness, inadequate statistical power in the comparisons, or the potency of such common factors as the practitioner-client relationship, contained in virtually all psychotherapies. But what about the assumption that some interventions work better than others? Bill Reid persisted; he continued to pit study against study, trying to discern what treatments emerged as more effective than others, trying to distill a critical mass of social work interventions that work. According to Bonnie Davis Kenaly, one of Bill's coauthors, the sweeping review article published in this issue reflects the product of his work to update the findings of the 1982 review article. In Do Some Interventions Work Better than Others: A Review of Comparative Social Work Experiments, I see traces of two of Bill's trademarks. First, Bill was optimistic about our profession and our potential for impact through intervention. He could pursue the question of works? decade after decade because he had faith in social work interventions. He was convinced that our interventions were effective and that their effects could be discerned and documented. Second, in this article Bill characteristically raised the bar on the question. He moved beyond the question, What works? Assuming that social work has a store of several effective interventions, he posed more challenging questions: interventions are effective for which problem? Which interventions are more effective? Dr. Reid assumed a high standard for our profession; if several effective interventions are available for a given problem, evidence-based practitioners will want to know and use the intervention that is most effective. The present study sought to review all experiments comparing social work programs that were published between 1990 and 2001 (n = 39). This article seems to be the only published review of comparative evaluations of social work programs to date. Most of the studies used true, or randomized, experimental designs. …