Whose Evidence and for What Purpose?

Witkin, Stanley L. ; Harrison, W. David

Evidence-based practice (EBP) is one of the predominant new ways of thinking about what social workers should do in their practice and how they should decide to do it. EBP involves using the "best available" evidence, often interpreted to mean research-based "knowledge," about specific types of practices with particular problems. Although its advocates tout EBP as an imperative for social workers, others have raised questions regarding potential drawbacks of this approach. This editorial is intended neither to advocate nor to oppose EBP, but rather to identify issues that we believe social workers should consider.

Origins and Characteristics of EBP

Before there was EBP, there was EBM--"evidence-based medicine"--a didactic approach first used with medical students in Canada and later applied to the solution of clinical problems. Widely adopted in the United Kingdom, and increasingly in the United States, EBM is used to determine the most desirable ways to promote health and especially to treat illnesses. Its more general form, EBP, has become a major dimension of professional education in the United Kingdom and a way of attempting to arrive at a consensus about what collective bodies of research findings have to recommend. Gambrill's (1999) thoughtful and informative article advocating EBP in social work documents some of these fundamental and influential British sources.

The medical origins of EBP are evident in the value placed on randomized clinical trials, similar to what social workers call experimental designs. Information generated by randomized clinical trials is taken to be the "gold standard" of evidence. Although results from studies using less traditional research controls such as case accounts are used by EBP, they occupy a lower status in the hierarchy of credible evidence. Judgments about evidence also are based on systematic reviews of treatment-outcome studies and meta-analyses that aggregate several research studies statistically. Assessing such evidence is a complex process requiring a high level of research sophistication and knowledge of the subject matter. For example, even with a large group of randomized clinical trials on a topic, small alterations in the definitions of problems or "interventions" can lead to changes in what is considered best practice. A review of readily accessible online reports of EBP or evidence-based medicine studies (see, for e xample, Research Triangle Institute, 2000) shows that various types of "psychosocial" treatments are sometimes aggregated across studies, and that medically precise definitions of "outcomes" may be hard to reconcile with social workers' espoused views of taking into account all relevant aspects of a social situation.

Social Work and EBP

Today, EBP has become a common term in many professions, including social work. Attempts to deal seriously with systematic evidence as a way to reduce uncertainty and improve practice have a long history in social work, as anyone familiar with the extended and legally oriented presentation concerning evidence as a basis for social work in Mary Richmond's Social Diagnosis (1917) will recall. Similarly, social surveys historically have provided evidence of the existence and effects of structural inequalities in society, often with suggestions for reform and documentation of the social benefits of reforms. The contemporary social work version of EBP (like its predecessor, empirical clinical practice) is focused more on weeding out ineffective therapies and practices and recommending interventions that logically are related to predetermined changes.

Social workers' current advocacy or adoption of EBP can be thought of as an expression of the profession's recent attention to research activities and ways of thinking. The idea of systematically basing our practice on scientific evidence is appealing in our "tell me what works" society. Paralleling medicine to a degree not seen in years, recent concerted efforts to place social work in the mainstream of scientifically oriented professions can be considered the enactment of cultural beliefs about what a profession should do and be. If only we have enough evidence, based on accurate specification of conditions, outcomes, and interventions, we should be able to solve our problems. What remains unclear is the range and types of problems for which the "what works" formulation is helpful. For example, it may be a useful formulation for problems that are believed to exist stably across time and contexts, where problems and interventions can be specified (operationalized) and replicated in other practice settings, where they exist in deterministic relationships to one another, and where outcome measures are seen as reflecting rather than generating the problems they address. It seems less useful for a world characterized by shifting, multiple identities and relational constructions, in which an "outcome" is at most the beginning of something else, or where the production and evaluation of knowledge for practice are considered mutual activities between professionals and people who use their services.

At its heart, EBP involves using the best evidence available to help practitioners make decisions and conduct their practice in concert with professional values. Ideally, "best evidence" would include integrating knowledge gained from practice experience with knowledge gained through research, a position that even some EBP advocates accept (see, for example, Sackett, Rosenberg, Muir Gray, Haynes, & Richardson, 1996). Practitioners should be able to understand, critically assess, and apply evidence that is relevant to the tasks and missions they undertake. They should be able to do so while appreciating the complexity and meaning of the situation to the people with whom they hope to reshape the situation in mutually desired ways.

Social work has produced few studies that exemplify randomized clinical trials. One reason for this may be the complex ways in which social workers understand their practice. Social workers see the heart of their practice as "person in situation," in expanding problem understanding to include social and environmental elements. Social work practice involves seeing people as much for their differences as for anything that links them to classifiable problems or diagnoses. It values the often subjugated perspectives of the people we serve and attempts to understand their individual and collective narratives of their situations and conditions. We learn to work with moral narratives, the "morass of goods and bads, rights and wrongs, evils and virtues, bearing little resemblance to the diagnostic labels or the balance sheet of assets or liabilities that the client inevitably earns" (Goldstein, 2000, p. 349). These interactive accounts of people in their situations are not just tools for understanding, but the essen tial components of the individual's experience of social problems, medical conditions, and behavior. We learn to listen for discrepancies between the public discourse of disadvantaged people dealing with more powerful systems and the internal discourse within groups and individuals that frequently offer different understandings. In this sense, social workers often are cultural bridges, able to deal in multiple worlds of understanding. Sometimes this involves using the logic of EBP with clients when there is credible evidence of some relevant knowledge available. Other times, however, the most important work is in educating decision makers or those who have control of resources about how irrelevant the best scientific evidence is to the world of people whose experiences brought them into contact with the professionals.

It is interesting to note that in the United Kingdom significant debates have arisen concerning the place of EBP in the professions. Even the British Medical Journal, where a great deal of EBP advocacy was centered, recently included a series of articles on "narrative based medicine" (for a good example, see Greenhalgh & Hurwitz). Thoughtful critiques also have begun to emerge in British social work, citing EBP's narrow approach to practice and the fact that social work remains contested philosophically. Webb (2001) offers a particularly thought-provoking caution that uncritically embracing the basic assumptions of EBP would be incompatible with the reality of social work practice, which is a reflexive, interactive, and rather unpredictable rather than regular and "rational-technical" process.

Language and EBP

Words like "evidence" and "outcome" are just that, words. They do not point to any pre-existing entities in the world. Rather, their meaning is derived from their use in particular contexts. In situations where being able to provide the grounds for one's beliefs or actions is normative, having "evidence" for a position provides a stronger warrant for that position than not having evidence. Evidence is the name given to a culturally preferred reason for an existential claim or the performance of an action. In the first case evidence functions as proof, in the second as justification.

But a claim of evidence must fit the rules of the social situation in which it is used and be negotiated with those who have the power to legitimate the claim. Therefore, what counts as evidence and the value of different types of evidence tell us much about cultural beliefs and power relations as about what is real. We are familiar with these notions in legal proceedings, where what can be entered "into evidence" is governed by rules and the determination of a judge, and where such decisions may determine the outcome of a case.

These notions are no less important to practice. Definitions of evidence, and their presumed relation to professional practice, support particular assumptive and ideological positions. By restricting the types of information to be used as evidence and defining the relationships among these types of evidence (for example, information from a randomized clinical trial versus practice experience), EBP supports certain practices (for example, those claiming to be "empirical") and undermines others.

Fortunately, because the social work world functions only some of the time like a court room, we are not bound by the meanings ascribed to these terms by any particular group. The value of particular meanings can be contested and alternative meanings and values considered. We might examine EBP from several perspectives. For example, consider the question of who gains and who loses by the adoption of an EBP approach? If you subscribe to the meaning of evidence as used in EBP and its importance to the justification of practice, then you might respond that clients gain and irresponsible practitioners lose. Alternatively, by keeping social work practice a highly complex technology that can be mastered by relatively few experts, adherence to EBP may weed out some incompetent practitioners, but also rid the profession of many people who are doing excellent work, dissuade potentially creative practitioners from entering the field, or create (or render invisible) new kinds of incompetence.

Different sources of evidence (for example, researchers, practitioners, and clients) also might be considered. EBP seems to presume a context in which practitioners do something to clients. This "doing" is considered to be most effective (and ethical), if based on "evidence." But what if practice is viewed as a mutual activity in which what is best (not necessarily effective) is co-generated by clients and practitioners? What is the relative value of different sources and types of evidence in this scenario? (Of course, clients will not ordinarily use the word evidence to justify or prove something, putting them at a disadvantage in a setting where evidence already is afforded high status.) These are merely a few examples of the type of inquiry we encourage social work practitioners to conduct as they wrestle with this important issue.

Conclusion

Evidence-based practice envisions a scientifically based social work that uses the best available evidence to guide practice decisions. It also aligns social work with other, more prestigious professions like medicine. Determining whether these are desirable directions for the profession presents social workers with considerable technical and conceptual challenges. Even social workers who like the idea of EBP need considerable expertise in how to evaluate research design, methodology, and analysis lest they accept on faith others' judgments about the "best evidence" (a position that would be inconsistent with EBP). But such evaluations, although important, are not sufficient. They are confined to criteria presumed important by EBP and may not address other issues that are important to social workers.

EBP, like all approaches, is ideological in the sense that it assumes certain beliefs and adheres to certain values. Understanding these assumptions, beliefs, and values requires examining EBP through a variety of lenses (for example, social reform) and questioning what is taken for granted or considered unproblematic. For example, we might question whether there are parallels between EBP's current affiliation with medicine and social work's affiliation with psychiatry in the early part of the 20th century. What do (did) we gain and what do (did) we lose by this association? Or we might inquire about what EBP-informed social work practice is actually like. How does it position the profession vis-a-vis its social values? What effect will it have on issues of relationship that are so central to social work practice?

It seems to us that social work in today's society is largely reactive in that it responds to social forces for the efficient management of lives rather than building communities, helping people determine their own goals and destinies, and standing with unpopular or unrecognized people. Much of practice is oriented to blending into the mainstream, going with the flow of professional acceptability where the prevalent, if unspoken, values and tasks center on diagnosing and treating people who are personally experiencing society's problems or coping with human conditions. Social work's social control functions have been reflected in the language of "servicing" people, assessing their "needs" for them, and, despite the widespread use of the term "empowerment," in conformity with the conventions of medical and legal descriptions of people and the work of the profession.

Given our view, we want to be mindful that the power of the ideas encompassed in EBP can be used in a variety of ways. These ideas are so compatible with other dominant professional trends and cultural narratives about how people and society function that they may be taken for granted and thus may lead to social work's losing its traditional emphasis on understanding people in context and understanding that individual problems and social problems are inseparable. Is it a coincidence that EBP is favored by managed care providers pushing practice toward an emphasis on specificity in problem identification and rapid responses to the identified conditions? Social work's vision of EBP must not be limited to a medicalized view of social problems and a clinical view of responses to them. Rather, as social workers seek to develop more credible practice, we hope they will be mindful of Gambrill's (2001) challenge to use EBP principles for purposes of mutuality and empowerment. If EBP is to have any lasting value in s ocial work, it will involve helping the profession take a more critical stance to all knowledge, including knowledge grounded in what is considered to be the best available evidence.

Stanley, L. Witkin, PhD, is professor, Department of Social Work, University of Vermont, and W. David Harrison, PhD, is dean, School of Social Work and Criminal Justice, East Carolina University, Greenville, NC.

References

Gambrill, E. (1999). Evidence-based practice: An alternative to authority-based practice. Families in Society, 80, 341-350.

Gambrill, E. (2001). Social work: an authority-based profession. Research on Social work Practice, 11, 166-175.

Goldstein, H. (2000). Joe the King: A study of strengths and morality. Families in Society, 81, 347-350.

Greenhalgh, T., & Hurwitz, B. (1999). Narrative based medicine: Why study narrative. British Medical Journal, 318, 48-50.

Research Triangle Institute. (2000). Assessing "best evidence": Grading the quality of articles and rating the strength of evidence [Online]. Available: ww.rti.org/epc/grading_article.html.

Richmond, M. (1917). Social diagnosis. New York: Russell Sage Foundation.

Sackett, D. L., Rosenberg, W. M. C., Muir Gray, J. A., Haynes, R. B., & Richardson, W. S. (1996). Evidence based medicine: What it is and what it isn't. British Medical Journal, 312, 72-73.

Webb, S. A. (2001). Some considerations on the validity of evidence-based practice in social work. British Journal of Social Work, 31, 57-79.