Potentially Harmful Therapy and Multicultural Counseling: Bridging Two Disciplinary Discourses

Abstract

In recent years psychologists have been increasingly concerned about potentially harmful therapy, yet this recent discourse has not addressed issues that have long been voiced by the multicultural counseling and psychotherapy movement. We aim to begin to bring these seemingly disparate discourses of harm into greater conversation with one another, in the service of placing the discipline on a firmer foothold in its considerations of potentially harmful therapy. After reviewing the two discourses and exploring reasons for their divergence, we argue that they operate according to differing assumptions pertaining to the sources, objects, and scope of harm. We then argue that these differences reveal the discipline’s need to better appreciate that harm is a social construct, that psychotherapy may be inherently ethnocentric, and that strategies for collecting evidence of harm should be integrated with a social justice agenda.

“First, do no harm”: it has long been an ethical priority among mental health practitioners (APA, 2002). And yet, in comparison to the profession’s predominant focus on treatment effectiveness, efforts to systematically identify and prevent potentially harmful or iatrogenic treatment have lagged behind efforts to discover and implement successful treatment (Barlow, 2010). Recent years, however, have seen an upsurge in concern for the identification and repudiation of potentially harmful therapy (PHT), as evidenced by the influence of Scott Lilienfeld’s review article, “Psychological Treatments that Cause Harm,” published in 2007 in Perspectives on Psychological Science. As of June 2014, Lilienfeld’s review has been cited by 190 publications in the PsycINFO database.

This burgeoning literature, however, has emerged in isolation from a longstanding, voluminous discourse of harm in psychology. The multicultural counseling and psychotherapy (MCP) movement, a visible and influential fixture in American psychology since the Civil Rights era, has repeatedly emphasized the potential harm of culturally insensitive or even oppressive therapy practices for diverse individuals (APA, 2003). In this article, we aim to begin to bring these seemingly disparate discourses into greater conversation with one another, in the service of placing the discipline on a firmer foothold in its considerations of potentially harmful therapy. First, we review and compare the PHT and MCP literatures, in terms of their considerations of harmful treatment. Second, we explore possible reasons for the separation of these discourses, concluding that pragmatic specialization or adjudication of evidence are not sufficient reasons. Rather, we argue, the two discourses operate according to differing assumptions about sources, objects, and scope of harm—and these differences point to the discipline’s limited conception of harmful treatment. Finally, we argue that these differences reveal the discipline’s need to better appreciate that harm is a social construct, that psychotherapy may be inherently ethnocentric, and that strategies for collecting evidence of harm should be integrated with a social justice agenda.

The Potentially Harmful Therapy Literature

In certain respects, concern with PHT has a long history. Allen Bergin (1966) sounded an alarm to the profession almost 50 years ago about the incidence of client “deterioration,” and in the years that followed he and other pioneering researchers made important contributions to conceptualizing “negative effects” of treatment (e.g., Strupp & Hadley, 1977). In recent decades, concern with harmful treatment has been occasionally addressed by psychotherapy outcome researchers (e.g., Lambert et al., 2003) or as an implicit appendage of the evidence-based treatment (EBT) movement of the past two decades (e.g., Chambless et al., 1996; McFall, 1991). However, sustained attention to harmful treatment has been lacking and greatly overshadowed by efforts to demonstrate the effectiveness of psychosocial interventions and investigate “what works.” This focus is evident, for example, in pioneering task force statements on EBT, in which harmful treatment is not explicitly addressed (see APA Presidential Task Force on Evidence-Based Practice, 2006; Chambless et al., 1993; cf. Lilienfeld, 2007).

In his 2007 review article, Lilienfeld argued that identification of PHTs ought to be prioritized alongside efforts to implement effective treatments, and that clinicians and researchers are likely to be in agreement about the importance and urgency of such an endeavor. This argument was made on both ethical and empirical grounds, and was contextualized in light of the widespread use of fringe or questionable treatments, combined with the lack of a watchdog for psychosocial services akin to the Food and Drug Administration. Lilienfeld’s review included a list of several PHTs, divided into two tiers (“probably harmful for some individuals” and “possibly harmful for some individuals”), including facilitated communication for autism, rebirthing and holding attachment therapies, D.A.R.E. programs for youth substance abuse prevention, and critical incident stress debriefing. Lilienfeld’s focus on specific treatments is a clear extension of the EBT movement, and his two-tiered categorization resembles early efforts to label EBTs as either “well-established” or “probably efficacious” (e.g., Chambless et al., 2003).

Lilienfeld’s review has arguably “reignited interest in negative effects from psychological treatments,” as evidenced by its many citations, as mentioned above (Barlow, 2010, p. 17). It is beyond the scope of our article to review this voluminous literature, and so we focus on four influential conceptual articles authored by prominent researchers: Lilienfeld’s review and a series of three articles in a 2010 special section of American Psychologist (Barlow, 2010; Castonguay, Boswell, Constanino, Goldfried, & Hill, 2010; Dimidjian & Hollon, 2010). In these latter three articles, the authors expanded on Lilienfeld’s recommendations by addressing specific ways to prevent and identify harmful practice. Barlow (2010) emphasized the need to clinically track harmful treatment effects on individuals, with greater attention to publishing case study reports and tracking clients with rigorous outcome measures. Dimidjian and Hollon (2010) recommended the routine monitoring and reporting of a wide range of unforeseen adverse events, even if it is unclear whether such harm is induced by treatment. And Castonguay et al. (2010) provided training recommendations to foster clinician awareness and prevention of systematic, predictable sources of harm, especially in the realm of psychotherapy processes.

In these articles, harmful treatment was conceptualized in similar ways, generally focusing on iatrogenic harm—that is, harm induced by the treatment or clinician. Lilienfeld (2007) argued that PHTs are harmful above and beyond short-term or innocuous increase in distress (e.g., short-term increase in anxiety due to exposure therapy), or opportunity costs from ineffective treatments; rather, PHTs are directly responsible for substantial and/or enduring harm to clients (e.g., worsening target symptoms or physical injury). Similarly, Dimidjian and Hollon (2010) differentiated harmful from merely unhelpful treatment, underscoring that the former causes clients to be worse than if they had received no treatment. Barlow (2010) conceptualized harm under the label of “negative effects,” leaving open whether such effects necessitate “deterioration during or after treatment or, perhaps, improving less than individuals in an untreated comparison group” (p. 17). Finally, Castonguay et al. (2010) operated from a broadened conceptualization of harmful effects, consisting of “either a decelerated rate of improvement that is the direct effect of the treatment or an opportunity cost reflected in participating in an unhelpful or protracted versus a helpful and parsimonious treatment” (p. 34). Each of these authors stated or implied that harm is often inadvertent from well-meaning clinicians.

Finally, we should stress that these authors each endorsed a lower bar of suitable evidence, relative to establishing EBTs, to warrant the discipline’s attention to potential harm. Lilienfeld (2007) and Dimidjian and Hollon (2010) argued that experimental evidence of harm is very difficult to obtain. Instead, indications of harmful treatment are likely overlooked when an intervention (a) assists clients on average while harming some clients, (b) improves target symptoms while worsening other symptoms less likely to be assessed, (c) benefits clients while harming their family members and friends, and (d) helps in the short term but not the long term, or vice-versa. Moreover, such harm is difficult to experimentally replicate for obvious ethical reasons. As a consequence, PHT researchers have emphasized identifying harm on a case-by-case or idiographic basis. Barlow (2010) argued,

Perhaps it is time to … attend to the responding of each and every individual to avoid burying potentially important negative effects in the group average of clinical trials, whether those negative effects are due to unrelated life events, untoward therapeutic influences, or the direct effect of a given psychological treatment. (p. 18)

Dimidjian and Hollon (2010) gave similar recommendations, adding that “qualitative research also can highlight possible indications of harm and directions for future research” (p. 26). These calls for a lower bar of evidence are consistent with Lilienfeld’s designation of the strongest category of PHTs being “probably harmful for some individuals” (p. 58, italics added).

The Multicultural Counseling and Psychotherapy Literature

The MCP literature has been concerned with harmful treatment for several decades. This movement developed in response to the recognition that “mainstream” interventions had long failed to adequately address the mental health needs of culturally diverse individuals (APA, 2003). Although this concern has been inclusive of many marginalized groups (e.g., women, sexual and religious minorities, and people with disabilities), a predominant focus has been directed to racial, ethnic, and cultural minority clients. For the pragmatic purpose of discussing the vast MCP literature in depth, we limit the breadth of our conceptualization of the MCP movement to this same focus, using the label “ethnoracial minorities.”

Recognition of ethnoracial minority concerns originated in the late 1960s and 1970s, as increasing numbers of minority psychologists entered the field and focused greater attention on limitations of mainstream interventions (Abreu, Chung, & Atkinson, 2000). These developments led to the formation of ethnic minority psychology associations and structural changes in the American Psychological Association (APA) for greater attention to ethnic minority issues. The changes soon materialized in a burgeoning literature on multicultural concerns, especially in the applied professions (Lau, Cisco, & Delgado-Romero, 2008). This literature has been most predominant in academic textbooks and counseling psychology journals, as well as through the creation of specialized journals devoted to ethnoracial minority topics. The MCP movement formed much of the basis of a 2001 report of the U.S. Surgeon General that drew attention to large mental health disparities between ethnoracial minorities and Whites, and its literature also served as the backbone for APA’s (2003) comprehensive multicultural guidelines.

The MCP literature is extensive and multifaceted, and a comprehensive review of such is beyond the scope of this article. Much of this discourse has focused on maximizing treatment utilization and effectiveness for minority clients, such as through racial and linguistic matching of clinicians and clients (see Cabral & Smith, 2011), culturally-adapted therapies (see Bernal & Domenech Rodríguez, 2012), and standards of cultural competence (see S. Sue, Zane, Hall, & Berger, 2009). Another prominent concern, however, has been negative experiences of ethnoracial minority clients who have been treated through mainstream European American interventions. APA’s (2003) multiculturalism guidelines state, for example, that “the traditional Eurocentric therapeutic and interventions models in which most therapists have been trained” may be ineffective or “do harm” to culturally diverse individuals (p. 390). This propensity to harm is perhaps best summarized by D. W. Sue and D. Sue (2008):

Counseling and psychotherapy have done great harm to culturally diverse groups by invalidating their life experiences, by defining their cultural values or differences as deviant and pathological, by denying them culturally appropriate care, and by imposing the values of a dominant culture upon them. (p. 34)

The MCP movement has addressed negative treatment experiences for ethnoracial minorities at multiple levels of analysis and with a diversity of methodological procedures (e.g., epidemiological research, experimental research, outcome research, and case studies). At the broadest level, this discourse documents significant and enduring treatment disparities for ethnoracial minorities in terms of access, availability, and utilization, even after controlling for income, insurance status, and education (e.g., Alegría et al., 2008; Ault-Brutus, 2012; U.S. Surgeon General, 2001). Among those who receive treatment, ethnoracial minorities are less likely than Whites to receive adequate care and more likely to drop out early (Ault-Brutus, 2012; Ridley, 2005; U.S. Surgeon General, 2001). In addition, researchers have repeatedly documented dissatisfaction with treatment and mistrust of clinicians among ethnoracial minorities (see, e.g., Chang & Berk, 2009; U.S. Surgeon General, 2001). These examples do not necessarily involve iatrogenic treatment in a strict sense (as we discuss below), but they are often presented in the context of clearly harmful societal racism and discrimination (see, e.g., Ridley, 2005; D. W. Sue & D. Sue, 2008; U.S. Surgeon General, 2001).

At the level of individual clinicians, the MCP literature has amply documented racist and discriminatory practices. Ridley (2005) cited 132 peer-reviewed journal articles that “have uncovered racism in American mental health care delivery systems” (p. 7; see table on pp. 8–9). Racism is implicated in these articles primarily in terms of actions of White clinicians working with Black clients, especially through overly pathologized diagnoses and assignment to briefer, less intensive, and lower quality interventions. The most frequent documentation of harm in the MCP literature is critical misdiagnosis, especially overdiagnosis of schizophrenia and other psychotic disorders among African American clients (Metzl, 2010; Pavkov, Lewis, & Lyons, 1989). A common reason given for overdiagnosis of ethnoracial minorities is clinicians imposing European American standards when determining “normality” (D. W. Sue & D. Sue, 2008), and thus attributing racial/ethnic differences from a deficit perspective related to Whiteness. For example, African Americans’ personality scores for suspiciousness, mistrust, and paranoia have been commonly misinterpreted as pathological rather than as functional survival mechanisms (Parham, White, & Ajamu, 1999). This overdiagnosis may result in enduring stigma associated with severe mental illness and severe side effects from unnecessary antipsychotic medication.

To a lesser extent, but increasingly so in recent years, the MCP literature has documented mis-attunements and micro-ruptures between clinicians and clients (e.g., Chang & Berk, 2009; Gaztambide, 2012). When pertaining to race, these processes may be categorized broadly as “racial microaggressions,” defined as “brief and commonplace verbal, behavioral, and environmental indignities, whether intentional or unintentional, that communicate hostile, derogatory, or negative racial slights and insults to the target person or group” (D. W. Sue, 2007, p. 273). One common effect of microaggressions is to invalidate or trivialize experiences of minorities, say, by invoking a “color-blind” stance in which racialized experiences purportedly do not matter (Neville, Awad, Brooks, Flores, & Bluemel, 2013; D. W. Sue et al., 2007). Although studies examining racial microaggressions in treatment settings are somewhat sparse, recent studies have shown that client perceptions of racial microaggressions are correlated with lower ratings of therapeutic alliance and treatment satisfaction (e.g., Chang & Berk, 2009; Constantine, 2007), as well as with lower psychological wellbeing and treatment outcomes (Owen et al., 2011).

Divergent Discourses in the Psychological Literature on Harmful Treatment

Given the duration and extent to which the MCP movement has addressed serious concerns regarding treatment of ethnoracial minorities, it is reasonable to expect that the recent PHT discourse would be conversant with at least some of these concerns. This expectation is heightened by the multiple levels of analysis (e.g., specific treatments, clinician processes, and relationship factors) the PHT literature has addressed. However, from our review, there appears to be little if any overlap between the two discourses. Among the four conceptual PHT articles reviewed above (Barlow, 2010; Castonguay et al., 2010; Dimidjian & Hollon, 2010; Lilienfeld, 2007), ethnoracial minority and cultural concerns have not been mentioned, let alone addressed and evaluated. The only exception is Castonguay et al.’s (2010) one-line recommendation for supervisors to help trainees “avoid relationship pitfalls when working with clients from different cultural backgrounds” (p. 45). Furthermore, of the 190 PsycINFO references that cite Lilienfeld’s (2007) article, zero can be identified by their abstracts as explicitly addressing racial, ethnic, and/or cultural concerns in the context of harmful or iatrogenic treatment.

How might this divergence be explained? One possible explanation is that ethnoracial minority client concerns are implied in the conceptualizations and recommendations of the PHT literature. Lilienfeld (2007), for example, addressed clinician and client factors that may moderate harmful treatment, which we assume could include race, ethnicity, and culture. Similarly, Barlow’s (2010) and Dimidjian and Hollon’s (2010) endorsements of idiographic research designs could easily be used to illuminate potential harm from the perspectives of ethnoracial minorities. Finally, Castonguay et al.’s (2010) focus on clinician-client relational concerns could be inclusive of processes such as racial microaggressions. Thus, one might conclude that omission of ethnoracial minority concerns is attributable to a pragmatic focus on general treatment and therapeutic process concerns.

However, this conclusion may seem unsatisfying inasmuch as PHT concerns are framed in ethical terms as a “solemn mandate” (Lilienfeld, 2007, p. 66). As an ethical matter, one might expect more proactive vigilance in ensuring that vulnerable and potentially marginalized populations are not disproportionately harmed. Concern for vulnerable individuals, after all, was a primary motivation and rationale for the Belmont principles that have profoundly shaped clinical practice guidelines and APA’s (2002) code of ethics (Fisher, 2006). This concern is perhaps illustrated most aptly by the notorious Tuskegee study of 400 low-income African American men with syphilis—a major catalyst, along with Nazi experiments during the Shoah (Holocaust), for modern ethical reforms (Evans, 2000). More recent cases, such as exploitative research practices surrounding informed consent with members of the Havasupai American Indian tribal nation (Sterling, 2011), highlight that such concerns are hardly a thing of the past. Although there may possibly be less concern for egregious exploitation in clinical practice as opposed to research, we worry that unless harmful treatment concerns for ethnoracial minority clients are positioned more in the foreground, these concerns may be “off the radar” for researchers and practitioners, as we discuss further below. Moreover, a conclusion of pragmatic specialization is insufficient because the broader EBT movement has included considerable attention to ethnoracial minority concerns (see Hall, 2001; Morales & Norcross, 2010)—including concern for such in the EBT movement’s infancy (e.g., Chambless et al., 1996).

A second possible explanation for these disparate discourses is that the MCP literature lacks adequate evidence of iatrogenic treatment for ethnoracial minorities. Although the MCP movement has repeatedly documented negative treatment experiences in a broad sense, as discussed above (e.g., denial of care or early drop-out, insensitive care, dissatisfaction with treatment, and critical misdiagnoses), it is not as definitive in terms of iatrogenic harm in a more direct or narrow sense (i.e., substantial harm induced by a treatment or clinician). Based on our extensive search (including review of prominent MCP texts, bibliographic database searches, listserv solicitations, and inquiries with multiple MCP experts), the MCP literature has not yet documented that a given sample of ethnoracial minorities was worse off as a result of receiving insensitive or inappropriate treatment. (As far as we are aware, only one attempt has even been made to do so: Cusack et al., 2007, conducted a retrospective study comparing the prevalence and negative outcomes of treatment-induced traumatic events in inpatient psychiatric settings between African Americans and Whites, in which results were generally inconclusive.) Even where the MCP literature has linked racial microaggressions to undesirable outcomes such as lowered therapeutic alliance, treatment satisfaction, or treatment outcomes (e.g., Chang & Berk, 2009; Constantine, 2007; Owen et al., 2011), these events have not been investigated in relation to measurable indices of client deterioration of interest in the PHT literature (e.g., increased symptomatology or trauma). In light of these limitations, one may conclude that MCP scholars put the cart before the horse and therefore PHT scholars are justified—pending more substantial evidence—in not formally addressing their concerns.

This second explanation, however, also seems unsatisfying, in that it appears to contradict the PHT discourse’s broader concern with proactively setting the agenda for needed lines of research. Given the extensive literature showing that ethnoracial minorities have regularly encountered culturally insensitive care and racial microaggressions (cited above)—alongside ample social science research documenting how racism and discrimination can cause mental and physical harm (e.g., D. W. Sue et al., 2007; Williams & Williams-Morris, 2000)—one may reasonably ask why PHT researchers have not assumed the responsibility of advocating for and conducting more conclusive research. This question is especially important given that the PHT literature has missed several opportunities to incorporate evidence-based claims from the MCP movement. For example, Dimidjian and Hollon (2010) stressed the importance of learning why some clients drop out early—as it may be due to iatrogenic treatment—but did not mention the extensive MCP literature on premature termination of psychotherapy at disproportionately high rates by ethnoracial minorities (e.g., U.S. Surgeon General, 2001). Likewise, Castonguay et al. (2010) discussed therapeutic processes (e.g., empathy and therapeutic alliance) at length, even while recognizing that these processes have not usually been empirically linked to outcomes of true deterioration; this discussion could have included racial microaggressions, and yet processes pertaining to race, culture, and ethnicity were not mentioned. These and other examples may indicate that rather than having been formally scrutinized and judged as wanting, ethnoracial minority concerns have been simply off the radar within the PHT literature.

We suggest, then, that the separateness of the PHT and MCP discourses likely extends beyond matters of pragmatic specialization or adjudication of evidence. Certainly, part of the problem is that contributors of the two literatures have different professional affiliations and identities. Based on our review, MCP scholars are more likely to be housed in counseling psychology programs in education departments, whereas PHT scholars appear to be more representative of the recent clinical science movement and housed within arts-and-sciences psychology departments. Moreover, whereas multiculturalism is more likely to be explicitly valued by counseling psychologists (Lau et al., 2008), clinical scientists have emphasized general treatment outcome effects and individual client differences (Hall, 2001). However justifiable these different emphases might be, it would greatly benefit the profession if representatives from both camps engaged each other on the question of harmful treatment. Toward this end, in the next section, we analyze how the two discourses have been operating from differing assumptions in their conceptions of harm.

The Concept of “Harm” in the PHT and MCP Discourses

Before analyzing how the PHT and MCP discourses have conceptualized harm differently, a brief discussion of the meaning of “harm” is in order. Although harm is “widely assumed to be self-evident,” its meaning has existed historically only in relationship to complex and evolving social values (Conaghan, 2002, p. 321). This fluctuating meaning is evident in Sharpe and Faden’s (1998) in-depth historical analysis of iatrogenic treatment in modern medicine, in which they demonstrate that the ethical mandate to do no harm “has been interpreted relative to historical circumstance and the value commitments of individuals and institutions” (p. 81). At the start of the 19th century, a physician’s duty was in the context of the individual patient, although differential emphasis was placed on avoiding harm caused by omission vs. commission of intervention, depending on the treatment philosophy. As scientific understandings of disease emerged in light of statistical probabilities and presumed universal mechanisms, the ethical mandate shifted from the particular patient to the average or statistical patient, with the recognition that some individual patients would inevitably be harmed; this shift had the advantage of protecting physicians from a skyrocketing number of malpractice suits. In the post-World War II era, a utilitarian approach to assessing harm relative to benefit quickly emerged in light of increasingly documented iatrogenic events. Physicians assumed a paternalistic role in determining what counted as beneficial and harmful, in accordance with empirically verifiable standards of “a normal or healthy state of the human organism” (p. 120). Finally, after the advent of the patients’ rights movement of the 1960s and 1970s, a focus on objective standards of beneficence and harm was—after much resistance from the medical profession—balanced with the autonomy, informed consent, and subjective interpretations of the patient as a “unique individual with a particular history, constellation of relationships and values, desired life plan, and moral point of view” (p. 121). In like manner, targeted outcomes evolved from the management of disease to the enhancement of quality of life, including physical functioning, cognitive functioning, social and role functioning, and emotional well-being.

This history shows that considerations of harm are not as self-evident as they might seem; they are multidimensional, historically contingent in relation to changing clinical and societal contexts, and open to interpretation by multiple parties with fluctuating levels of power (see also Strupp & Hadley, 1977, regarding conflicts in appraisals of harm from clinicians, clients, and society). Awareness of the roles of these contexts, interpretations, and power dynamics is helpful in appreciating differing conceptions of harmful treatment between the PHT and MCP discourses. As we discuss, whereas the PHT literature’s treatment of harm is generally congruent with a medical-model frame within the health services, the MCP literature addresses harm primarily from a broader societal context. We summarize these differing conceptions here, according to three dimensions outlined by Sharpe and Faden (1998): sources, objects, and scope of harm.

First, these differing frames are evident in terms of sources of harm. A major focus of the PHT discourse is identifying possible harm to clients in the context of specific treatments. This focus includes identifying potentially harmful mediators of treatments, such as “premature termination of exposure, vicarious exposure to negative role models, and induction of false traumatic memories” (Lilienfeld, 2007, p. 65). To a lesser extent, attention is given to harm from procedural errors (e.g., improper treatment administration or inferential errors), as well as clinician variables (e.g., low empathy) and client variables (e.g., psychopathy) that might be associated with iatrogenic treatment. In contrast, the MCP literature has rarely addressed harm at the level of specific treatments, nor has the focus typically been general psychotherapy processes (an exception is the recent linking of therapeutic alliance with racial microaggressions; e.g., Constantine, 2007; Owen et al., 2011). Rather, the focus is on the interplay of psychotherapy with cultural, social, and institutional factors that contribute to racism, discrimination, and associated ills of particular relevance for ethnoracial minority groups. These factors include “unintentional racism” and “victimization” (Ridley, 2005); “institutional racism,” “racial prejudice,” and “discrimination” (Thompson & Neville, 1999); “cultural oppression” (e.g., D. W. Sue, 1977); and “dominance,” “manipulation,” and “social control” (Hall & Malony, 1983).

Second, there are important differences between the two discourses in their conceptualizations of objects of harm (i.e., who/what is harmed). For the PHT literature, the predominant focus is on the individual client (as is the case for most psychotherapy practice), which lends itself more readily to a medical model. Although brief attention is also given to avoiding harm in others (e.g., friends or relatives of clients; Dimidjian & Hollon, 2010; Lilienfeld, 2007), the object of harm is usually positioned as within generic individuals as opposed to among specific groups, communities, or societies (an exception is Dimidjian and Hollon, 2010, who identified “society as a whole” as an arguable object of harm resulting from “conversion therapy” for gay and lesbian clients; p. 22). In contrast, for the MCP literature, the object of harm is framed more collectively in terms of ethnoracial minority groups. A common approach, for example, is to analyze theory and data at the level of one or more ethnoracial minority groups, primarily African Americans, Latino/as, American Indians/Alaska Natives, and Asian Americans—but also for aggregates and subsets of these groups as well (see U.S. Surgeon General, 2001). This focus on specific groups reflects the realities of marginalization and discrimination that ethnoracial minorities have endured on the basis of out-group categorization.

Finally, the two discourses significantly differ in what they imply about the scope of harmful treatment. “Scope” refers to what is assumed in terms of “some positive or normative conception of proper functioning, well-being or interests” against which a client has experienced “some sort of injury or setback” (Sharpe & Faden, 1998, pp. 119–120). As explained by Strupp and Hadley (1977), “a judgment of ‘worse’ is always made in relation to an implicit or explicit standard, which also presupposes a definition of the meaning of ‘better’” (p. 187). For both the PHT and MCP literatures, scope of treatment is rarely addressed explicitly but can be inferred on the basis of how harmful treatment is discussed. The PHT literature has repeatedly used phrases such as “adverse effects,” “client deterioration,” “negative change,” “decelerated rate of improvement,” and “negative treatment outcomes,” but often with little clarity about just what these effects, changes, and outcomes are. Lilienfeld (2007) is the most explicit, stating that forms of harm may include “symptom worsening, the appearance of new symptoms, heightened concern regarding extant symptoms, excessive dependency on therapists, reluctance to seek future treatment, … and even physical harm” (p. 56). Here a medical model is clearly evident, in that harm is conceptualized predominantly in terms of general symptoms that could presumably apply to any client—by which we can presume that broadly shared professional assumptions about mental health and illness are implicit in the PHT literature’s scope of harm. Conversely, the MCP literature has been much more concerned with specific, contextual aspects of harm that differentially affect ethnoracial minorities. Treatment can be harmful because it may alienate, invalidate, stereotype, or oppress ethnoracial minorities, and thereby reinforce or exacerbate these ills as they occur outside the treatment setting (D. W. Sue et al., 2007). In this way, the MCP literature is informed by a broader social context characterized by legacies of racism, discrimination, and inequality.

Discussion

In light of these differing conceptualizations of harm, how might the PHT and MCP discourses be placed in greater conversation with each other? We offer three recommendations—each of which require critical re-examination of taken-for-granted assumptions in both discourses—in service to greater disciplinary integration of concepts of harm. First, we posit that harm should be explicitly theorized as a social construct, requiring critical reflection about its various meanings and how they might affect diverse constituencies. Second, we argue that mental health interventions should be appreciated as cultural artifacts, and as such may be ethnocentric. Finally, we propose that strategies for collecting evidence of harm should be integrated with a social justice agenda that recognizes that ethnoracial minorities are at potentially greater risk of being harmed in treatment. Space constraints permit us to provide only broad recommendations, leaving in our wake many unsettled questions and thorny dilemmas. Although we strive to anticipate some of these issues, our principal intent is to simply begin a conversation at the juncture of the PHT and MCP literatures; it is beyond our scope to attempt to resolve these questions and dilemmas.

Harm is a Social Construct

The history of modern medicine, as reviewed above, suggests that “harm” does not have a straightforward, objective, or obvious meaning. Rather, harm is a social construct that is interpreted and negotiated, however informally, in relation to norms about well-being, clinical treatment, and social relations. As recognized by both PHT and MCP scholars, harm is often conceptualized differently by individuals and groups with competing values and interests. A given treatment can be seen as harmless by the profession—and even corroborated as such by various outcome measures—while still being experienced as harmful by clients, their families, coworkers, communities, and/or cultures. Professional considerations of harm are further complicated by dynamic social and economic factors, such as clinicians’ and institutions’ worries about litigation for allegedly iatrogenic treatment, the ability of clients or third parties to pay for services, and legislation relevant to clinical services (Sharpe & Faden, 1998).

Both the MCP and PHT discourses can improve in their attention to harm as a social construct. The MCP movement, to be sure, is sensitive to the social construction of harm, in light of it having routinely deconstructed ethnocentric norms of much clinical practice (see, e.g., APA, 2003; D. W. Sue & D. Sue, 2008). Indeed, a great strength of the multiculturalism literature is a focus on social context, as evidenced by its having explored ethnocentrism and racism from multiple situations and perspectives, and in the context of both micro (e.g., intrapsychic) and macro (e.g., societal and organizational) processes (see D. W. Sue, Ivey, & Pedersen, 1996). Nonetheless, MCP scholars could facilitate greater communication with the PHT literature—or, better yet, contribute to this literature themselves—by also explicitly addressing medical model considerations, which are reflective of an arguably inescapable social context of harm relative to individuals seeking psychosocial services in the U.S. We recommend for MCP scholars to clarify, both conceptually and empirically, any relationship of racist and discriminatory practices (e.g., racial microaggressions) to worsened psychopathology in individual clients (e.g., increased depression and anxiety symptoms, decreased self-esteem or self-confidence, or increased hypervigilance) after receiving specific forms of treatment. We also recommend for MCP scholars to engage more vigorously in empirical research that seeks to identify specific mechanisms and moderators of harm in the context of treatment. These research approaches, while not precluding attention to broader dimensions of harm, would greatly contribute to the discipline’s understanding of iatrogenic treatment.

In contrast, the PHT discourse, while being clearer in its conceptualizations of iatrogenic treatment, has been more limited in exploring harm in terms of broader social values and societal context. Although acknowledgement has been given to the existence of “value-laden considerations,” including “extrascientific factors, such as judgments about the appropriate goals of psychotherapy” (Lilienfeld, 2007, pp. 55–56), the PHT literature has seemingly assumed that overt articulation of these values and factors is not necessary for addressing iatrogenic treatment. Lilienfeld (2007), for example, did not disclose or articulate any value-laden considerations underlying his empirical review of PHTs, even while proceeding to identify treatments that “have demonstrated harmful psychological or physical effects in clients or others (e.g., relatives)” (p. 57). A possible problem with this approach is that any conceptualization of harm, including an individualized focus using empirical data, relies at the outset on social values and “extrascientific” considerations. If these values and considerations are not recognized and explicitly identified, then they are at greater risk of reflecting taken-for-granted societal assumptions. In the case of the PHT literature, we worry that an implicit, supreme value of the generic individual is unintentionally crowding out concerns of harm towards vulnerable and marginalized groups.

In order to address these shortcomings, we recommend for psychological researchers and clinicians to conceptualize harmful treatment with explicit attention to harm as a social construct, including both individual and collective sources and objects of harm. Such conceptualizations ought to be reflective of medical-model treatment considerations as well as political realities and power structures about who decides what is harmful in a given situation. Questions of course would remain about how to navigate conflicting conceptions of harm, as well as what might be done to avoid and minimize such harm, but this broad recommendation may at least help researchers and clinicians to better recognize that no consideration or investigation of harm is value-free, and that assumptions of value-neutrality may inevitably be associated with overlooking unique concerns of culturally diverse and marginalized clients and communities.

Psychological Interventions May Be Ethnocentric

By being upfront that harm is a social construct, the discipline is better prepared to recognize ways that psychological interventions, as cultural artifacts, may be harmful owing to ethnocentrism. Interventions are not only practical tools to help individuals with distress, but carry with them cultural assumptions concerning the nature of the self, how one relates to and interacts with others and the cosmos, and the nature of illness and cure (Kirmayer, 2007). As several critical observers have noted, mainstream psychotherapeutic interventions typically are based on a Western cultural concept of the self as “agentic, rationalistic, monological, and univocal,” and thus rely heavily on socialization to a Westernized worldview of individuality and interiorized identity and control (Kirmayer, 2007, p. 240; cf. Cushman, 1995). Such socialization is construed as benign “psychoeducation” for clients with deficient “psychological mindedness,” but there may be a fine line in some instances between pragmatic education and ethnocentric proselytization. Indeed, some researchers have considered evidence that clinicians act as “crypto-missionaries” (Meehl, 1959, p. 257) by subtly or inadvertently “converting” their clients to their own professional, social, and even moral values (see, e.g., Slife, Smith, & Burchfield, 2003; Tjeltveit, 1986).

Both the PHT and MCP discourses can improve in their recognition of potential harm owing to ethnocentrism. The PHT literature addresses multiple dimensions and aspects of harm, including the importance of idiographic research and the recognition that certain clients may be more likely to be harmed. It is seemingly silent, however, concerning any possible relationship between harmful treatment and ethnocentric therapists or therapies. On the other hand, the MCP movement has long recognized that ethnoracial minority clients have been vulnerable to clinician promotion of the values and behaviors of the dominant culture—arguing, for example, that mainstream psychotherapy “often is a form of social control toward majority norms” (Hall & Maloney, 1983, p. 139) or even an “opiate or instrument of oppression” (Pinderhughes, 1973, p. 99). However, a review of the MCP literature suggests a predominant focus on treatments that are usually adapted from conventional therapies (see Griner & Smith, 2006; Smith, Domenech Rodríguez, & Bernal, 2011; van Loon, van Schaik, Dekker, & Beekman, 2013), preserving and reproducing the embedded orientations and assumptions of these treatments in ways that may attenuate the movement’s overt critiques of ethnocentrism. Although this focus has been extremely valuable in successfully improving outcomes on target symptoms for ethnoracial minorities, it has perhaps not fully appreciated the extent to which even adapted treatments partake of dominant cultural sensibilities in pervasive fashion (e.g., through engagement in reflexive talk-based services, rendered in confidential spaces, situated in indoor clinics or hospitals, administered by credentialed professionals, for a precise duration of time, in accordance with a guiding theoretical orientation, with reference to a particular diagnostic nomenclature, and for which the primary targets of intervention are the thoughts, feelings, beliefs, and/or behaviors of an individual “client”; see Wendt & Gone, 2012). Furthermore, multicultural clinicians may themselves be prone to unwittingly impose certain cultural values (e.g., individualist ideals of happiness, productivity, open-mindedness, secularism, and tolerance) on clients with differing values (Fowers & Richardson, 1996; Slife et al., 2003).

We suggest that more thorough and consistent critical concern with the cultural encapsulation of professional psychological interventions—including aspects of such that may be harder for professionals of all ethnoracial backgrounds to see—may be needed in order for the discipline to begin to grapple with potential harm from ethnocentric therapies. Of course, we recognize that change is what therapy is all about, and this may defensibly include some clients changing deeply-held cultural values. We simply recommend greater attention to and empirical documentation of when and how cultural values of clients change over the course of therapy, whether these changes come about through a clinician’s unquestioned values, and whether these changes are seen as harmful to the needs of ethnoracial minority individuals and communities.

One fruitful way for the discipline to better understand ethnocentrism in mainstream treatment is to give greater consideration and empirical investigation of more radically cultural-divergent healing approaches that may be more likely to be used by ethnoracial minorities. These include partnerships with indigenous healers, churches, and community organizations in designing and delivering culturally-centered interventions (e.g., Aten, Topping, Denney, & Bayne, 2010); integration of various indigenous approaches, such as Mestizo spirituality, African-centered frameworks, and Reiki and Qigong healing, into psychotherapy (e.g., Cervantes, 2010; Parham et al., 1999; Yeh, Hunter, Madan-Bahel, Chiang, & Arora, 2004); and “culture-as-treatment” approaches, involving intensive experiential and educational efforts to return to one’s cultural roots (e.g., Gone & Calf Looking, 2011). Of course, critical questions remain about the therapeutic efficacy and potential harm of some if not most of these alternative interventions. However, it would behoove the discipline to recognize potential advantages of these interventions, such as less risk of covert cultural proselytization and greater access to and choice of treatment. Perhaps most important, greater attention to these alternative interventions can help the discipline to recognize through contrast how mainstream interventions may be ethnocentric.

Evidence of Harm Should Be Linked to Social Justice

As reflected in both the PHT and MCP discourses, it is not easy to empirically document iatrogenic treatment. This difficulty is compounded in the context of ethnoracial minority groups. Consider, by way of analogy, the extent to which EBTs cannot always be generalized to ethnoracial minority groups: Minorities often are not represented in clinical trials, and therefore the generalizability of outcomes in too many instances remains an open empirical question (Miranda et al., 2005); however, it is unfeasible to expect that treatments will be validated with every conceivable demographic category (Kazdin, 2009). In comparison, evidence of harm towards minority groups would be even more difficult to demonstrate, irrespective of its prevalence, in light of the difficulty of firmly establishing instances of harmful treatment. We especially stress—in light of concern that interventions may benefit clients on average or with target symptoms, while harming other individuals or in peripheral domains (Dimidjian & Hollon, 2010; Lilienfeld, 2007)—that aggregate treatment effectiveness for ethnoracial minorities does not mean that some ethnoracial minorities are not being harmed.

An important corrective for these concerns, as advocated in the PHT literature, is a lower bar of evidence for establishing PHTs (relative to EBTs), including greater attention to idiographic research such as case studies and interviews (see above). Idiographic research, with its ability to reveal harmful treatment for problems or clients that are not easily measured or represented in nomothetic trials, is a promising paradigm for beginning to demonstrate harm for ethnoracial minorities. Inasmuch as researchers and clinicians follow Lilienfeld’s (2007) recommendation to “report the full scope of outcomes so that readers could draw their own conclusions regarding the presence of harmful effects” (p. 56; cf. Dimidjian & Hollon, 2010), one could reasonably expect that evidence for harm towards ethnoracial minorities would be better understood.

However, if an idiographic research paradigm is not connected with explicit attention to ethnoracial minority concerns, then subtle forms of culturally-related harm may be overlooked, for at least two reasons. First, culturally-related forms of harm, such as racial microaggressions or covert cultural proselytization, are often subtle and may require asking clients about them directly and in a culturally competent manner in order for them to be revealed (see D. W. Sue & D. Sue, 2008). Specific efforts may need to be taken to observe or inquire about harmful effects relevant to the full scope of clients’ lives, such as their relationships with their families, communities, and cultural groups. Second, ethnoracial minority clients may be less likely to be included in case studies unless active efforts are made for their inclusion. Consider a recent 12-article special issue in Cognitive and Behavioral Science presenting individual case studies of treatment failure (Dimidjian & Hollon, 2011): cultural diversity concerns were conspicuously absent and clients’ races were either listed as White (N=7) or were undisclosed (N=4).

In light of these concerns, we recommend for empirical investigations of potential harm to be linked more often with a social justice agenda, resulting in more proactive searches for potential harm that may be overlooked or unduly dismissed due to the marginalized status of many ethnoracial minorities. Case studies and qualitative research designs are promising starting points, so long as ethnoracial minorities are intentionally included. One recommended preliminary research approach—requiring an integration of the two discourses—is to interview minority clients who report having had negative treatment experiences. Even if these experiences are retrospective, this research may be an important first step to understand how minority clients talk about and make meaning of harm, and thus may serve to document the existence of alleged harm and/or reveal previously ignored dimensions of harm. Such an approach was used by Shidlo and Shroeder (2002) in the context of self-defined harm among 202 gay and lesbian clients who had received conversion therapy. Although this design was unable to demonstrate the prevalence of harm or indicate specific ethical violations, the study was used as evidence in APA’s (2009) resolution against sexual orientation change efforts. We are surprised that a similar study has not already been conducted in the context of ethnoracial minority concerns, given the MCP movement’s concern with harm from mainstream mental health services (concerning limitations of the MCP literature’s research on iatrogenic treatment, see above).

We also recommend for the discipline to more proactively raise the banner of potential harm for vulnerable populations such as ethnoracial minorities. As we have discussed, much is unknown about the extent and consequences of harmful treatment for ethnoracial minorities. Some may suggest that limitations of this evidence mean that our call for greater integration of the PHT and MCP discourses is premature—and we certainly advocate for empirical research to proceed apace in the documentation of iatrogenic treatment for ethnoracial minorities. However, a conservative burden-of-proof approach, according to a high standard of evidence, is not only inconsistent with the broad concern and anticipatory stance of the PHT literature (reflected in the qualifier “potentially”)—it could forestall timely attention to potential harm for ethnoracial minorities, irrespective of the extent to which it may be occurring. Given the likely low base rates of iatrogenic treatment, the discipline’s predominant focus on aggregate treatment outcomes, the marginalized status of research funding on socio-cultural rather than intra-individual factors (Nezu, 2005), and continued underrepresentation of ethnoracial minority concerns in the psychological literature (Hartmann et al., 2013), we anticipate that progress for such an empirical endeavor will be slow. This uphill climb is all the more reason for taking seriously, here and now, our call for a serious conversation between these divergent discourses.

One possible action in this regard is for the discipline to immediately identify certain forms of racist or discriminatory practices (e.g., racial microaggressions) as potentially harmful. Such a response is defensible, we would argue, in light of the considerable research on the prevalence and negative effects of racist and discriminatory practices in society (as cited above). Moreover, assuming that these practices are viewed as repugnant by the vast majority of therapists and their clients, we foresee little if any downside to this swift, proactive action. (It is worth noting, in this regard, the work of Norcross, Koocher, and Garofalo, 2006, in establishing expert consensus of “discredited psychological treatments and tests,” irrespective of empirical evidence.) Clearly, the designation of such practices as potentially harmful involves value-laden considerations, but such is the case with any consideration of harm, as we have argued above. How the discipline proceeds in these matters, then, is inescapably a reflection of its values in regards to proactive concern for marginalized individuals and groups, not simply a matter of tough-minded scientific skepticism.

Conclusion

Both the PHT and MCP discourses contribute to psychology’s understanding of harmful treatment. However, the two discourses are not in conversation with each other, and they operate according to differing conceptions of the sources, objects, and scope of harm. Whereas the PHT literature has been consistent with a medical model of iatrogenic harm, the MCP literature has been concerned with the interplay between broader societal factors (e.g., racism and discrimination) and the treatment setting. Integration of the two literatures requires the discipline’s critical interrogation of taken-for-granted assumptions, in better appreciating that harm is a social construct, mental health interventions as cultural artifacts may be ethnocentric, and evidence of harm should be integrated with a social justice agenda. In light of the ethical mandate to do no harm and the heightened vulnerability of ethnoracial minorities receiving treatment, it is crucial and urgent for adherents of both discourses to be in greater conversation with one another. We hope at least for this article to serve as an invitation for future engagement.