Get Adobe Flash player


CRA, ACRA and CRAFT has been subject of multiple studies in the past four decades. Some of those studies have been conducted in the Netherlands or are the result of collaborations with international research groups. On this page a selection of key research findings are presented highlighting CRA.

Why using the Community Reinforcement Approach (CRA)[1]?

Counselor training is a critical question for using any Empirically Supported Treatment (EST). We elected to examine training issues using CRA as a model because CRA meets the definition of an EST and has advantages relative to several other ESTs:

  1. CRA has been found to be among the most strongly supported treatment methods in nearly every review and meta-analysis of the alcohol treatment outcome literature during the past decade (e.g., Chambless & Ollendick, 2001; Finney & Monahan, 1996; Holder et al., 1991; Miller & Heather, 1998). Support of CRA interventions from randomized clinical trials is particularly strong (Kirby et al., 1999; Meyers & Miller, 2001; Miller et al., 1999; Roozen et al., 2004 ; Sisson & Azrin, 1986).
  2. CRA provides a comprehensive, integrated approach for treating substance use disorders, yet it is consistent with practice in a managed care environment. It can be delivered in 30 to 60 minute individual sessions: a format that is widely available in treatment programs (even though group formats are used heavily) and typically is reimbursed by the payers of substance abuse treatment. (Note: CRA has also been adapted to group formats; however, we will initially examine methods for training CRA delivered in individual sessions because this format has the best evidence of efficacy.)
  3. CRA is flexible, and is delivered in a manner similar to standard clinical practice. Whereas therapist manuals often prescribe a fixed length of treatment and specific content for each session, CRA is a menu-driven approach that is individualized to patient needs, allowing flexibility in the spacing, number, and extensity of sessions. This allows clinicians to retain a sense of doing it their own way.
  4. CRA is rarely implemented. While it has been in the literature for almost 40 years, it is still relatively unknown and unused among substance abuse treatment professionals.
  5. We find that once they have been introduced to CRA, most clinicians are interested in learning it. They typically view the approach as potentially advantageous, compatible, learnable, and easy to try out (cf. Henggeler et al.,2002).
  6. CRA incorporates Cognitive Behavioral Treatment (CBT) strategies along with strategies consistent with a more basic operant behavioral approach. Its basic therapeutic style is consistent with Motivational Interviewing (MI), but like CBT, it involves the application of specific behavioral principles and requires implementation of a specific technology.
  7. CRA procedures are well-specified, with published therapist manuals that have already been tested in clinical trials with alcohol and drug problems (Meyers et al., 1996; Meyers & Smith, 1997).
  8. Reliable measures and methods have already been developed to assess the amount and quality of CRA implementation and to provide counselors with feedback and coaching.
  9. Train-to-criterion methods that allow counselors to be certified in CRA have already been developed and implemented in various countries.
  10. CRA training and certification is already available in the US and other countries around the world and practical training methods for both counselors and supervisors have been developed so that community-based treatment programs can sustain the implementation of high-quality CRA services. As such, CRA provides an excellent model for evaluating the efficacy of training methods.

What Is CRA and What is the Empirical Support For It?

CRA is a comprehensive behavioral program for treating substance abuse problems. It is fundamentally based in operant learning theory, aiming to achieve abstinence by diminishing positive reinforcement for substance abuse and increasing positive reinforcement for sobriety. It integrates several treatment components, including motivation building and increasing of positive expectations; functional analysis of the client’s drinking; pro-social, or relapse behavior; training variety of skills including communication and drug refusal skills; sobriety sampling; problem solving; couples therapy and relapse prevention; and using shaping and positive reinforcement to build skills during in-session role-play.

Numerous controlled studies of CRA have demonstrated relatively large effects with clients suffering from alcohol problems (Azrin, 1976; Azrin et al., 1982; Hunt & Azrin, 1973; Mallams et al., 1982; Miller et al., 2001; Smith et al., 1998), opiate dependence (Abbott et al., 1998; Bickel et al., 1997) and cocaine dependence (Higgins et al., 1993, 1995, 1997, 2000a, 2000b; Secades-Villa et al., 2008).

Moreover, during the past two decades a number of meta-analytic reviews of treatments for alcoholism have consistently ranked CRA among the top programs in terms of effect sizes (Holder et al., 1991; Miller et al., 1995). In fact, in one of these (Finney & Monahan, 1996) CRA was ranked first. In a more recent systematic review of CRA, Roozen et al. (2004) identified and reviewed 11 CRA studies of mainly high methodological quality and concluded that there was moderate to strong evidence that CRA was more effective than usual care in the treatment of alcohol and cocaine disorders, and had promising support for the treatment of opiate disorders. Evidence was strongest for CRA in the treatment of alcohol disorders, and CRA in combination with abstinence-based vouchers for the treatment of cocaine dependence.

Why Hasn’t CRA Been Used More Broadly?

There are several possible reasons why CRA has not been broadly applied. First, the U.S. has embraced the disease model, which looks within the individual for the cause of alcoholism and drug addiction (Miller, 1986; Miller & Hester, 1995) while CRA is based in operant theory, which emphasizes interactions with the environment. Second, counselors often have well-established patterns of practice that may require unlearning before new innovations can be learned. Third, when correctly implemented CRA is an extremely active approach. The therapist must give examples, model behaviors, role-play scenes and teach new skills. Some therapists are accustomed to a more passive role and may have trouble shifting to an active therapy approach. Although these reasons probably account for some of CRA’s limited distribution, we believe that perhaps the most important reason is that until very recently there have been no systematic training programs in CRA and few experienced senior clinicians to promote and teach these approaches. For decades CRA had been primarily accessible through scientific journals. The CRA manual was published in 1995, but provision of manuals is known to be a very ineffective method to promote implementation and dissemination. Effective methods are needed that provide training and support for the development and implementation of CRA skills.

[1] This text is written by Kimberly Kirby, PhD and adapted for this website.