Tailoring Interventions to Clients: Effects on Engagement and Retention

Tailoring Interventions to Clients: Effects on Engagement and Retention Larry E. Beutler, Heidi Zetzer, and Elizabeth Yost Drug abuse represents a major social and behavioral health problem. National Institute of Drug Abuse (NIDA) statistics suggest that 37 percent of the U.S. population has used illicit drugs (Smith 1992) and as many as 23 percent of the work force regularly do so in the workplace (Barabander 1993). Drug use in the workplace is not limited to benign, recreational drugs. The NIDA estimates indicate that 3 percent of the work force abuse heroin (Browne 1986) and over twice that figure abuse cocaine (Abelson and Miller 1985). Chemical abuse affects family functioning, work performance, and the health of newborn children (Levy and Rutter 1992). Of contemporary concern, opiate and cocaine abuse probably represent the most significant problem because of their severe addictive properties, the high likelihood of polydrug abuse among their users, and the decline of social functioning that accompanies their abuse (Almog et al. 1993). Three tasks face those who attempt to develop treatment programs for drug abuse disorders: (1) developing procedures that facilitate patient engagement in treatment, (2) developing procedures that increase the likelihood of retention of individuals in treatment, and (3) establishing the conditions under which even effective treatments work best and least well. While this chapter will focus on the status of research on the first two of these tasks, the last one cannot be ignored in this process. This is true both because of the necessity of continually testing the relationship between treatment engagement and dropout on one hand and treatment efficacy on the other, and because there are promising developments in the area of treatment efficacy that may improve awareness of the significance of engagement and retention. There is considerable contemporary interest in tailoring aspects of the intervention to fit individual patient needs. Individualized treatments offer the hope that sensitivity and specificity of interventions will increase treatment retention and engagement rates as well as improve the likelihood of clinical efficacy. Unfortunately, literature is sparse on the application of tailored interventions to chemical abuse problems. Thus, efforts to tailor handling of patients who suffer from these problems must borrow and extract from research on other problems. Two basic methods of matching treatments and patients have shown enough promise in psychotherapy research to suggest that they may be 85 transportable to research on chemical abuse treatment. The first of these methods focuses on matching personal qualities (e.g., personalities, backgrounds, beliefs, and other characteristics) of patients and therapists to increase the likelihood that they will develop a working relationship that will enhance therapeutic progress. The second seeks to match specific treatment procedures or strategies to particular characteristics and needs of patients. A brief review of the status of these matching procedures may help in the assessment of their applicability to substance abuse treatment research. PATIENT-THERAPIST MATCHING AND THERAPEUTIC OUTCOMES One way to tailor interventions to clients of different types is to match them to therapists who represent the most appropriate fit. Such patienttherapist matching has a long tradition in psychotherapy, covering the gamut of matching variables from sociodemographic characteristics to value and belief systems. Indices of treatment engagement have included patient preferences and satisfaction levels. One line of research, for example, has suggested that some ethnic minority clients prefer to work with counselors who are of the same race, but this finding tends to vary according to the client’s ethnicity and his or her level of ethnic identity development and acculturation (Parham and Helms 1981; Sanchez and Atkinson 1983). Summaries of this research usually conclude that while African-American clients prefer African-American therapists over white therapists (Atkinson 1983; Harrison 1975; Sattler 1977), a preference for an ethnically similar therapist is not as evident among other ethnic minority groups (Atkinson 1983). Atkinson (1983) observed that inattention to within-group differences and constricted ranges of acculturation might have prevented the emergence of preferences as a distinguishing effect of ethnic matching among nonblack minority groups. Sanchez and Atkinson (1983) remedied some of these problems by including level of acculturation as an independent variable in their investigation of racial similarity and therapist preference. They found that Mexican-American college students with strong commitments to the Chicano culture preferred racially similar counselors, while those who did not have strong cultural identities did not. Interestingly, however, Vietnamese refugees (presumably not acculturated) showed no preference for a racially similar counselor (Atkinson et al. 1984). Somewhat stronger evidence has accumulated to suggest that patient-therapist ethnic similarity is associated with lower dropout rates during therapy (Krebs 1971; Terrell and Terrell 1984; Yamamoto et al. 1967). 86 Finally, therapist-patient socioeconomic status (SES) similarity has been related to more positive perceptions of the therapist, but the relationship between SES and either therapy process or outcome has not been investigated adequately. Despite evidence that therapeutic engagement (e.g., therapist preference and treatment retention) is enhanced by the assignment of a racially similar therapist to minority patients, the extent to which this similarity or dissimilarity affects engagement in the therapy process is far from clear (Atkinson and Schein 1986). About half the studies examined by Atkinson and Schein (1986) support the hypothesis that patient-therapist ethnic similarity can enhance certain qualities of the therapy process, including the working relationship, and at times even lead to improved therapy outcomes. The other half of the research studies in those authors’ review found no relationship among these factors. It may be, however, that ethnic similarity is not the important variable in studying the cultural fit of patients and therapists. Perhaps a broader dimension of shared cultural values would provide a better test of the role of ethnicity and background in psychotherapy change among certain disenfranchised groups. In the most systematic effort to address this latter issue, Smith and colleagues (1980) assigned an index of cultural similarity to each of the 475 studies in their meta-analysis of psychotherapy outcome, based upon the degree to which client and therapist samples shared a common educational, economic, and upwardly mobile history. An effect size of only 0.10 was obtained using this variable, suggesting that little variance in outcomes could be attributed to similarity of culturally derived attitudes. There is also little evidence in contemporary literature to indicate that gender similarity, age similarity, or physical ability similarity have an effect on psychotherapeutic engagement, relationship enhancement, or treatment outcome (Atkinson and Schein 1986), though patients, regardless of their status on these dimensions, tend to prefer female (Stricker and Shafran 1983), middle-aged (Simon 1977), and physically disabled therapists (Brabham and Thoreson 1973; Mitchell and Frederickson 1975). Collectively, these findings suggest both that different ethnic groups respond differently to demographically similar therapists and that sharing certain attitudes, rather than simply demographic background, may be more important in facilitating patient engagement and commitment. This con-clusion may be seen as supportive to the position, often taken in chemical abuse treatment programs, that therapists who share the 87 patient’s history of substance abuse may be better able to provide help than therapists who do not have a history of chemical abuse. The results of research on this topic have shown some support for this view, at least for enhancing the therapeutic alliance (Argeriou and Manohar 1978; Lawson 1982). For example, Lawson (1982) found that counselors who were in alcohol recovery were judged by their alcoholic patients to show higher regard and greater unconditionality than counselors without alcohol problems. Results regarding attrition and outcome in these treatment programs are not available. Nonetheless, this work raises the importance of considering patient and therapist fit along dimensions of personality and attitude. Viewing literature on broad personality dimensions reveals evidence of some relationship between the degree of patient-therapist similarity and the strength of the therapeutic relationship (Atkinson and Schein 1986). Though not replicated extensively in contemporary research, similar personality styles, as measured by the Myers-Briggs Type Indicator, have been associated with improved treatment retention rates (Mendelsohn and Geller 1963, 1965). However, treatment duration does not necessarily indicate treatment benefit. For example, Swenson (1967) found that dissimilarities on the specific dimensions of dominance and submission were directly related to treatment gains. Likewise, several reviews conclude that dissimilar, rather than similar, personality traits are associated with better outcomes and greater patient satisfaction levels (Atkinson and Schein 1986; Beutler 1981). Therapist conceptual level is another personality trait that has been the focus of both theoretical and empirical work over several years. Unlike findings related to dominant and submissive traits, however, this research generally supports the value of patient-therapist similarity. Similarity in conceptual level has been particularly related to retention in treatment (Holloway and Wampold 1986; Lamb 1977; McLachlan 1972; Stein and Stone 1978). For example, Hunt and colleagues (1985) explored the effects of cognitive style match among consecutive admissions to the University of Washington Psychiatric Outpatient Clinic. They found that 60 percent of the premature terminations occurred in mismatched or dissimilar dyads while only 24 percent occurred in dyads with similar cognitive styles. Among clients who continued therapy, similar dyads were associated with more symptomatic changes at the end of 12 weeks of treatment than dissimilar dyads, an effect that disappeared quickly when mismatched groups caught up to the similar group. These findings suggest that similarity of cognitive style facilitates retention in therapy and may even speed improvement in the early sessions of therapy. 88 As the contradictory findings from these studies on conceptual level and dominance-submissiveness would suggest, it is difficult to generalize from any one variable within the domains of personality and belief systems to others, even within similar domains. No single conclusion is warranted that applies to a broad range of beliefs or personality characteristics regarding whether similarity or difference is more facilitative of retention, engagement, or outcome in psychotherapy. However, a relatively consistent theme does emerge to suggest that increasing patient-therapist similarity on a wide variety of these variables over the course of therapy is associated both with indicators of engagement and improvement. For example, Foon (1985, 1986) reported that among a diagnostically heterogeneous group of 78 adult outpatients and their 21 therapists, end-of-therapy similarity of locus of perceived control, but not pretreatment similarity, was positively associated with improvement, indicating that convergence of client-therapist perceptions is a factor in achieving benefit. Patient-therapist convergence has been a particularly consistent observation in research on patient belief and value systems, a conclusion that partially reflects and has been supported by a relatively large series of studies from the authors’ laboratory (Beutler et al. 1991a). Six critical qualitative reviews of this research have all concluded that clients tend to adopt the personal values of their therapists during the course of successful psychotherapy, independently of the type of problem presented (Atkinson and Schein 1986; Beutler 1981; Beutler and Bergan 1991; Beutler et al. 1991b; Kelly 1990; Tjelveit 1986). According to these reviewers, patients who become converted to the therapist’s beliefs or values have good outcomes. A review of the six most methodologically sound studies on this topic (Kelly 1990) suggests that value conversion may be related most closely to therapist ratings of improvement, but the effect, nevertheless, does extend to the outcomes assigned by external raters and the patients themselves. A second generalizable, but somewhat weaker, conclusion also emerges from this literature. Three of the six reviews (Beutler 1981; Kelly 1990; Tjelveit 1986) inspected the relationship between initial therapist-client similarity and subsequent improvement, noting that initial dissimilarity of client and therapist values was associated with the subsequent adoption of the therapists’ values and beliefs. None of the six qualitative reviews reached a clear determination as to whether initial global value similarity or dissimilarity is more conducive to client improvement. Likewise, Foon (1985, 1986) found that initial patient-therapist similarity on the dimension of perceived locus of control was not predictive of treatment outcome. It appears that while 89 global value and personality convergence are associated with positive outcomes, and while initial dissimilarity on these global dimensions is associated with convergence, neither global value/belief similarity nor dissimilarity are consistent predictors of treatment response (Beutler 1981; Tjelveit 1986). Some effort has been devoted to looking at and comparing specific value and personality characteristics within the patient-therapist dyad that will identify a pattern of initial similarities and dissimilarities that will relate to positive treatment outcomes. Such a pattern, if identified, could serve as a template for assigning patients to therapists with whom they would work well. Some relatively weak but promising conclusions seem to be emerging from this literature. The findings indicate that psychotherapy improvement may be enhanced by a complex pattern of similarity and dissimilarity between client and therapist belief and value systems (e.g., Beutler et al. 1974; Cheloha 1986). In the authors’ laboratory, a series of studies has suggested that treatment success is enhanced when clients and therapists are similar in the relative value placed upon such qualities as wisdom, honesty, intellectual pursuits, and knowledge (e.g., Arizmendi et al. 1985; Beutler et al. 1974). At the same time, client-therapist discrepancies in the value placed on personal safety (Beutler et al. 1978), interpersonal values, social status and friendships (Arizmendi et al. 1985; Beutler et al. 1974, 1983) have been found to facilitate improvement. At least some of these findings have been supported by independent research programs (Charone 1981; Cheloa 1986). In sum, while demographic similarity and conceptual level appear to be related to dropout, value conversion and a complex pattern of individual belief and value similarities and dissimilarities appear to be related to improvement. To date, however, none of these lines of research has inspected retention and improvement rates within drug-abusing populations. Typically, patient samples have been diverse outpatient groups with very heterogeneous problem types. The research to date is promising, especially in its implications to retention rates. It provides a fertile field in which treatment retention rates among chemical abuse patients might be explored. Clearly, more research on the types of similarity dimensions that are positively, nonsignificantly, and negatively associated with different types of retention and improvement will be necessary to apply these promising relationships to problems of chemical abuse. MATCHING PROCEDURES TO PATIENT NEEDS 90 Efforts to define various psychotherapy methods that are effective either for those with a common diagnosis, such as drug abuse, or with a specific symptom, such as depression, have largely concluded that all psychotherapy approaches produce similar mean effects (e.g., Beutler et al. 1986; Lambert et al. 1986; Smith et al. 1980). The studies that support these conclusions are largely based on a randomized clinical trials methodology that has been borrowed from psychopharmacological research. The methodology of this research paradigm is to study a single, diagnostically homogeneous sample of patients and to compare the efficacies of one or more packaged, reliably applied, and brand-named treatments. In this paradigm, nondiagnostic patient characteristics are usually studied as a secondary, post hoc variable when, as is usually the case, no significant differences are observed between two packaged treatments. Depression and its various subtypes have been the target of most studies of psychotherapy. Fortunately, there are reasons to believe that there are links between depression and substance abuse that will allow this body of research to transfer to chemical-abusing populations. For example, depres- sion is a frequent coexisting condition both in drug abuse and during drug withdrawal (Weiss et al. 1992). Whether cause or consequence, the coexistence of drug abuse and depressive symptoms suggests that psycho- behavioral interventions that are effective in treating depression also may be effective treatments for chemical abuse. Because of the demands of randomized clinical trials research, a number of well-established treatments of depression also have evolved, many of which are potentially transportable to the area of chemical abuse. For example, mounting evidence suggests that even in the case of endogenous depression, the condition most often thought to be weighted toward the role of biological precipitators, manualized forms of cognitive therapy are effective in both relieving depression and preventing its recurrence (Corbishley et al. 1990; Jarrett et al. 1990; Simons and Thase 1992). Moreover, cognitive therapy, though initially formulated and manualized as a treatment for depression (Beck et al. 1979), in recent years has successfully been adapted as a treatment both for drug abuse (Wright et al. 1993) and alcoholism (Wakefield et al., in press). Similarly, treatment manuals based upon relationship-oriented therapies (both psychodynamic and interpersonal models) have been successfully extrapolated from research on the treatment of depression and anxiety disorders for application in the treatment of opiate abuse (McLellan et al. 1983; Rounsaville et al. 1987). The use of manuals within the context of the usual clinical trials research model is well adapted to revealing which 91 systematic therapies are effective in treating specific diagnostic groups, but is of limited value for assessing questions of matching treatments to patients. Nonetheless, the translations of established manuals to chemical- abusing populations have provided a foundation for explorations of the conditions under which different psychotherapies and psychotherapy procedures are maximally effective. The effort to fit treatments—variously called "eclectic," "integrative," and "prescriptive" psychotherapies—to patients has evolved largely in the last decade. Two approaches to matching patients to treatments have been employed in these prescriptive models. One has been to develop different manualized therapies for patients with different diagnostic conditions. In this approach, an effort is made to construct a theoryconsistent therapy that can be applied in a somewhat different form to several different and diagnostically distinct patient groups. The foundation studies for this method usually concentrate on demonstrating the clinical efficacy of each within the patient samples for which it was designed rather than on comparing the efficacy of the different manuals. A second approach to matching patients to treatments has been to define characteristics of treatment procedure and strategy that distinguish different theoretical approaches to psychotherapy, and then to identify the patient characteristics on which these procedures are differentially effective. The foundation studies for this approach to prescriptive matching have been those in which two or more manualized therapies are applied to two or more patient groups. Rather than being selected solely on diagnostic grounds, the patient groups for this approach usually are stratified on the basis of a variable that is thought on empirical or theoretical grounds to be differentially responsive to the therapies studied. Differentiating aspects of each treatment are related to differential efficacy on the diverse samples of patients, yielding conclusions about those treatment characteristics that best fit the patient characteristics. In these studies, the patient characteristics of interest are often extradiagnostic in nature. While the first approach works within a single theoretical system to develop variations that fit different diagnostic groups, a major aim of the latter approach is to develop guidelines for mixing and combining procedures from across theoretical models to maximally tailor interventions to specific patient characteristics that are not captured well in diagnosis. Both types of studies provide leads to indicate the patient and treatment dimensions that will make the most effective matches. 92 STATUS OF INTEGRATIVE TREATMENTS To compare the relative value of the two prescriptive treatment methodologies described above, Beutler and Crago (1987) compared studies that used a variety of methodologies to calculate the percentages of explained, within-subject variance accounted for by each approach— different treatment models applied to patients with different diagnoses versus contrasting treatment models applied to nondiagnostic patient variables. The value of the two approaches was assessed against a base rate expectation of 10 percent, the amount of variance attributable to different treatment types when patient variables are not considered. The base rate figure of 10 percent was derived from a variety of research reviews of comparisons of different psychotherapy models (Lambert 1989; Lambert and DeJulio 1978; Smith et al. 1980). The comparisons indicated that the interaction effects attributable to combinations of psychotherapy types by patient diagnoses increased the amount of outcome variance accounted for from 10 to 15 percent, a very modest increase over the base rate of 10 percent. This finding confirmed the suggestion (Howard 1989) that even in manualized treatments of diagnostically homogeneous patient groups, the variability of outcomes among treatments is very broad. In any defined and uniformly applied treatment, there appears to be a relatively large number of patients who get better and a smaller but substantial number who do not. Apparently, diagnostic variables are insufficient to reduce the wide variance in outcomes that are secured by all treatments. When Beutler and Crago considered studies that matched patients and treatments in ways other than through patient diagnosis or brand-named therapies, they had more success in establishing the presence of differential effects among treatments. Matching treatments to select nondiagnostic variables increased the amount of attributable variance to an average of 30 percent, and some variables accounted for as much as 60 percent of the variance in patient responses. Some patient variables that have been found to interact most successfully with treatment procedures are coping styles (Beutler 1979; Beutler and Mitchell 1981; Sloane et al. 1975), levels of resistance (Beutler et al. 1991c, 1991d; Shoham-Salomon and Hannah 1991), cognitive organization (McLachlan 1972), and aspects of problem severity and distress (Imber et al. 1990; Luborsky et al. 1985). Promising Matching Dimensions 93 Drawing from studies representing each of the foregoing approaches, an inspection of the most promising findings suggests that: • Experiential therapies often are more effective than cognitive and dynamic therapies either when used early in treatment or when applied to those who are insufficiently distressed about their problems to support emotional growth (Beutler and Mitchell 1981; Greenberg and Safran 1987; Mohr et al. 1990; Orlinsky and Howard 1986); • Nondirective and paradoxical interventions are more effective than therapist-directed ones among patients with high levels of pretherapy resistance (i.e., resistance potential or reactance) (Beutler et al. 1991c, 1991d; Forsyth and Forsyth 1982; Shoham-Salomon and Hannah 1991); and • Therapies that target cognitive and behavior changes are more effective among impulsive, externalizing patients than those that attempt to facilitate insight. The latter effect has often proven to be reversed among patients with internalizing coping styles (Beutler et al. 1991c, 1991d; Calvert et al. 1988; Sloane et al. 1975). All of these relationships have been found to be sufficiently robust to be revealed in a variety of diagnostic disorders. Thus, they carry implications for the prescription of psychotherapeutic strategies and procedures that are extracted and combined across theoretical models (see Beutler and Consoli 1992; Beutler and Hodgson 1993; Gaw and Beutler 1995). The robust effect of many of these parameters suggests that combined treatments, based upon these relationships, may be applicable to chemical abuse dependencies as well as to a variety of mental health disorders. On a more negative note, even the best among the available research studies on this topic have oversimplified the complexity of matching patients and treatments. With few exceptions, the long-term effects of these variables on relapse and efficacy have not been investigated. Neither has research, to date, investigated the interdependence and joint effects of two or more patient and treatment dimensions operating at once. The best studies include only one or two manualized variations of therapy procedure as applied to a group of patients who are selected to vary along a single dimension. No study to date has had the resources to address the implications of assigning a treatment package composed of several interventions to patients who vary on several indicators/ contraindicators at once. 94 The importance of considering multiple patient and treatment parameters at once cannot be overstated. Even manualized treatments, if they are not sensitive to the complexities of individual proclivities and the treatments with which they fit (and this includes most contemporary manuals), may unintentionally include treatment components for a given patient that are offsetting when applied to a patient who embodies a constellation of characteristics that do not fit. For example, the positive effects of cognitive therapy applied to an externalizing patient may be offset by the limited effective-ness of this same therapy with patients who have high levels of resistance to therapist leadership or control. Thus, comparisons of two or more treatments may fail to reveal important differences within diagnostically homogeneous groups because these groups include patients who have counterbalancing but unassessed differences in indicating characteristics. Extracting from research on the patient-treatment dimensions previously identified, it is conceivable that treatment outcomes could be maximized among substance abusers by fitting specific procedures from several different treatment models to the unique combination of extradiagnostic characteristics that is presented by the individual patients (e.g., combining the symptom focus of cognitive therapy with the arousal-induction procedures of relationship-oriented therapies for use with an externalizing, nondistressed patient). This is an area where research is needed. Selecting the most promising combinations of procedures from among those valued by the several hundred available theories (Corsini 1981) requires that several patient and treatment dimensions be varied at once. The complexity of patient and treatment variables (e.g., Lazarus 1981; Orlinsky and Howard 1986; Parloff et al. 1978) makes it unlikely that studies in which one brand of psychotherapy is pitted against another, without regard for the patient characteristics that fit and fail to fit with these treatments, will yield much information about treatment efficacy. ILLUSTRATIVE RESEARCH For descriptive purposes, findings that have been obtained and the implications of some patient-therapy matching components will be illustrated. For example, in a prospective test of the independent effects of two matching dimensions among depressed outpatients, Beutler and colleagues (1991c) selected manualized therapies that contrasted in defined ways to provide greater and lesser fits to patients varying in coping style and resistance potential. 95 Following 20 sessions of treatment, therapies that were directive and therapist-guided were found to have opposite effects from a therapy that was designed to be self-directed and nondirective when contrasting patient groups were studied. Resistance-prone patients did poorly in the directive therapies but well in the nondirective/self-directed therapy, while patients who were not prone to high levels of resistance did comparatively better with the directive therapies than the nondirective one. This finding was independently crossvalidated on a sample of anxious and depressed patients at the University of Bern utilizing a variety of alternative measures of defensive anxiety (Beutler et al. 1991d). Corollary work (Horvath 1989; Seltzer 1986; Shoham-Salomon et al. 1989; Shoham-Salomon and Rosenthal 1987) has confirmed the conclusion that trait-like indicators of resistance may be a specific indicator for the use of nondirective, paradoxical, and self-help procedures. Shoham-Salomon and colleagues (1989) have demonstrated that college students who were predicted to be highly resistant by voice tone measures became worse when they were directly told to change habits of procrastination. However, paradoxical assignments (i.e., "observe but don’t change your habits") resulted in a decrease in symptoms. Confirmations in different populations (Shoham-Salomon and Jancourt 1985) suggest that either low directive or paradoxical (don’t change) instructions are indicated for patients judged to have high propensities for resistance, while directive assignments are contraindicated (Forsyth and Forsyth 1982). Studying a second dimension, Beutler and colleagues (1991c) found that a therapy that was designed to directly induce a change in symptoms of depression worked more effectively among patients who had adopted impulsive and other externalizing coping styles, but it was relatively ineffective among those who were prone to internalize conflict and to be excessively self-reflective. In contrast, therapies that were designed to provoke self-awareness and insight worked best among internalizing individuals but relatively less well among those who externalized their conflicts and acted out. Outcomes. Outcome analyses revealed that two of three outcome measures were affected differently across treatments as a function of patient coping style. The symptom-focused procedures of cognitive therapy exerted their strongest effects among depressed patients who exhibited externalizing coping styles. Less impulsive, depressed patients did best with the insight-oriented procedures of a client-centered therapy. This finding was subsequently independently crossvalidated on a sample of 96 depressed and anxious subjects at the University of Bern, utilizing a variety of alternative measures of externalization (Beutler et al. 1991c). Relapse. Another finding emerged from these studies that is important to the development of matching dimensions. Beutler and colleagues (1993) tabulated followup data from the completed study of depression. They found that while relapse rates for depressed patients in all treatments were very low (averaging 12 percent) over a 1-year period, return of clinically significant depressive symptom levels was a function of interactional matches between type of treatment and patient characteristic. Externalizing patients in cognitive therapy had lower relapse rates than either externalizing patients in the other treatment conditions or than nonexternalizing patients in cognitive therapy. In contrast, nonexternalizing patients in two insight-oriented psychotherapies had lower relapse rates than nonexternalizing patients in cognitive therapy. Likewise, high-resistance patients in directive, cognitive, and experiential therapies relapsed at a higher rate than resistant patients in a nondirective therapy. Low-resistance patients in directive therapies also relapsed at a relatively low rate. APPLICATIONS TO SUBSTANCE ABUSE With the exception of Project MATCH (alcohol abuse), randomized clinical trials of psychotherapy for either depression or substance abuse have not mirrored the integrative effort that characterizes clinical practice. Kazdin (1983, 1986) has suggested that comparative studies of packaged psychotherapies must be accompanied by dismantling and combining strategies to refine the potency of interventions. The authors believe that a variety of psychotherapies have now adequately demonstrated their clinical efficacy and that more specific combined strategies are now needed to define the dimensions of differential treatment selection. However, without prospective, hypothesis-driven research designs, the accumulation of empirical knowledge is likely to be slow (Goldfried and Padawer 1982). There have been several interesting studies of substance abuse that have used aspects of treatment matching to look at ways of enhancing treatment efficacy. Most notable among these, from the authors’ perspective, have been those of Kadden and colleagues. Kadden and colleagues (1990) evaluated the relative effects of interactional (insight/interpersonal therapy) and skills training (symptomatic/ behavioral) aftercare groups among 96 inpatient alcoholic patients who were differentiated by their propensities toward sociopathic behaviors (an externalizing quality). Interactional therapy proved to be most 97 efficacious among those with low sociopathic qualities—more internalizing—while behavioral skills training was most efficacious among those with high sociopathic qualities. After 2 years, results were obtained that paralleled the findings of Beutler and colleagues (1991d) on patients with major depression. Cooney and coworkers (1991) found that longterm relapse rates were also associated with matches of therapy and patient types. Correctly matched groups produced less relapse and better long-term gains than did poorly matched patients. Though consistent with the research on depression, these latter findings have not received universal support among substance abusers. For example, Woody and colleagues (1985) found that sociopathy did not differentiate between those patients in a methadone maintenance program who responded to cognitive therapy and those who responded to an insight-oriented therapy. In this study, however, the measure of sociopathy was categorical and diagnosis specific rather than being continuous and symptomatic. Moreover, it was obtained by clinician ratings rather than self-report. Thus, it is likely that the elements of coping style reflected in this measure were different and more diagnosis specific than measures used in studies of coping style. Woody and colleagues (1983) did find some results that bear on the effort to discover indicators and contraindicators for types of intervention. They investigated the role of problem severity and level of impairment as a con-tributor to differential outcomes of psychotherapies versus drug counseling. They found that methadone-maintained, opiate-abusing patients with severe problems did better in professionally run psychotherapy programs. In contrast, those with less severe problems of opiate abuse were able to benefit from less intensive drug counseling. Alterman and colleagues (1991) have incorporated these findings and others into a systematic set of suggestions for the differential treatment of substance abusers. The authors’ research group (Beutler et al. 1993) is currently implementing a research program that was designed to demonstrate the advantages of matching patient characteristics with psychotherapeutic techniques. The program compares the differential effectiveness of family systems couples therapy (Rohrbaugh et al. 1995) and cognitivebehavioral couples therapy (Wakefield et al., in press) in treating men or women with a primary "Diagnostic and Statistical Manual of Mental Disorders," 3d ed. revised (DSM-III-R) (American Psychiatric Association 1987) diagnosis of alcohol abuse or dependence. The participants are engaged in 20 sessions of treatment with their partners over a period of 6 months. In addition to evaluating the efficacy of the two treatments in 98 reducing or eliminating alcohol intake, improving the quality of the couple’s relationship, and alleviating psychological symptomatology, the authors are also examining the differential effects of treatment for men with two different drinking styles (episodic versus steady), two different coping styles (internalizing versus externalizing), and varied levels of interpersonal reactance. It is hoped that the importance of some of these variables in both the selection of systems- versus symptom-focused treatments and the application of other intervention strategies that distinguish the treatments will be confirmed. This study reflects the authors’ belief that the field is ripe for developing methods of combining treatment procedures across theoretical models. By combining aspects of different treatments into a single treatment package based upon the patient indicators revealed in these studies, it may be possible to improve treatment efficacy far over that obtained using manualized, single-theory models. In support of this conclusion, several recent theoretical and methodological articles have appeared in the literature advocating a search for treatment by patient interaction dimensions. A special series in the April 1991 "Journal of Consulting and Clinical Psychology" was devoted to aptitude by treatment interaction (ATI) in psychotherapy and posed some methodological suggestions as well as advocating for an integrative, conceptual position. Likewise, several textbooks of eclectic and integrative models of psychotherapy have been published in recent years and are well received (Norcross 1986, 1987; Norcross and Goldfried 1992; Striker and Gold 1993). This level of activity indicates that this is both an exciting and fruitful area of investigation, and one with many potential applications, including applications to drug and alcohol abuse. Recommendations In this chapter, treatment research studies on a variety of patient and diagnostic conditions have been reviewed in an effort to find some dimensions that may be extrapolated to the treatment of substance abuse disorders. The focus of the review has been on efforts to match patients either to specific therapists or to types of psychosocial treatments to reduce dropout rates and increase treatment-related gains. It is impressive that there is a growing body of research demonstrating meaningful, but largely extradiagnostic, differences in the types of patients for whom different treatment strategies and methods are effective. Moreover, the advent of treatment manuals from randomized clinical trials research, and the demonstration that some treatment strategies and procedures from these manuals are more effective than others when 99 applied to distinguishable patient groups, have laid the foundation for combining some of the procedures used in a variety of treatments in order to tailor therapies to the needs of different chemical abusers. Collectively, the review of literature suggests several promising directions for future research. 1. Patient-therapist similarity on various aspects of background and demographic variables appears to slow the rate and frequency of premature termination. Some of these variables, most notably gender and ethnic similarity, may also contribute to reductions in focal symptoms. In the case of drug abuse, this literature suggests that retention in treatment and declining use of drugs may be enhanced by selecting and assigning therapists whose backgrounds are similar to those of patients. The mechanism of this action is uncertain, but at least conceptually it is associated with the patient’s ability to identify with the therapist and to find the therapist to be a credible and believable individual. 2. Aspects of patient and therapist dissimilarity may also be important, especially for facilitating symptomatic change. For example, the patient’s ability to accept and adopt the therapist’s general view of life appears to be associated with improved functioning, especially as rated by the therapist. Concomitantly, the presence of contrasting attitudes and values between therapist and patient seems most conducive to the emergence of this conversion process. 3. Certain patterns of initial patient-therapist similarity and dissimilarity of viewpoint and personality also are conducive to facilitating improvement, irrespective of the process of attitude convergence. Similarity of cognitive conceptual level, social values, and intellectual values, combined with dissimilarity of interpersonal needs for closeness or ascendence, appear to be an optimal pattern. 4. Defining a fit between patient characteristics and therapy procedures also appears to be possible. Among the best studied patient-therapy dimensions are the effects of matching patient coping style to the symptom or insight orientation of the therapy. Impulsive, charactero-logical patients seem to do best in behavioral and cognitive therapies, while overcontrolled, internalizing patients do best with insight-oriented therapies. 5. Similarly, it appears that patients who have strong tendencies to resist external control through oppositional behaviors do best when treated 100 with nondirective and paradoxical therapies. Conversely, patients who exhibit more cooperative and less resistant reactions to external demands are likely to benefit from therapies led and directed by the therapist. 6. Therapies that combine a number of procedures from several different models in order to accommodate both the patient’s coping style and level of interpersonal resistance may be maximally effective. This cross-theory eclecticism may entail a number of other dimensions, as well, with the expectation that it may have increasing effects on symptom reduction. Combining patient-therapist assignment and patient-therapy treatment selection may create an opportunity to both decrease dropout rates and to increase treatment efficacy. REFERENCES Abelson, H.I., and Miller, J.D. A decade ends in cocaine use in the household population. In: Kozel, M.S., and Adams, E.H., eds. Cocaine Use in America: Epidemiologic and Clinical Perspectives. National Institute on Drug Abuse Research Monograph 61. DHHS Pub. No. (ADM)87-1414. Washington, DC: Supt. of Docs., U.S. Govt. Print. Off., 1985. Almog, Y.J.; Anglin, M.D.; and Fisher, D.G. Alcohol and heroin use patterns of narcotic addicts: Gender and ethnic differences. Am J Drug Alcohol Abuse 19:219-238, 1993. Alterman, A.I.; O’Brien, C.P.; and McLellan, A.T. Differential therapeutics for substance abuse. In: Frances, R.J., and Miller, S.I., eds. Clinical Textbook of Addictive Disorders. New York: Guilford Press, 1991. pp. 369-390 American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. 3d ed. rev. Washington, DC: American Psychiatric Association, 1987. Argeriou, M., and Manohar, V. Relative effectiveness of nonalcoholics and recovered alcoholics as counselors. J Stud Alcohol 39:793799, 1978. Arizmendi, T.G.; Beutler, L.E.; Shanfield, S.; Crago, M.; and Hagaman, R. Client-therapist value similarity and psychotherapy outcome: A microscopic approach. Psychother Theory Res Practice 22:16-21, 1985. Atkinson, D. Ethnic similarity and counseling psychology. Couns Psychol 11:79-92, 1983. Atkinson, D.R.; Ponterotto, J.G.; and Sanchez, A.R. Vietnamese and AngloAmerican EOP students’ attitudes toward counseling. J College Student Personnel 25:448-452, 1984. 101 Atkinson, D.R., and Schein, S. Similarity in counseling. Counsel Psychol 14:319-354, 1986. Banikiotes, P.G., and Merluzzi, T.V. Impact of counselor gender and counselor sex role orientation on perceived counselor characteristics. J Counsel Psychol 28:342-348, 1981. Barabander, C.S. Alcohol and drugs in the workplace. In: Straussner, S.L.A., ed. Clinical Work with Substance-Abusing Clients. New York: Guilford Press, 1993. Beck, A.T.; Rush, A.J.; Shaw, B.F.; and Emery, G. Cognitive Therapy of Depression. New York: Guilford Press, 1979. Beutler, L.E. Toward specific psychological therapies for specific conditions. J Consult Clin Psychol 47:882-897, 1979. Beutler, L.E. Convergence in counseling and psychotherapy: A current look. Clin Psychol Rev 1:79-101, 1981. Beutler, L.E.; Arizmendi, T.G.; Crago, M.; Shanfield, S.; and Hagaman, R. The effects of value similarity and clients’ persuadability on value convergence and psychotherapy improvement. J Soc Clin Psychol 1:231-245, 1983. Beutler, L.E., and Bergan, J. Value change in counseling and psychotherapy: A search for scientific credibility. J Counsel Psychol 38:16-24, 1991. Beutler, L.E.; Clarkin, J.; Crago, M.; and Bergan, J. Client-therapist matching. In: Snyder, C.R., and Forsyth, D.R., eds. Handbook of Social and Clinical Psychology: The Health Perspective. New York: Pergamon Press, 1991b. Beutler, L.E., and Consoli, A.J. Systematic eclectic psychotherapy. In: Norcross, J.C., and Goldfried, M.R., eds. Handbook of Psychotherapy Integration. New York: Basic Books, 1992. pp. 264-299. Beutler, L.E., and Crago, M. Strategies and techniques of prescriptive psychotherapeutic intervention. In: Hales, R.E., and Frances, A.J., eds. American Psychiatric Association: Annual Review. Vol. 6. Washington, DC: American Psychiatric Association, 1987. Beutler, L.E.; Crago, M.; and Arizmendi, T.G. Therapist variables in psychotherapy process and outcome. In: Garfield, S.L., and Bergin, A.E., eds. Handbook of Psychotherapy and Behavior Change. 3d ed. New York: John Wiley and Sons, 1986. Beutler, L.E.; Engle, D.; Mohr, D.; Daldrup, R.J.; Bergan, J.; Meredith, K.; and Merry, W. Predictors of differential and self directed psychotherapeutic procedures. J Consult Clin Psychol 59:333-340, 1991c. Beutler, L.E.; Engle, D.; Shoham-Salomon, V.; Mohr, D.C.; Dean, J.C.; and Bernat, E.M. University of Arizona Psychotherapy Research 102 Program. In: Beutler, L.E., and Crago, M., eds. Psychotherapy Research: International Programmatic Studies. Washington, DC: American Psychological Association, 1991a. Beutler, L.E., and Hodgson, A.B. Prescriptive psychotherapy. In: Striker, G., and Gold, J., eds. Comprehensive Handbook of Psychotherapy Integration. New York: Plenum, 1993. pp. 151-164. Beutler, L.E.; Jobe, A.M.; and Elkins, D. Outcomes in group psychotherapy: Using persuasion theory to increase treatment efficiency. J Consult Clin Psychol 42:547-553, 1974. Beutler, L.E.; Machado, P.P.P.; Engle, D.; and Mohr, D. Differential patient X treatment maintenance of treatment effects among cognitive, experiential, and self-directed psychotherapies. J Psychother Integration 3:15-31, 1993. Beutler, L.E., and Mitchell, R. Psychotherapy outcome in depressed and impulsive patients as a function of analytic and experiential treatment procedures. Psychiatry 44:297-306, 1981. Beutler, L.E.; Mohr, D.C.; Grawe, K.; Engle, D.; and MacDonald, R. Looking for differential effects: Cross-cultural predictors of differential psychotherapy efficacy. J Psychother Integration 1:121-142, 1991d. Beutler, L.E.; Patterson, K.M.; Jacob, T.; Shoham, V.; Rohrbaugh, M.; and Yost, E. Matching treatment and alcoholism subtypes. Psychotherapy 30:463-472, 1993. Beutler, L.E.; Pollack, S.; and Jobe, A.M. "Acceptance," values and therapeutic change. J Consult Clin Psychol 46:198-199, 1978. Brabham, R.E., and Thoreson, R.W. Relationship of client preferences and counselor’s physical disability. J Counsel Psychol 20:10-15, 1973. Brooks, V.R. Sex and sexual orientation as variables in therapist’s biases and therapy outcomes. Clin Soc Work J 9:198-210, 1981. Browne, A.C. Drug and alcohol abuse among employees: Critical issues. Employee Assistance Q 2:13-22, 1986. Calvert, S.J.; Beutler, L.E.; and Crago, M. Psychotherapy outcome as a function of therapist-patient matching on selected variables. J Soc Clin Psychol 6:104-117, 1988. Cantor, D.W. Women as therapists: What we already know. In: Cantor, D.W., ed. Women as Therapists: A Multitheoretical Casebook. New York: Springer Publishing Co., 1991. pp. 3-19. Charone, J.K. Patient and therapist treatment goals related to psychotherapy outcome. Diss Abst Int 42:365B, 1981. Cheloha, R.S. The relationship between client-therapist mental health, value similarity, and psychotherapy outcome. Diss Abst Int 47:171bB., 1986. 103 Cooney, N.L.; Kadden, R.M.; Litt, M.D.; and Getter, H. Matching alcoholics to coping skills or interactional therapies: Two-year followup results. J Consult Clin Psychol 59:598-601, 1991. Corbishley, M.A.; Beutler, L.E.; Quan, S.; Bamford, C.; Meredith, K.; and Scogin, F. Rapid eye movement density and latency and dexamethasone suppression as predictors of treatment response in depressed older adults. Current Ther Res 47:856859, 1990. Corsini, R.J. Handbook of Innovative Psychotherapies. New York: John Wiley and Sons, 1981. Foon, A.E. Similarity between therapists’ and clients’ locus of control: Implications for therapeutic expectations and outcome. Psychotherapy 22:711-717, 1985. Foon, A.E. Locus of control and clients’ expectations of psychotherapeutic outcome. Brit J Clin Psychol 25:161-171, 1986. Forsyth, N.L., and Forsyth, D.R. Internality, control ability, and the effectiveness of attributional interpretation in counseling. J Counsel Psychol 29:140-150, 1982. Gaw, K.F., and Beutler, L.E. Integrating treatment recommendations. In: Beutler, L.E., and Berren, M., eds. Integrative Assessment of Adult Personality. New York: Guilford, 1995. pp. 280-319. Goldfried, M.R., and Padawer, W. Current status and future directions in psychotherapy. In: Goldfried, M.R., ed. Converging Themes in Psychotherapy. New York: Springer, 1982. pp. 3-49. Greenberg, L.S., and Safran, J. Emotion in Psychotherapy. New York: Guilford, 1987. Harrison, D.K. Race as a counselor-client variable in counseling and psychotherapy: A review of the research. Counsel Psychol 5:124-133, 1975. Hart, L.E. An investigation of the effect of male therapists’ views of women on the process and outcome of therapy with women. Diss Abst Int 42:2529B, 1981. Holloway, E.L., and Wampold, B.E. Relation between conceptual level and counseling-related tasks: A meta-analysis. J Counsel Psychol 33:310-319, 1986. Horvath, A. "There Are No Main Effects, Only Interactions." Paper presented at the annual meeting of the Society for Psychotherapy Research, Toronto, Ontario, Canada, June 1989. Howard, K.I. "Patient and Therapy Interaction Effects in Drug Abuse Research." An invited paper presented at the National Institute on Drug Abuse, Washington, DC, May 1989. 104 Hunt, D.D.; Carr, J.E.; Dagodakis, C.S.; and Walker, E.A. Cognitive match as a predictor of psychotherapy outcome. Psychotherapy 22:718721, 1985. Imber, S.D.; Pilkonis, P.A.; Sotsky, S.M.; Elkin, I.; Watkins, J.T.; Collins, J.F.; Shea, M.T.; Leber, W.R.; and Glass, D.R. Mode-specific effects among three treatments for depression. J Consult Clin Psychol 58:352-359, 1990. Jarrett, R.B.; Rush, A.J.; Khatami, M.; and Roffwarg, H.P. Does the pretreatment polysomnogram predict response to cognitive therapy in depressed outpatients? A preliminary report. Psychiat Res 33:385-399, 1990. Kadden, R.M.; Cooney, N.L.; Getter, H.; and Litt, M.D. Matching alcoholics to coping skills or interactional therapies: Posttreatment results. J Consult Clin Psychol 57:698-704, 1990. Kazdin, A.E. Treatment research: The investigation and evaluation of psychotherapy. In: Hersen, M.; Kazdin, A.E.; and Bellack, A.S., eds. Clinical Psychology Handbook. New York: Pergamon, 1983. pp. 265-284. Kazdin, A.E. The evaluation of psychotherapy: Research design and methodology. In: Garfield, S.L., and Bergin, A.E., eds. Handbook of Psychotherapy and Behavior Change. 3d ed. New York: John Wiley, 1986. Kelly, T.A. The role of values in psychotherapy: Review and methodological critique. Clin Psychol Rev 10:171-186, 1990. Krebs, R.L. Some effects of a white institution on black psychiatric outpatients. Am J Orthopsychiat 41:589-596, 1971. Lamb, K.L. Matching persons and counseling environments to foster selfcontrol of test anxiety. Diss Abs Int 38:3890B-3891B, 1977. Lambert, M.J. The individual therapist’s contribution to psychotherapy process and outcome. Clin Psychol Rev 9:469-485, 1989. Lambert, M.J., and DeJulio, S.S. "The Relative Importance of Client, Therapist and Technique Variables as Predictors of Psychotherapy Outcome: The Place of ‘Nonspecific’ Factors." Paper presented at the mid-winter meeting of the Division of Psychotherapy, American Psychological Association, March 1978. Lambert, M.J.; Shapiro, D.A.; and Bergin, A.E. The effectiveness of psychotherapy. In: Garfield, S.L., and Bergin, A.E., eds. Handbook of Psychotherapy and Behavior Change. 3d ed. New York: John Wiley, 1986. pp. 157-211. Lawson, G. Relation of counselor traits to evaluation of the counseling relationship by alcoholics. J Stud Alcohol 43:834-839,1982. 105 Lazarus, A.A. The Practice of Multimodal Therapy. New York: McGraw- Hill, 1981. Levy, S.J., and Rutter, E. Children of Drug Abusers. New York: Lexington Books, 1992. Luborsky, L.; McLellan, A.T.; Woody, G.E.; O’Brien, C.P.; and Auerbach, A. Therapist success and its determinants. Arch Gen Psychiat 42:602-611, 1985. McLachlan, J.C. Benefit from group therapy as a function of patient-therapist match on conceptual level. Psychother Theory Res Prac 2:317-323, 1972. McLellan, A.T.; Luborsky, L.; Woody, G.E.; O’Brien, C.P.; and Druley, K.A. Predicting response to alcohol and drug abuse treatments. Arch Gen Psychiat 40:620-625, 1983. Mendelsohn, C.A., and Geller, M.H. Effects of counselor-client similarity on the outcome of counseling. J Counsel Psychol 10:71-77, 1963. Mendelsohn, C.A., and Geller, M.H. Structure of client attitudes toward counseling and their relation to counselor-client similarity. J Consult Psychol 29:63-72, 1965. Mitchell, D.C., and Frederickson, W.A. Preferences for physically disabled counselors in hypothetical counseling situations. J Counsel Psychol 22:477-482, 1975. Mohr, D.C.; Beutler, L.E.; Engle, D.; Shoham-Salomon, V.; Bergan, J.; Kaszniak, A.W.; and Yost, E. Identification of patients at risk for non-response and negative outcome in psychotherapy. J Consult Clin Psychol 58:622-628, 1990. Norcross, J.C., ed. Handbook of Eclectic Psychotherapy. New York: Brunner/Mazel, 1986. Norcross, J.C., ed. Casebook of Eclectic Psychotherapy. New York: Brunner/Mazel, 1987. Norcross, J.C., and Goldfried, M.R., eds. Psychotherapy Integration. New York: Basic Books, 1992. Orlinsky, D.E., and Howard, K.I. Process and outcome in psychotherapy. In: Garfield, S.L., and Bergin, A.E., eds. Handbook of Psychotherapy and Behavior Change. 3d ed. New York: John Wiley, 1986. pp. 311-384. Parham, T.A., and Helms, J.E. The influence of Black students’ racial identity attitudes on preference for counselor’s race. J Counsel Psychol 28:250-257, 1981. Parloff, M.B.; Waskow, I.E.; and Wolfe, B.E. Research on therapist variables in relation to process and outcome. In: Garfield, S.L., and Bergin, A.E., eds. Handbook of Psychotherapy and Behavior Change. 2d ed. New York: John Wiley, 1978. pp. 233-282. 106 Rohrbaugh, M.; Shoham, V.; and Spungen, C. Family systems therapy in practice: A systematic couples therapy for problem drinking. In: Bongear, B., and Beutler, L.E., eds. Comprehensive Textbook of Psychotherapy. New York: Oxford University Press, 1995. pp. 228-253. Rounsaville, B.J.; Dolinsky, Z.S.; Babor, T.F.; and Meyer, R.E. Psychopathology as a predictor of treatment in alcoholics. Arch Gen Psychiat 44:505-513, 1987. Sanchez, A.R., and Atkinson, D.R. Mexican-American cultural commitment, preference for counselor ethnicity, and willingness to use counseling. J Counsel Psychol 30:215-220, 1983. Sattler, J.M. The effects of therapist-client racial similarity. In: Gurman, A.S., and Razin, A.M., eds. Effective Psychotherapy: A Handbook of Research. New York: Pergamon Press, 1977. pp. 252-290. Seltzer, L.F. Paradoxical Strategies Psychotherapy: A Comprehensive Overview Guidebook. New York: John Wiley, 1986. Shoham-Salomon, V.; Avner, R.; and Neeman, R. "You are changed if you do and changed if you don’t": Mechanisms underlying paradoxical interventions. J Consult Clin Psychol 57:590598, 1989. Shoham-Salomon, V., and Hannah, M.T. Client-treatment interactions in the study of differential change processes. J Consult Clin Psychol 59:217-225, 1991. Shoham-Salomon, V., and Jancourt, A. Differential effectiveness of paradoxical interventions for more versus less stress-prone individuals. J Counsel Psychol 32:443-447, 1985. Shoham-Salomon, V., and Rosenthal, R. Paradoxical interventions: A metaanalysis. J Consult Clin Psychol 55:22-27, 1987. Simon, W.E. Age, sex, and title of therapist as determinants of patients’ preferences. J Psychol 83:145-149, 1977. Simons, A.D., and Thase, M.E. Biological markers, treatment outcome, and 1year follow-up in endogenous depression: Electroencephalographic sleep studies and response to cognitive therapy. J Consult Clin Psychol 60:392-401, 1992. Sloane, R.B.; Staples, F.R.; Cristol, A.H.; Yorkston, N.J.; and Whipple, K. Psychotherapy Versus Behavior Therapy. Cambridge, MA: Harvard University Press, 1975. Smith, P.H. The political economy of drugs: Conceptual issues and policy options. In: Smith, P.H., eds. Drug Policy in the Americas. Boulder, CO: Westview Press, 1992. Smith, M.L.; Glass, G.V.; and Miller, T.I. The Benefits of Psychotherapy. Baltimore: Johns Hopkins University Press, 1980. Stein, M.L., and Stone, G.L. Effects of conceptual level and structure on initial interview behavior. J Counsel Psychol 25:96-102, 1978. 107 Striker, G., and Gold, J.R., eds. Comprehensive Handbook of Psychotherapy Integration. New York: Plenum Press, 1993. Stricker, G., and Shafran, R. Gender and psychotherapy: A review of empirical literature. In: Murray, J., and Abramson, P.R., eds. Bias in Psychotherapy. New York: Praeger, 1983. Swenson, C.H. Psychotherapy as a special case of dyadic interaction: Some suggestions for theory and research. Psychother Theory Res Prac 4:7-13, 1967. Terrell, R., and Terrell, S. Race of counselor, client sex, cultural mistrust level, and premature termination from counseling among black clients. J Counsel Psychol 31:371-375, 1984. Tjelveit, A.C. The ethics of value conversion in psychotherapy: Appropriate and inappropriate therapist influence on client values. Clin Psychol Rev 6:515-537, 1986. Wakefield, P.; Yost, E.; Williams, R.I.; and Patterson, K.M. CognitiveBehavioral Therapy of Alcoholic Couples. New York: Guilford, in press. Weiss, R.D.; Griffin, M.L.; and Mirrin, S.M. Drug abuse as self-medication for depression: An empirical study. Am J Drug Alcohol Abuse 18:121-129, 1992. Woody, G.E.; Luborsky, L.; McLellan, A.T.; O’Brien, C.P.; Beck, A.T.; Blaine, J.; Herman, I.; and Hole, A. Psychotherapy for opiate addicts: Does it help? Arch Gen Psychiat 40:639-645, 1983. Woody, G.E.; McLellan, A.T.; Luborsky, L.; and O’Brien, C.P. Sociopathy and psychotherapy outcome. Arch Gen Psychiat 42:10811086, 1985. Wright, F.D.; Beck, A.T.; Newman, C.F.; and Liese, B.S. Cognitive therapy of substance abuse: Theoretical rationale. In: Onken, L.S.; Blaine, J.D.; and Boren, J.J., eds. Behavioral Treatments for Drug Abuse Dependence. National Institute on Drug Abuse Monograph 137. NIH Pub. No. 93-3684. Washington, DC: Supt. of Docs., U.S. Govt. Print Off., 1993. Yamamoto, J.; James, Q.; Bloombaum, M.; and Hattem, J. Racial factors in patient selection. Am J Psychiat 124:630-636, 1967. ACKNOWLEDGMENT Work on this chapter was supported by grant no. AA 08970 from the National Institute on Alcoholism and Alcohol Abuse awarded to Larry Beutler (PI), Theodore Jacob, and Varda Shoham (co-PIs). 108 AUTHORS Larry E. Beutler, Ph.D. Professor of Education and Psychology Heidi Zetzer, Ph.D. Project Coordinator Elizabeth Yost, Ph.D. Visiting Associate Professor Couples Alcoholism Treatment Project Counseling/Clinical/School Psychology Program Department of Education University of California Santa Barbara, CA 93106 Click here to go to page 110 109 View publication stats