Strategies for training counselors in evidence-based treatments.

Evidence-based treatments (EBTs) for substance abuse and dependence have demonstrated superiority over treatment as usual when applied with strict fidelity in controlled clinical trials. Effective counselor training is critical if substance abuse programs are to realize these interventions' full potential to enhance client outcomes in community practice. Although few empirical evaluations of training in EBTs have been conducted to date, the existing data warrant tentative conclusions concerning the appropriate roles and effectiveness of workshops, clinical supervision, distance learning, and blended learning. Among several outstanding research issues are questions of benchmarks for counselors' performance in training and the relationships between such performance and clients' substance abuse outcomes.

More than a decade ago, the Institute of Medicine (1998) challenged addiction professionals to implement evidence-based treatments (EBTs) in community programs. Although EBTs have been defined in various ways (Miller, Zweben, and Johnsen, 2005), in general they are treatments that have been shown to improve client outcomes in more than one randomized clinical trial (Chambless and Ollendick, 2001). In practice, counselors use their clinical expertise to apply these treatments in a manner that addresses their clients’ unique characteristics, cultures, and preferences (American Psychological Association Presidential Task Force on Evidence-Based Practices, 2006). EBTs may be pharmacological (i.e., methadone, buprenorphine, naltrexone, and disulfiram) or psychosocial (i.e., cognitive-behavioral therapy, contingency management, motivational interviewing, and 12-step facilitation) and typically are the best treatments counselors have to offer clients.

As addiction counselors’ awareness of EBTs has grown, their attitudes toward these treatment strategies, particularly psychosocial ones, have become increasingly positive (Garner, 2009), and they have increasingly begun to seek training (Miller et al., 2005). This demand has raised questions about how best to train addiction counselors in EBTs; how to evaluate their performance of the interventions; and how counselor, client, and organizational factors influence their learning and performance. This article describes current empirical knowledge on these issues and critical areas for research.

TRAINING STRATEGIES

Workshops are the most frequently used format for training counselors in EBTs. Clinical supervision is an important tool for helping counselors apply what they have learned in a workshop, or by other means, in practice with clients (Carroll and Rounsaville, 2007; Center for Substance Abuse Treatment, 2007; Miller et al., 2005, 2006). Distance learning and blended learning methods are relatively new training options that expand the possibilities for EBT dissemination. Each approach is described below, along with the evidence for its effectiveness.

Workshops

In a typical workshop, an expert or experts provide instruction via lecture and slide presentation, reinforced with reviews of treatment manuals and handouts. The workshop usually lasts hours or 1–2 days and often involves opportunities for participants to practice applying the EBT principles and skills in experiential and role-play activities. Most workshops include participants from different programs and regions; programs that provide their own workshops may tailor the training to their own specific context and issues (Baer et al., 2009).

Research indicates that workshop training improves counselors’ attitudes, knowledge, and confidence but does not adequately equip them to deliver EBTs to patients (Walters et al., 2005). For example, counselors consistently exhibit small increases in motivational interviewing (MI) skills after a workshop but revert quickly to pre-workshop levels, sometimes after only 2 months (Baer et al., 2004; Miller et al., 2004; Mitcheson, Bhavsar, and McCambridge, 2009). Although workshops are insufficient EBT training mechanisms in themselves, they are useful, and may be necessary, to inculcate basic skills and principles that counselors can further develop and hone in supervised practice experiences.

Clinical Supervision

Traditionally, substance abuse programs providing clinical supervision have relied on generic supervision principles such as those described by Powell and Brodsky (2004) or in the publication Competencies for Substance Abuse Treatment Clinical Supervisors (Center for Substance Abuse Treatment, 2007). These include establishing a supervisory alliance, recognizing supervisee stage of professional development, setting supervisory goals, and understanding organizational context and administrative functions.

Recent supervisory practice has emphasized competency-based approaches that (1) explicitly identify the knowledge, skills, and values that form the basis of competency in a particular EBT and (2) use specific learning strategies and evaluation procedures to sequentially build the counselors’ skills appropriate to their clinical settings (Falender and Shafranske, 2007). The core elements of high-quality competency-based supervision are the same activities that have been used to train counselors in the clinical trials that established treatments as evidence-based: direct observation of counselors’ sessions and the use of performance feedback and individualized coaching (Baer et al., 2007). Supervisors listen to audiotapes of counselors’ client sessions and rate the frequency with which the counselors use specific treatment strategies, their skill when implementing the strategies, and any intrusion of counseling strategies that are incompatible with the EBT (Waltz et al., 1993). The supervisors review their observations with counselors, give advice for improvement, and sometimes suggest practice scenarios or exercises (e.g., role-play during supervision) or model or demonstrate skills. A typical clinical supervision schedule calls for biweekly discussions, some with individual counselors and some with groups of counselors, extending over several months.

Several studies have shown that competency-based clinical supervision improves counselors’ ability to deliver EBTs. Miller and colleagues (2004) found that counselors who received mailed feedback, phone-based coaching, or both after attending a workshop on MI gained more proficiency than others who attended a workshop with no supervisory followup. Notably, only when counselors received the most intensive level of supervisory input—both feedback and coaching—did their clients exhibit significant improvements in motivation for change within their sessions. Sholomskas and colleagues (2005) similarly found that counselors who attended a cognitive-behavioral therapy (CBT) training workshop with 3 months of followup supervision showed greater improvements in skill and received more “adequate” ratings for their delivery of the intervention than counselors who trained themselves in the therapy using a manual. Smith and colleagues (2007) described successful use of an innovative approach to MI training in which supervisors listen to client sessions over a phone and provide performance feedback and coaching via a modified telephone headset worn by the counselor (“bug-in-the-ear”). In another study, counselors’ performance of contingency management improved when supervisors offered drawings for cash prizes to those who met criteria for adherence to the protocol (Andrzejewski et al., 2001).

Despite these promising findings, it can be difficult to engage counselors in intensive supervision. Many trainers have noted that counselors are reluctant to provide their supervisors with recordings of their client sessions or to participate in session reviews, even when they are offered at no cost (Baer et al., 2004; Miller et al., 2004; Mitcheson, Bhavsar, and McCambridge, 2009). Moreover, many substance abuse programs, pressed by time constraints, omit supervisory reviews or focus them solely on administrative issues and case review (Center for Substance Abuse Treatment, 2007). Hence, the addiction treatment community needs additional feasible EBT training options. Distance learning methods may answer this need.

Distance Learning

Distance learning methods include computer-assisted and Web-based training and simulation programs (see Weingardt, 2004, for review of this literature). They typically use a variety of media (e.g., text, video, audio instruction, animation, and interactive exercises); tailor content to meet specific training needs (e.g., by allowing learners to select from a menu of learning modules); put learners in control of the order and speed of presentation; repeat material as needed; and feature built-in opportunities to practice newly learned skills, sometimes with performance feedback. Overall, distance learning modules offer individualized training with few geographic or temporal constraints, standardized quality, and low cost.

In the field of education, distance learning has been shown to produce equivalent gains in knowledge and skill relative to traditional workshop trainings (Weingardt, 2004). However, few data are available on its effectiveness for training addiction counselors. One randomized trial investigated the efficacy of supplementing training that used the Twelve-Step Facilitation Therapy Manual (Norwinski, Baker, and Carroll, 1992) with a computer-assisted CD-ROM that was keyed to the manual (Sholomskas and Carroll, 2006). Two groups of counselors spent a minimum of 10 hours over 3 weeks reviewing the manual; at the end of this period, the group that also used the CD-ROM demonstrated significantly improved twelve-step facilitation (TSF) Therapy performance and greater gains in knowledge than the group that relied on the manual alone. The same research center reported that supplementing manual-based training with a Web-based training module produced superior CBT skills improvements throughout a 12-week followup period (Sholomskas et al., 2005). In this study, however, a workshop followed by clinical supervision yielded the best results (Figure 1).

FIGURE 1

In Training, Method Matters

Community-based substance abuse counselors who attended a seminar and received clinical supervision achieved greater gains in cognitive-behavioral therapy (CBT) adherence and skill than counselors who only studied a CBT manual. Counselors who worked with a Web-based CBT training program achieved adherence and skill ratings in between these two groups. Performance scores were assigned by independent raters based on observing the clinicians in a role-play exercise. The solid horizontal line indicates the level of performance that typically receives certification in clinical efficacy trials.

Weingardt and colleagues (2009) developed an online eight-module CBT course ( based on the Therapy Manual for Drug Addiction, Manual 1 (Carroll, 1998). They tested the course with two groups of counselors, one of which took it in a prescribed progression of two modules per week, while the other completed the modules in their own chosen order. Both groups also received weekly Web counseling supervision, the former on the material in their assigned modules, the latter largely unstructured. Both groups showed improvements in CBT knowledge and self-efficacy, suggesting that a flexible training model might be as effective as a more structured approach. Skills improvement was not measured in this study.

A new strategy for training uses computer programs to provide instruction via a virtual coach, who then provides feedback during an on-screen simulated practice session, praising good performance and offering corrective advice as needed. Hayes-Roth and colleagues (2004) developed and pilot-tested a program of this type to teach medical and nursing students a brief intervention for individuals who screen positive for having a substance abuse problem. Students who trained on the program performed the intervention significantly better in a test with standardized patients than did others who studied the intervention with a self-paced e-book.

These studies suggest that technology-based distance learning strategies can effectively teach counselors EBTs. However, none of them tested how well the strategies may affect counselors’ abilities to deliver EBTs with real clients in community program practice settings, nor did they examine effects on client treatment outcomes. Pending further studies, it may be prudent to consider distance learning methods as promising adjuncts to traditional counselor training strategies, rather than replacements for them (Weingardt, 2004).

Blended Learning

Blended learning involves combinations of training techniques and strategies to help counselors learn EBTs (Cucciare, Weingardt, and Villafranca, 2008). These may include traditional approaches (reading manuals, workshops, face-to-face supervision) and distance learning approaches (computer- and Internet-based courses and seminars, audio podcasts, online or telephone-based supervisory support). The most appropriate mix for a particular program will depend upon its counselors’ needs and interests and its trainers’ familiarity with multiple methods and skill in blending them. Blended learning typically sequences its component strategies over time, scheduled at the discretion of the trainer or according to the counselors’ preferences. This approach goes beyond one- or two-session trainings and usually involves extended contact to promote ongoing practice and skill development.

The experimental interventions that were compared in three above-cited training studies are examples of blended learning, combining a workshop with mailed or telephone feedback (Miller et al., 2004); a workshop with clinical supervision (Sholomskas et al., 2005); and Web-based CBT instruction with weekly teleconference supervision (Weingardt et al., 2009). Liddle and colleagues (2006) described a comprehensive blended approach to teach day-program counselors and other staff to perform multidimensional family therapy for adolescent substance abuse. Training was divided into two phases. An initial 6-month formal phase included group didactic sessions about adolescent development, families, and drug abuse treatment; skill-building workshops; and completion of daily diaries about key principles and skills. A 14-month implementation phase followed and involved regular supervision with feedback and coaching, co-therapy sessions with experts, and booster skill-building sessions. The researchers documented significant improvement in the use of the family therapy strategies, client satisfaction ratings, and self-reported drug abstinence from baseline to followup. The study did not have a control condition.

Blended learning is appealing because combining multiple learning methods with guided practice to teach complex psychosocial EBTs makes intuitive sense. However, research has not yet determined which combinations and sequences of training strategies over what periods of time are best for which types of EBTs, counselors, and organizational contexts. Also, the cost-effectiveness of these ambitious programs has not been determined. Counselor training and client outcome improvements might need to be large to justify the substantial investment required to develop and deliver these systems (Cucciare, Weingardt, and Villafranca, 2008).

EVALUATING PERFORMANCE

Studies have shown that counselors often overestimate their ability to deliver EBTs in the early stages of learning (Carroll, Nich, and Rounsaville, 1998; Martino et al., 2009; Miller et al., 2004). As self-evaluation is misleading, training efforts need to adequately prepare supervisors to properly judge counselors’ ability to use EBTs and provide them with teaching resources to further develop their skills. Motivational Interviewing Assessment: Supervisory Tools for Enhancing Proficiency (Martino et al., 2006) is an example of a supervision tool-kit designed for this purpose. One of the best methods for assessing counselors’ performance is for an observer to rate it with a reliable and valid scale that researchers developed to certify counselors in clinical trials. Some of these scales are specific to one EBT; others may be adapted to evaluate counselors’ performance of several EBTs. An example of the latter is the Yale Adherence and Competence Scale (YACS) (Carroll et al., 2000), which is designed to rate counselors’ delivery of CBT, contingency management, TSF, MI, and interpersonal therapy. The YACS identifies the key strategies and techniques that define each treatment and interventions inconsistent with them. Typically, supervisors use the ratings to give counselors feedback on their performance and help them improve. The YACS also features parallel self-report versions that counselors can use to make self-assessments of their own performance that they can compare with their supervisors’ assessments. Counselors who do this may be more accepting of supervisory feedback and likely to provide suggestions for their own skill development (Sobell et al., 2008).

The reliability and validity of EBT counselor performance rating scales have been well established in clinical trials. However, the scales take time to learn, and the degree to which community supervisors or consultants can apply them accurately in real world treatment settings has not been well established. In one study that yielded promising results, Martino and colleagues (2009) found that community program-based supervisors who were trained to use a YACS-based scale to evaluate counselors’ use of MI could accurately identify the presence or absence of many strategies consistent with this treatment. The supervisors gave the counselors slightly higher marks than independent observers did, but lower marks than the counselors gave themselves, on the frequency of use of MI strategies.

Another method for evaluating counselors’ delivery of EBTs is to examine evidence of clients’ responses to treatment, such as changes in symptoms or outcome data. Although competent and faithful EBT delivery does not guarantee client improvement—for example, sudden and severe psychosocial stressors may cause setbacks—this trend should be generally present in counselors’ caseloads. Monitoring clients’ progress with standardized measures on an ongoing basis may be a valuable way to evaluate how counselors’ implementation of EBTs relates to positive behavior change. Lambert and colleagues (2001) provide an example of this method: Supervisors administered the Outcome Questionnaire 45 to obtain weekly self-reports of psychotherapy patients’ symptom distress, interpersonal relationships, and social role performance. Based on the results, the supervisors classified each patient’s status in one of four color categories: green (adequate functioning with no change in treatment strategy needed), yellow (less than adequate functioning with need for treatment adjustments), red (poor functioning with need for substantial treatment adjustments), or white (normal functioning and possible treatment completion). The supervisors disclosed these assessments to the patients’ therapists and, where indicated, helped the therapists review and find ways to improve their methods and interactions with their patients. In several studies, patients whose therapists received this feedback and input had more positive treatment responses (Lambert et al., 2005), although the specific impact on therapists’ actions within sessions was not studied.

This same methodology with appropriate standardized measures might be adopted for evaluating and improving addiction counselors’ implementation of EBTs. For example, cocaine abusers’ performance of homework assignments might be an informative indicator of how well their counselors are implementing CBT, inasmuch as the extent of homework completion corresponds to coping skills and cocaine use outcomes, even after controlling for treatment retention and baseline client motivation (Carroll, Nich, and Ball, 2005). As another example, the tenor and pattern of clients’ talk about change may be a good measure of counselors’ performance of MI, an intervention in which developing and supporting change talk is an essential counselor skill (Miller and Rose, 2009).

FACTORS INFLUENCING LEARNING AND PERFORMANCE OF EBTs

To learn an EBT well, counselors likely must be receptive and have beliefs, orientations, and core counseling skills that are consistent with the treatment (Ball et al., 2002). For example, McGovern and colleagues (2004) found that counselors who endorsed a 12-step model as their primary orientation were more likely than counselors who did not have this allegiance to use TSF and less likely to use MI, CBT, or behavioral couples counseling. Similarly, Sholomskas and colleagues (2005) and Baer and colleagues (2009) found that counselors who endorsed a 12-step model showed fewer gains in CBT and MI skills, respectively, although Miller and colleagues (2004) did not find that counselor characteristics affected MI training outcomes. A relatively unexplored area has been examination of counselors’ personal characteristics and predispositions that might affect their ability to learn EBTs. For example, Miller and colleagues (2005) speculated that counselors might require a minimum level of empathic ability to learn MI and later suggested that programs wishing to implement a particular EBT might screen counselors to identify those whose personal qualities are best suited to learning and delivering it (Miller et al., 2006).

With EBTs, as with other interventions, counselors may find a particular treatment model easier to apply with some clients than with others. Gaume and colleagues (2009) found that counselors’ skill in delivering MI depended on the clients’ stated ability to change their drinking. While all counselors demonstrated effective use of MI when clients expressed high levels of ability to change, some showed markedly poorer overall MI performance with clients who said they found it difficult to not drink. Thyrian and colleagues (2007) found that counselors who were treating postpartum women for smoking demonstrated greater adherence to MI with those women who had stopped smoking than those who were daily smokers. These studies show that counselors’ EBT skills may waver when the clinical going gets tough and suggest the need to provide ongoing training support as counselors apply EBTs in real-world practice.

Organizational factors can help or hinder counselors’ efforts to learn, implement, and sustain an EBT over time. Fuller and colleagues (2007) found that counselors were more likely to support EBTs when they felt that their programs had a greater need for improvement and when their programs had more Internet access, better opportunities for professional growth, a clearer sense of organizational mission, and higher organizational stress. Baer and colleagues (2009) found that training improved counselors’ MI skills most when they felt their programs were open to change and not supportive of counselors functioning completely independently. Other studies have demonstrated that the influence of peer opinion leaders who advocate for EBTs is essential for successful implementation, and loss of these key individuals may derail training efforts (Squires, Gumbley, and Storti, 2008). Careful consideration of these and other critical organizational issues—e.g., staffing stability, policies and procedures, financial resources, administrative support—should be part of the planning process when preparing EBT training for counselors (Center for Substance Abuse Treatment, 2009; Simpson and Flynn, 2007).

QUESTIONS FOR RESEARCHERS

As the foregoing suggests, the empirical basis for training counselors in EBTs is developing but still far from complete. We have partial answers to some of the questions that follow and no answers to others.

Which Training Strategies Are Optimal to Teach Counselors EBTs?

Given the limited evidence about training strategy effectiveness, it is premature to specify guidelines for how to best train counselors in EBTs. Nonetheless, the research findings to date warrant provisional conclusions that:

distance learning methods appear to develop counselors’ knowledge;

workshops may serve as a platform for establishing basic skills;

counselors likely require clinical supervision that includes observation, feedback, and coaching to become proficient in using EBTs with real patients;

blending some or all of these strategies is probably the best way to approach counselor training.

On this basis, programs might consider sequential training that begins with the least resource-intensive methods. Thus, counselors who are new to an EBT might first complete a Web- or computer-based training to understand the basic concepts, then attend a workshop to obtain initial hands-on experience using multiple practice exercises, then implement the treatment under clinical supervision.

As noted earlier, some counselors may “take to” an EBT quickly, while others may struggle to learn it. Some researchers have accordingly proposed adaptable training approaches that can provide each counselor with the intensity of training that he or she needs (Collins, Murphy, and Bierman, 2004). Such approaches will require performance criteria or benchmarks for identifying counselors who need additional training.

What Should the Criteria Be for Evaluating Counselor Learning?

No empirically derived training criteria or benchmarks linked to client outcomes exist in the literature for any EBT. With the exception of Liddle and colleagues’ (2006) descriptive training trial, none of the studies noted above examined whether training counselors to use EBTs actually improved clients’ substance use outcomes. However, Sholomskas and colleagues (2005) and Sholomskas and Carroll (2006), in their above-mentioned studies of CBT and TSF training, reported the percentage of counselors who reached the same YACS-based standards of performance used to certify counselors in studies in which those interventions were efficacious. Miller and colleagues (2004) used the Motivational Interviewing Skills Code (MISC)—a validated instrument for measuring fidelity to MI, although not one linked to outcomes—to establish standards for counselor proficiency in MI. More research with outcomes-linked performance measures is needed to establish empirically derived training goals.

How Does Training Affect Counselors’ Interactions With Clients?

To date, most studies that have evaluated counselors’ ability to implement EBTs have done so by assessing their performance in demonstration sessions with actors portraying clients. It is not clear that such assessments accurately predict how counselors will use EBTs with real clients. Real clients will likely vary more in clinical presentation and responsiveness to interventions than client actors (Miller et al., 2004). Resolution of this issue will require randomized controlled trials that evaluate how well counselors use EBTs with real clients and that collect practice samples with more than one client at each assessment point to gain a more valid measure of the counselors’ skills in using EBTs.

How Well Do Training Strategies Sustain Counselors’ Skills?

Most training studies to date have tested the effectiveness of interventions that were relatively brief (1–4 months) with posttraining followup periods of only a few months. Evidence that training strategies produce initial skill increases does not mean that these effects are durable, nor is it likely that the counselors’ skills will improve further without subsequent training and guided practice. While extended blended training interventions have promise for teaching counselors complex psychosocial EBTs, many questions remain unanswered, such as:

What mixture of blended strategies should be used?

How intensive should each strategy be, and how long should the intervals be between strategies?

How long does it take for performance standards to be met?

In addition, the order of training targets might also be important. Miller and Moyers (2006) proposed eight ordered stages for learning MI; for example, becoming familiar with the underlying philosophy of the intervention precedes recognizing and reinforcing client statements that support change. While they acknowledge that their exact ordering may not be desirable in all cases, their stages do suggest a possible progression for extended blended training programs. Further development and testing may advance EBT training.

What Qualifications Should Trainers Have?

Quality assurance standards need to be developed for trainers to ensure high-quality counselor training. These standards should include competence in performing the EBT and other skills necessary to be an effective trainer—for example, facilitating discussions, organizing materials, and adapting content and methods to meet trainee needs. Direct observation is necessary for establishing counselors’ competence in delivering EBTs, and the same holds true for trainers. Martino and colleagues (2007) found that one-third of the applicants for a training-of-trainers workshop in the use of an MI supervision product were unable to demonstrate modest standards of MI proficiency based on an independent review of recorded client sessions. The specification of competencies for EBT trainers and methods to train them to these standards requires future development.

How Well Does Formal Coursework Prepare Students to Implement EBTs?

Most professionals first learn how to provide substance abuse treatment through formal coursework in graduate or certification programs. Coursework usually is coupled with clinical experiences to help students learn how to apply the material. Unfortunately, this training often does not emphasize EBTs, and most new substance abuse professionals enter practice unprepared to implement these interventions (Weissman et al., 2006). Work is needed to develop and evaluate curricula for substance abuse EBTs, possibly incorporating some of the counselor training strategies described above. Given that graduate and certification courses typically are designed to conceptually build upon one another, the progression of coursework might offer a unique opportunity to study the developmental process or stages by which counselors learn EBT skills.

How Cost-Effective Are the Different Training Strategies?

Many have noted that the cost of counselor training must be weighed against the benefits expected from using EBTs (Cucciare, Weingardt, and Villafranca, 2008; Miller et al., 2005). This is particularly true for technology-based approaches and comprehensive blended learning approaches that require a substantial investment of resources. Studies to date have not estimated the costs involved with training or how much benefit consumers of training programs are likely to get for their money. Cost-effectiveness analyses need to be part of future counselor training trials.

What Studies Are Under Way That Might Advance Our Understanding of How to Train Counselors in EBTs?

Two NIDA-funded trials are currently testing counselor training strategies. Moyers is conducting a randomized clinical trial to determine if workshop and supervision training in MI can be streamlined by emphasizing the elements that presumably make the intervention work. The study will examine whether training counselors to recognize, reinforce, and elicit change language increases clients’ use of such language in counseling sessions. Martino is currently conducting a randomized clinical trial to test the effectiveness of supervising counselors in MI using the Motivational Interviewing Assessment: Supervisory Tools for Enhancing Proficiency. This study aims to determine the impact of clinical supervision on client outcomes and the extent to which the counselors’ adherence and competence in using MI mediates these outcomes.

CONCLUSIONS

While evidence-based practices are now established and widespread in substance abuse treatment, the knowledge base regarding counselor training methods has just begun to form. Although the evidence is far from complete, it indicates that ongoing training with supervisory support and rating-based feedback and coaching—spaced over time and individualized to the counselors’ training needs—is effective. New distance learning approaches have potential to extend training to more counselors and may be particularly useful when blended with traditional approaches to help learners master the complexities of psychosocial substance abuse treatments. Research has shown that counselors vary in their capacity to learn EBTs, depending on their treatment orientations, the types of clients they treat, and the nature of the organizations within which they work. More effort is needed to better understand these relationships. Finally, while training increases EBT skills and fidelity, the impact on client outcomes and cost-effectiveness are unknown.