Alcohol-Focused Behavioral Couple Therapy

Abstract

Alcohol Behavioral Couple Therapy (ABCT) has emerged over the last 30 years as a highly efficacious treatment for those with alcohol use disorders. This review highlights the historical and conceptual underpinnings of ABCT, as well as the specific treatment elements and structure. Proposed active ingredients, moderators, and mediators of treatment outcome are discussed. Efficacy is evaluated for reductions in identified patient drinking, improved relationship functioning, and reductions in intimate partner violence. Adaptations of ABCT for substances other than alcohol are described. Other adaptations, including brief interventions, interventions addressing PTSD and TBI along with alcohol use, and interventions deliverable via technology platforms are described. Additional cost-benefit and cost-effectiveness findings supporting the economic value of ABCT are noted. Future directions for research in this area include possible adaptations for female identified patients, non-traditional couples, LGBT partners and dyads involving non-intimate partner relationships. The development of more flexible models and enhanced dissemination strategies may improve clinical uptake and utility as well as increasing the feasibility of this treatment for integrated healthcare settings.

Description of Alcohol Behavioral Couple Therapy (ABCT)

Historical Roots

Concerns about the impact of alcohol on families and the engagement of families in alcohol treatment date back to the temperance movement in the 1800s (inspired in large part by women's concerns about the impact of male drinking in taverns on the family) and efforts in the late 1800s to engage families in treatment in early residential treatment programs for “dipsomania and inebriety” (McCrady, Owens, & Brovko, 2013). More contemporary family-focused treatment approaches began with efforts by caseworkers to assist women married to men with drinking problems (e.g., Baldwin, 1947) and the development of parallel therapy groups for husbands with alcohol use disorders (AUDs) and their wives (e.g., Gliedman, Rosenthal, Frank, & Nash, 1956; Pattison et al., 1965). Table 1 summarizes major characteristics of couple therapy studies for AUDs (see also Table S1 for a complete listing of early treatment studies). Many of these early approaches drew on psychodynamic principles, positing that marriage to a man with an AUD represented a neurotic resolution of psychological conflicts by these wives, and that therapy, therefore, should focus on the woman's own psychological issues.

Table 1.

Major Characteristics of Couple Therapy Studies for Alcohol Use Disorders

Therapy Type	Number of Studies	Number of Participants	Years Conducted	Relationship Outcome Evaluated & Improved?
				Yes	No	No Diff.
Non-Behavioral Couple Treatment	12	565	1956 - 1982	4	7	1
Group Treatment for Wives Only	4	159	1958 - 1969		4
Behavioral Couples Therapy	24	1186	1985 - present	15	3	6

The application of family systems and behavioral models to the treatment of AUDs began in the late 1960s and early 1970s, when clinicians began to report the use of conjoint therapy for AUDs (e.g., Burton & Kaplan, 1968a). The earliest studies described family-systems based models with some cognitive-behavioral elements, and many reported comparisons of conjoint therapy to other approaches using non-randomized groups (e.g., comparing outcomes for men in treatment whose wives did or did not participate in sessions). Follow-ups varied widely in duration, from 6 to 39 months, and typically reported substantially more positive drinking outcomes for men whose wives participated in the treatment than those who did not, as well as improvements in relationship functioning (Burton & Kaplan, 1968a, 1968b; Gallant, Rich, Bey, & Terranova, 1970; Smith, 1967, 1969). By the mid-1970s, descriptions of behavioral approaches to conjoint therapy for AUDs began to appear in the literature, and controlled outcome studies of cognitive-behavioral approaches began in the late 1970s (e.g., McCrady et al., 1986; O'Farrell, Cutter, & Floyd, 1985).

Conceptual Model

Alcohol-focused Behavioral Couple Therapy (ABCT) is a cognitive-behavioral treatment model based on the assumption that multiple factors maintain the identified patient's (IP's) drinking, including individual, dyadic, familial, and other social/environmental variables. The ABCT model assumes a reciprocal relation between drinking and relationship functioning, and that interventions focused on both will be most effective. The model assumes that (a) external antecedents to drinking have a lawful relation to drinking, developed through repeated pairings with positive or negative reinforcers; (b) internal physiological, cognitive, and affective states mediate the association between external antecedents and drinking behavior; (c) expectancies about the reinforcing value of alcohol play an important role in determining subsequent drinking behavior; (d) drinking is maintained by its more immediate, positive consequences, which may be physiological, psychological, or interpersonal; and (e) negative consequences of drinking tend to be delayed and therefore have less impact on drinking behavior (see McCrady & Epstein, 2015).

Interventions in ABCT focus on familial antecedents and consequences of drinking. Familial antecedents may include typical family celebrations or daily rituals as well as familial attempts to influence the IP's drinking. Families in which alcohol problems are present often have evolved poor patterns of communication and problem solving and have developed a variety of relationship, sexual, financial, and child-rearing problems over time. All of these can serve as antecedents to further drinking.

Families inadvertently play a large role in both beneficial and aversive consequences of drinking. Some beneficial consequences include sharing of positive activities that include alcohol, caretaking when the IP has been drinking, or being particularly gentle and nonconfrontational during drinking episodes. Although these behaviors can be understood as normal reactions when a family member is sick or in a bad mood, such behavior in families with alcohol problems may serve to reinforce drinking. Families also provide a number of aversive consequences for drinking, such as withdrawal and avoidance of the drinking member, negative verbal comments about the drinking (either during or after a drinking episode), and, in some families, physical violence directed at the drinking member. These aversive consequences may lead the drinker to avoid family interactions or attempt to hide the drinking, or may serve as cues to further drinking.

McCrady and Epstein's approach to ABCT combines three major components into an integrated treatment program (McCrady & Epstein, 2015) to affect the drinking and negative patterns of couple interactions. These include (a) cognitive behavior therapy (CBT) to target the IP's drinking; (b) CBT to enhance significant other (SO) skills to support change; and (c) behavioral couple therapy (BCT) to enhance relationship functioning. Other ABCT approaches (e.g., O'Farrell & Fals-Stewart1, 2006) typically have focused primarily on the SO skills training and BCT aspects of ABCT, with the primary alcohol treatment provided in a separate program. To distinguish between ABCT and the O’Farrell and colleagues’ treatment approach we have labeled their treatment as BCT-A for AUD populations, and BCT-D for other drug dependent populations throughout this paper.²

Treatment Elements

Table S2 provides an outline for ABCT. Specific treatment elements include:

CBT for drinking

Similar to other CBT approaches to AUDs (e.g., Epstein & McCrady, 2009), ABCT includes a number of strategies designed to help the IP decrease and/or stop drinking, including: (a) self-monitoring of drinking through daily logs; (b) functional analysis of drinking, including examination of antecedents to drinking, internal reactions to external antecedents (physiological, cognitive, and affective), the actual behavioral response to the antecedent (e.g., drinking, other response), and positive and negative consequences of the drinking; (c) development of a plan to reduce or stop drinking; (d) self-management planning; (e) development of strategies to manage negative cognitions and negative affect; (f) development of alternative behavioral coping strategies; and (g) relapse prevention.

CBT for partner coping

SO-focused interventions are similar to those developed in Unilateral Family Therapy (Thomas & Ager, 1993) and the Community Reinforcement and Family Training approach (Meyers, Smith, & Lash, 2005), and include: (a) self-monitoring through daily logs; (b) functional analysis of SO behaviors that might serve as antecedents or beneficial consequences of drinking; (c) self-management plans for behavior change; (d) skills training for coping with drinking-related situations and feelings; (e) skills training to provide positive support for IP behavior change; and (f) partner-focused relapse prevention.

BCT for relationship enhancement

Couple-focused interventions are similar to those provided in BCT for relationship distress (e.g., Epstein & Baucom, 2003), but also include specific alcohol-focused couple interventions. Couple interventions include: (a) increasing shared positive activities; (b) increasing observation and feedback about positive partner behaviors; (c) developing communication skills around alcohol-focused topics such as whether to keep alcohol in the house, or how to jointly manage situations in which the IP is offered alcoholic beverages; (d) increasing communication and problem-solving skills training; and (e) developing couple-focused relapse prevention strategies. O'Farrell and Fals-Stewart's approach (2006) also includes “sobriety contracts” that may include daily use of medications such as Antabuse.

Structure of Treatment

ABCT is a structured treatment, typically guided by a therapist manual and workbook for the couple. Assessment at the beginning includes a 2-hour conjoint semi-structured clinical interview and self-report questionnaires to determine whether the couple is a good candidate for ABCT and a short individual meeting with each partner to assess for intimate partner violence (IPV). Couples who are interested, willing, and able to attend treatment sessions together and who do not show significant levels of IPV are good candidates for ABCT. Daily self-monitoring by both partners is introduced in the first session and continues throughout the treatment.

ABCT is designed to include both partners in all treatment sessions, although recent research suggests that providing a combination of ABCT and individual CBT tended to yield better treatment attendance and comparable treatment outcomes (McCrady, Epstein, Hallgren, Cook, & Jensen, in press). Sessions typically are 90 minutes in length, and the model has been tested with varying lengths of treatment, ranging from 12-20 sessions. When present, both partners are actively engaged in all aspects of the treatment, providing information and feedback even during the more individually-focused interventions.

Efficacy Research

Table S1 provides a comprehensive overview of the efficacy research that laid the foundation for ABCT interventions for AUDs as well as studies specifically of the efficacy of ABCT. From 1956-1982, non-BCT group interventions for couples were tested with samples comprised primarily of male IPs and their female partners, with IP sample sizes ranging from nine to 183; follow-up periods ranged from none (post-treatment) to four years. From 1958-1969, several investigators conducted studies of group therapies designed to support wives of men with AUDs, again with a range of sample sizes (six to 80) and follow-up periods (post-treatment to six months). Studies of BCT for AUDs began in 1985 and continue today, again with mostly male IPs, and have produced promising results on both drinking and relationship outcomes, with sample sizes ranging from nine to 303 IPs and follow-up periods ranging from post-treatment to 30 months.

Overview of Outcomes of ABCT Research

Typically, ABCT research has focused on two desired outcomes: reduced IP drinking and improved relationship functioning between the partners (e.g., McCrady & Epstein, 2009; O'Farrell & Fals-Stewart, 2006). ABCT has been shown to positively impact both of those outcomes. Research has shown that ABCT benefits both male and female drinkers in intimate relationships in reducing drinking, reducing drinking severity, and improving the overall quality of the relationship (O'Farrell et al, 1997; McCrady et al., 1999). Additionally, certain IP or SO characteristics may be predictive of positive outcomes for couples in ABCT. Having an SO who is particularly supportive of the IP, and having an SO without a personal history of problematic alcohol use both are related to better ABCT outcomes (O'Farrell, Kleinke, Thompson & Cutter, 1986).

Foundational Research

As noted, initially most efforts to impact couples affected by AUDs focused on separate therapy groups for male IPs and their wives, or groups for wives whose husbands were in ongoing alcohol treatment (Gliedman et al., 1956; Igersheimer, 1959; MacDonald et al., 1958). This work focused on males who had problematic drinking, and were largely intended to help women cope with the effects that partner drinking had on their families. Although results were somewhat mixed, overall these early studies had promising findings. For example, Gliedman et al. (1959) showed reduced drinking and improved sexual adjustment over the 16 week treatment period in male IPs. Igersheimer (1959) showed improved emotional expression over the course of five months in treatment for wives of men with AUDs, illustrating that involving partners in treatment could be beneficial for couples in distress. This foundational work served a number of functions. These early studies demonstrated the feasibility of these interventions, examined their usefulness for couples struggling with AUDs, and formed the basis for developing and testing interventions that could help couples in distress.

Over the next few decades, interest in specific behavioral interventions began to expand. The focus began to shift to interventions specifically designed to effect change in couples rather than just the SO, and several early RCTs of couple therapy approaches reported positive results (e.g., Burton & Kaplan, 1968a; Cadogan, 1973; Corder, Corder, & Laidlaw, 1972; Hedberg & Campbell, 1972; McCrady, Paolino, Longabaugh, & Ross, 1979). Over time, couple therapy approaches drew more on cognitive-behavioral treatment approaches, both to alcohol problems and to relationship distress. Overall, ABCT and BCT-A have a strong research base supporting their efficacy (O'Farrell & Schein, 2011; McCrady, 2012; Epstein & McCrady, 1998) and have been shown to lead to greater improvements in abstinence in the IP and relationship functioning of the couple compared to individually-focused treatments; two research groups have provided the most sustained contributions to the ABCT/BCT-A literature and their work is reviewed in some detail.

McCrady has reported the results of several clinical trials of ABCT. In a small initial randomized clinical trial (RCT) of males and females with AUDs, McCrady and her colleagues (McCrady et al., 1986; McCrady, Noel, Stout, Abrams, & Nelson, 1991) tested the active ingredients of ABCT by comparing CBT with the spouse present (minimal spouse involvement) to CBT with treatment focused on spouse coping (alcohol-focused spouse involvement, AFSI) and with ABCT. Outcomes 18 months post-treatment suggested that couples receiving ABCT showed greater improvements in relationship satisfaction and maintained positive changes in drinking better than couples in the comparison treatments. In a second RCT of males with AUDs and their partners, McCrady's group tested ABCT against ABCT enhanced either with relapse prevention (RP) interventions or with engagement with Alcoholics Anonymous and Alanon (McCrady et al., 1996, 1999; 2004). Drinking and relationship outcomes were comparable across the three treatments, but relapses were shorter in duration in the combined ABCT/RP treatment condition. McCrady and her colleagues also have tested ABCT in two studies with women with AUDs and their male partners (McCrady et al., 2009; McCrady et al., in press). The first of these RCTs (McCrady et al., 2009) compared ABCT to individual CBT, and found a higher percentage of abstinent days and a lower percentage of heavy drinking days in ABCT than individual CBT in the 12 months after treatment. The second study with women with AUDs (McCrady et al., in press) built on findings suggesting that women with AUDs often prefer individual treatment (McCrady, Epstein, Cook, Jensen, & Ladd, 2011) and used an RCT design to compare ABCT to a blend of ABCT and individual CBT sessions. Although the groups did not differ significantly on attendance or drinking outcomes, small to moderate effect sizes favored the blended treatment over stand-alone ABCT for this population.

In his studies of BCT-A, O'Farrell and colleagues (1985) found that in couples with a male IP, those assigned to the BCT-A condition rather than a no conjoint treatment control group or an interactional couple therapy group had fewer drinking days than either of the comparison groups. Additionally, couples receiving either interactional couple therapy or BCT-A also showed improved communication and marital adjustment, whereas the couples receiving no conjoint treatment did not. In a second study O'Farrell and his colleagues (O'Farrell, Choquette, Cutter, Brown, & McCourt, 1993; O'Farrell, Choquette, & Cutter, 1998) evaluated the effects of combining BCT-A with relapse prevention for couples with a male IP. After receiving 20 sessions of BCT-A, couples receiving an additional 15 RP sessions over the next year showed greater improvements in both alcohol use and relationship adjustment up to 18 months post-baseline. O'Farrell's findings, combined with McCrady's findings on ABCT plus RP, suggest that teaching couples specific tools to deal with potential relapse is helpful to couples with a male IP. O'Farrell and his colleagues also have tested BCT-A in samples of women with AUDs, and have found, compared to women receiving individual treatment that women receiving BCTA have been shown to have significantly reduced heavy drinking, more days of abstinence, and greater relationship satisfaction (Schumm et al., 2014).

RCTs from other research groups (see Table S1) also have reported better drinking outcomes for ABCT than comparison conditions (e.g., Bowers, 1990; Schumm et al., 2014, 2015; Walitzer & Derman, 2004). However, Vedel, Emmelkamp & Schippers (2008) found no differences in outcomes between ABCT and individual treatment, and Zweben (1988) found no differences in outcomes between a one-session advice and an eight-session conjoint treatment protocol.

Gender and ABCT Research

The majority of the research on AUDs in couples has focused on male IPs and their female partners, although research with female samples also has found that involving partners in treatment typically has led to reduced drinking and improved relationship functioning. There are several possible explanations for the overrepresentation of males in ABCT. First, the prevalence of AUDs is lower in women than men. Additionally, however, social mores continue to regard AUDs as an issue that affects only men and the greater stigma experienced by women with AUDs may affect their help-seeking. Also, male partners of women with AUDs may be more reluctant to engage in treatment, making it more difficult for women to access ABCT.

More recently, women have emerged as a population of interest in this area. Though more men suffer from AUDs than women, the consequences of problematic drinking behaviors disproportionately affect women. Women are more likely to die as a result of their drinking (Smith & Weisner, 2000), and are more likely to have severe medical problems as a result of their drinking. In addition, the reasons women drink may also differ from those of men. For example, women are more likely than men to drink as a result of discord and stress in their intimate relationships (McCrady, Epstein, Cook, Jensen & Hildebrant, 2009), women are also more vulnerable to relapse by drinking with their partners (Connors, Maisto, & Zywiak, 1998), and women are more likely than men to drink to cope with negative emotions (Annis & Graham, 1995). By addressing these unique challenges, adaptations of ABCT for women might improve treatment entry and retention in ABCT, as well as improve treatment outcomes.

Effectiveness Research

Although there is a substantial body of ABCT efficacy research, there are no true effectiveness studies of ABCT. A number of studies (e.g., Vedel et al., 2008) have been conducted in real-world community treatment program studies, but because these studies have had strict study inclusion and exclusion criteria, relatively small sample sizes, and short follow-ups, they cannot be considered to be true effectiveness studies (Gartlehner, Hansen, Nissman, Lohr, & Carey, 2006).

Recently, the Veterans Administration Healthcare System initiated a program to disseminate BCT for alcohol and other substance use disorders in the VA system. Unfortunately, the program was discontinued because of changes in budget priorities within the VA system, resulting in very limited effectiveness data on the program. However, O'Farrell and colleagues (2015) reported on the initial phase of the VA BCT-A dissemination project, which included a three-day training workshop followed by a six month consultation phase to guide therapists in learning how to implement BCT. Beginning in 2012, 92 therapists were enrolled in the training program; 68 completed program requirements. Therapist ratings of the initial workshop phase of the training were very positive, indicating that the training was successful in providing a better understanding of BCT theory and strategies, and teaching them couple therapy skills. Subsequently, a non-randomized outcome study of the implementation of BCT-A in the VA setting was conducted with 40 patients with AUDs (80% of sample) or other substance use disorders. Days of drinking and drinking-related consequences both decreased significantly from baseline to the end of treatment; SO relationship satisfaction increased significantly as well. No post-treatment follow-up data were reported, however. This preliminary implementation and effectiveness research project suggested the feasibility of training front-line clinicians in the use of BCT-A and potentially positive outcomes; it is unfortunate that the project was discontinued.

Process Research: Moderators, Active Ingredients, and Mediators

Examining moderators, active ingredients, and mediators in randomized clinical trials is valuable in elucidating for whom and under what circumstances treatments work and do not work, as well as why treatments work or do not work. Moderators are individual difference variables that may impact how a treatment works for different individuals or couples. Examining moderators is particularly important because knowledge of individual differences may allow clinicians to determine which treatment will be most effective for which clients, and for which clients other treatments should be sought. Active ingredients are the specific elements of a treatment that account for positive results. Active ingredients may be specific to one type of treatment or may be common to more than one treatment. Identifying active versus inactive or ineffective treatment elements may allow treatments to be streamlined. Mediators are client processes impacted by the active ingredients, which lead to desired behavior change. Examination of mediators allows for the identification of client processes that that should be enhanced in treatment. The result of studying moderators, active ingredients, and mediators, potentially, is a more potent and efficient treatment. Process research for ABCT is still in its nascence, but there are a few moderators, active ingredients, and mediators that have been examined to date. More work is needed in this area.

For Whom ABCT Works or Does Not Work - Moderators

Psychopathology

ABCT may provide additional benefits to individuals with additional psychopathology (in DSM-IV terminology, both Axis I and Axis II disorders). For example, in a study of women with AUDs and their male partners, women with a co-morbid Axis I disorder receiving ABCT had a higher percentage of abstinent days at 18 months post-treatment than those receiving individual CBT. Similarly, women with co-morbid Axis II psychopathology who received ABCT reported a higher percentage of abstinent days at the end of treatment and a lower percentage of days of heavy drinking at 18-month follow-up than those who received individual CBT (McCrady et al., 2009). It was not clear, however, if there was an effect of ABCT on psychopathology itself or if this association was mediated through improvements in relationship stability and satisfaction.

Drinking severity

To date, no single study of ABCT has included participants with a wide range of drinking severity, thus precluding direct analyses of drinking severity as a potential moderator. However, although there is a paucity of direct studies of drinking severity as a moderator, findings from one study suggest indirectly that ABCT may be more efficacious for drinkers with more severe alcohol dependence. Walitzer and Dermen (2004) found that ABCT and alcohol-focused spouse involvement (AFSI) treatment both were more efficacious than CBT in drinking outcomes both at post-treatment and at follow-up in couples with a male problem drinker but, in contrast to McCrady et al. (1991), outcomes did not differ between ABCT and AFSI. The authors concluded that the addition of relationship focused interventions in ABCT did not provide any additional benefit. It may be, however, that because the sample in Walitzer and Dermen's study only included problem drinkers and not alcohol dependent drinkers, the havoc that more severe alcohol dependence often wreaks on interpersonal relationships had not occurred in the relationships of this study sample.

Pre-treatment relationship satisfaction

McCrady et al. (2009b) found that women with higher relationship satisfaction at a baseline measurement had a lower percentage of heavy drinking days in ABCT treatment compared to individual CBT at 12-month post-treatment follow-up. The better baseline relationship functioning may allow ABCT to capitalize on the existing goodwill in the relationship, which allows both partners to focus on the aspects of the treatment related to reducing alcohol use. With more distressed couples, data suggest that more extended treatment may be more effective than standard ABCT or BCT-A. For example, O'Farrell et al. (1998), found that men with poorer relationship functioning had better drinking outcomes if they received BCT-A plus RP than BCT-A alone.

Proposed Active Ingredients

Four specific active therapist ingredients/interventions have been proposed for ABCT: (a) motivational enhancement; (b) drinker skills training; (c) partner skills training; (d) relationship enhancement. Of these, only partner skills training and relationship enhancement interventions have been studied. In addition, two active ingredients (adherence to the treatment manual, empathy) common to many treatments (not just ABCT) have been studied.

Dose-response relationship (amount of treatment)

ABCT appears to be as effective as control treatments in producing both positive drinking and relationship satisfaction outcomes, regardless of the number of treatment sessions (Powers, Vedel & Emmelkamp, 2008). In their randomized clinical trial of ABCT compared to individual CBT, McCrady et al. (2009) reported that participants in the CBT group attended significantly more treatment sessions than participants in ABCT, but women in ABCT evidenced better drinking outcomes. Findings such as these suggest, at least for women, that the dose-response relationship often seen in AUD treatment may not hold for ABCT, perhaps because addressing relationship functioning in addition to problematic alcohol use attenuates the need for more extensive treatment.

Therapist common factors

McCrady (2014) reported on a study in which ABCT therapy sessions were coded to examine time-ordered relations between therapist behaviors at the start of treatment, and drinking outcomes at three time points (mid-treatment, end of treatment, 6 months post-treatment). Therapist adherence to the ABCT treatment manual and a composite measure of common factors accounted for a significant but small percentage of IP drinking across the first half of treatment.

Partner skills training

O'Farrell et al. (1998) taught SOs how to reinforce IP use of Antabuse through the implementation of a daily sobriety contract, comparing the use of Antabuse contracts for couples receiving BCT-A or BCT-A + RP. Those in the latter group used Antabuse contracts more in the first twelve months after treatment.

Relationship enhancement

Studies have demonstrated that there may be a temporal relation between alcohol use and relationship satisfaction (Powers et al., 2008) in which relationship satisfaction gains occur before improvements in alcohol use or consequences of alcohol use. Such findings suggest that improved relationship functioning may facilitate improvements in drinking outcomes (Powers et al, 2008). However, studies of post-treatment drinking and relationship functioning have found a concurrent association but not a temporally ordered relation. Additionally, women reported attending more treatment sessions and were more engaged in treatment if they were in more satisfying relationships (Graff et al., 2009). Improvements in communication and problem solving both have been reported. Walitzer, Derman, Shyhalla and Kubiak (2013) observed improvements in both drinking and reductions in negative and harmful communication patterns. The improvements in communication appeared to also positively affect problem-solving for couples. Couples in a couples-focused alcohol treatment engaged in more collaborative problem-solving than couples in an individual-focused alcohol treatment condition (Walitzer et al, 2013).

Change is thought to occur through a number of pathways; not only is abstinence from alcohol actively rewarded by the non-drinking partner, but both partners also are encouraged to develop a deeper repertoire of shared enjoyable experiences and to actively work on improving communication patterns.

Proposed Mechanisms of Behavior Change - Mediators

Four mechanisms of behavior change have been proposed for ABCT (McCrady & Epstein, 2015): (a) IP motivation; (b) IP coping skills; (c) SO support; (d) couple interactions. A small body of research has addressed all but the impact of ABCT on IP coping skills.

IP motivation

Hunter-Reel, McCrady, and Hildebrandt (2009) proposed that pre-treatment social support from the SO and others may lead to better treatment outcomes by impacting IP motivation. In an empirical test of this hypothesis with a sample of women receiving either CBT or ABCT, Hunter-Reel, McCrady, Hilderbrand, & Epstein (2010) found that pre-treatment social support for not drinking from the SO and others predicted greater IP motivation at the end of treatment, which in turn predicted a lower percentage of drinking days six months post-treatment. Thus, female IP motivation mediated the relation between social support and drinking six months after treatment. This is clearly an important variable that warrants investigation in male IP samples. Given that alcohol use between partners is highly correlated (Leonard & Das Eiden, 1999; Leonard & Mudar, 2003; McLeod, 1993; Windle, 1997), it would be important to know if the direction of influence also holds for female SOs and male IPs, if couples demonstrating this benefit are discordant in the drinking to begin with, or if ABCT influences the drinking of both partners leading to improved outcomes for the IP.

SO support

Two studies have examined SO supportive behaviors as mediators of behavior change in ABCT. O'Farrell et al. (1998) found that greater use of the Antabuse contract correlated with a higher percentage of abstinent days (for 12 months post-treatment) and better relationship adjustment (for six months post-treatment). In their research using coded ABCT sessions, McCrady et al. (2014) did not find that SO behaviors as a set (including SO support, giving of general or alcohol-specific information, or change and counter-change talk) predicted drinking outcomes, but did find that the specific behavior of giving information during mid-treatment predicted a greater percentage of abstinent days in the second half of treatment.

Couple interactions

The same two studies of SO support also examined couple interactions as mediators of behavior change in ABCT, but results are somewhat contradictory. O'Farrell et al. (1998) reported that greater use of couple interaction skills taught during treatment was associated with a higher percentage of abstinent days and better relationship adjustment throughout three years from the beginning of treatment. McCrady et al. (2014) found that lower levels of confrontation from the IP during mid-treatment predicted a lower percentage of abstinent days in the six months after treatment. Clearly, more research on couple level interactions as mediators of treatment outcome is needed.

Other Research

Research on ABCT also has included substances other than alcohol, including other drugs and nicotine. Additionally, new adaptations to the ABCT protocol have been, or are being investigated, including brief interventions for alcohol use, brief interventions for drug use, ABCT for military families, and adaptations using web and smartphone delivery platforms. There also is a small body of literature investigating areas such as the cost-effectiveness of ABCT.

ABCT for Other Substances

In a study of 80 married and/or cohabiting males seeking treatment for a primary substance of abuse other than alcohol, Fals-Stewart and colleagues (1996) found significantly greater improvements in both substance use and relationship adjustment in males randomized to the BCT-D condition versus the control (individual and group cognitive behavioral coping skills training) condition over the 12-month follow-up period post-treatment. Though group differences in relationship adjustment and dyadic functioning generally disappeared by the 6-month post-treatment follow-up, group differences in percent days abstinent from drugs continued to be significant out to the 9 and 12 month follow-up time-points.

Epstein et al. (2007) adapted the McCrady ABCT model for males with other SUDs. This treatment development study examined pre- to three month post-treatment effect sizes in a group of 24 male IPs receiving stand-alone BCT-D with their female SOs. Drug and alcohol use decreased, as did drug-related consequences, and the majority of male IPs showed a significant increase in relationship satisfaction.

Early research demonstrated that specific partner behaviors are supportive of efforts at smoking cessation (e.g., Cohen & Lichtenstein, 1990). Some research has examined the efficacy of BCT for smoking cessation. Results to date have not suggested a benefit over traditional individual-based treatment. In an early study, McIntyre-Linsolver, Lichtenstein, and Mermelstein (1986) tested a couples-based behavioral approached to smoking cessation, finding no differential efficacy of the couple-based intervention. Similarly, LaChance and colleagues (2015) randomized 29 individuals smoking more-than-ten cigarettes a day into either a BCT condition consisting of seven conjoint therapy sessions and a subsequent eight weeks of nicotine replacement therapy, or a control condition consisting of seven individual sessions and eight weeks of nicotine replacement therapy. No significant differences in smoking cessation rates were found at the end of treatment, or at the three- and six-month follow-up time-points.

ABCT and Intimate Partner Violence

In addition to reducing substance use, BCT-A has been found to be of potential benefit to couples with a male IP with a history of intimate partner violence (IPV). In the 24 months after attending treatment, IPs who received BCT-A were shown to have fewer instances of IPV against their partners (O'Farrell, Murphy, Stephan, Fals-Stewart, & Murphy, 2004). Additionally, Schumm, O'Farrell, Murphy, and Fals-Stewart (2009) suggested that BCT-A appeared to be more effective than individual therapy at reducing both male-to-female and female-to-male physical and verbal aggression in couples with female partners with an AUD (Schumm, O'Farrell, Murphy, Fals-Stewart, 2009). A subsequent randomized clinical trial of BCT-A for women with alcohol dependence showed, however, that BCT and individual treatment were equally effective at reducing both male-to-female and female-to-male physical aggression (Schumm, O'Farrell, Hahler, Murphy & Muchowski, 2014). The authors did note, however, that baseline physical aggression was higher for individuals in the BCT group.

Secondary analyses by Fals-Stewart et al. (2002) also revealed a significant group difference in intimate partner violence (IPV) during the 12 months post-treatment. The percentage of couples endorsing at least one act of male-to-female physical aggression in the previous 12 months significantly decreased in the BCT-A condition from 43% at baseline to 17% at 12 months post-treatment. No such significant reduction was seen in the control condition (48% at baseline, 43% 12 months post-treatment). The group difference between these follow-up indicators of IPV was significant, and found to be mediated by frequency of drug use, frequency of heavy drinking, and relationship adjustment.

Adaptations of ABCT

Adaptations of ABCT and BCT-A interventions with treatment protocols using fewer than the traditional 12-15 sessions also have been investigated in recent years. A pilot study of a one-session brief family intervention (BFT) to encourage male drug abusers to attend aftercare post-detoxification showed a promising but non-significant improvement in treatment engagement over treatment as usual (TAU). The magnitude of this difference (r = 0.40) between the groups represents a medium effect size, and as such may be clinically meaningful (O'Farrell, Murphy, Alter, & Fals-Stewart, 2007). A similar study of the same one-session BFT intervention for patients in a detoxification unit for AUD showed a significant difference between the 24 patients in the BFT condition and the 21 patients in the TAU condition in terms of likelihood of entering aftercare post-detoxification. Ninety-two percent of BFT cases entered a continuing care program, whereas only 62% of TAU cases entered continuing care (O'Farrell, Murphy, Alter, & Fals-Stewart, 2008). McCrady and colleagues currently are investigating a three-session brief family-involved treatment with grant support from the National Institute on Alcohol Abuse and Alcoholism (NIH Project Number: 5R34AA023304).

Additional ongoing research points to novel and innovative adaptations of the ABCT protocol. Epstein and colleagues are currently investigating an adaptation of ABCT for post-deployment military personnel. This adaptation includes new modules specifically addressing Post-Traumatic Stress Disorder (PTSD), Traumatic Brain Injury (TBI), depression, and intimate partner violence (NIH Project Number: 5R34AA023027). Smelson and colleagues are testing a couple-based program for alcohol risk reduction in the National Guard and are adapting the ABCT protocol to be deliverable via telehealth (NIH Project Number 1R34AA023589).

Adaptations for other technologies also are being investigated. Woodall and colleagues are adapting the core concepts and content of ABCT to be deliverable to DWI offenders and their families via a smartphone application (NIH Project Number 1R41AA022850-01A1). Additionally, a recent study of a web-based coping skills program for women who have partners with an AUD resulted in significantly higher coping skills, significantly fewer depressive symptoms, and significantly lower situational anger when compared to wait-list controls (Rychtarik, McGillicuddy, & Barrick, 2015).

Economic Research

Cost-benefit and cost effectiveness analyses of BCT-A and its derivations generally have been supportive of the economic value of BCT-A. O'Farrell and colleagues (1996) found cost savings in one-year healthcare utilization that were five times greater than the cost of delivering BCT-A to AUD-diagnosed veterans. The delivery of a more intensive BCT-A plus relapse prevention protocol resulted in a higher number of days abstinent; however, the increased cost of delivering the more intensive protocol made BCT-A alone the more cost-effective intervention. A cost outcomes analysis of BCT-D delivered to polysubstance-abusing males found increased cost savings for BCT-D participants than for participants in an individual-based therapy (IBT) condition. The costs of delivering both interventions were equivalent; however, the reduction in total social costs (public assistance costs, justice system utilization costs, substance abuse treatment costs) was, on average, $6,600 in the BCT-D condition and only $1,900 in the IBT condition (Fals-Stewart, O'Farrell, & Birchler, 1997).

Future Directions

ABCT Efficacy and Effectiveness Research

There is no doubt that the picture of a “typical” couple has changed. Historically, empirical research on ABCT focused on a fairly specific demographic: heterosexual, non-Hispanic white dyads. As noted, the initial focus was on male IPs; this gave way to exploration of ABCT with female IPs. More recently, other populations of interest have emerged. The demographics of the United States are changing, and as such, culturally sensitive interventions continue to be an important area of research. The U.S. population is anticipated to continue to grow; by 2044, it is estimated that over half of all Americans will belong to a minority culture (U.S. Census, 2014). Different cultures have different customs, experiences, and expectations, not just around marriage and intimacy, but also around drinking, the role and effects of alcohol consumption, and help-seeking. The importance of linguistic conventions should be considered as well. Patterns of communication can vary tremendously in different cultures, and conventional ABCT approaches may promote a specific type of communication and couple-based problem-solving that is inconsistent with the mores of some cultures. Future ABCT research should address the greater cultural and racial diversity of contemporary couples.

Future research also should be expanded to include diversity of sexual orientations. This may be of particular importance as it has been shown that, relative to the general population, gay, lesbian and transgender individuals have higher rates of substance use issues, and that heavy use is more likely to persist over time (Centers for Disease Control, 2015). To date, there is a paucity of research testing the efficacy of BCT for gay and lesbian couples. A single study assessing the utility of inviting the SO of LGBT clients to attend at least one substance abuse treatment session found an association between partner attendance and higher abstinence rates, greater treatment satisfaction, and increased program completion (Senreich, 2010).

To date, ABCT research has not moved from the efficacy to effectiveness stage. Given the consistently positive findings for treatment efficacy, models are needed to adapt ABCT to enhance uptake in real-world treatment settings, and to test the effectiveness of the treatment in these settings. Given that a minority of clinical programs uses ABCT in any form (e.g., Forcehimes et al., 2010), research to identify and address barriers to utilization is needed as well.

ABCT Process Research

Process research on ABCT is in its nascence, and there are several moderating and mediating variables and proposed active ingredients that remain unexplored. For example, although there has been some investigation of the moderation of relationship satisfaction on treatment outcomes, relationship stability has remained completely unexplored. Although the influence of relationship satisfaction on stability appears straightforward, the association between relationship satisfaction and stability is actually influenced by a variety of factors, and satisfaction accounts for only 8% of the variance in stability for men and 18% of the variance in stability for women (Karney & Bradbury, 1995). Also unexamined are the influence of race/ethnicity, age, and length of the relationship as moderators of response to ABCT. Previous research has shown that non-Hispanic White couples report relationship satisfaction as a main reason for dissolution, while African-American couples report other factors such as substance use, infidelity, and spending money as reasons for dissolution (Amato & Rogers, 1997). Moderating factors such as these may play important roles in determining which couples choose ABCT and which couples benefit most from ABCT.

Future research also should further explore proposed active ingredients and mechanisms of change of ABCT. For example, do improvements in relationship functioning and alcohol use occur through motivational enhancement, IP or SO skills training, or a combination of those variables? In addition, are improvements in IP and SO coping skills active treatment ingredients? Lastly, are SO support and engagement necessary in the beginning stages of treatment for ABCT to be effective, or does ABCT improve SO support and engagement through active ingredients such as relationship enhancement?

Other Research

Because several adaptations of ABCT currently are underway, the results of these studies will provide important guides for future research. In addition to needs in efficacy, effectiveness, and process research, several other future directions would be important to explore. First, additional research is necessary to confirm the promising preliminary findings supporting efficacy for substances other than alcohol, and beneficial reductions in intimate partner violence. Second, with the high rates of co-occurrence of other psychiatric disorders with AUDs or other SUDs, conjoint models that are explicit in addressing alcohol and drug use along with other disorders are needed. Third, the integration of conjoint models into AUD treatment in primary care settings is largely unexplored. With the increasing trend toward health care homes to address both medical and behavioral health needs, development and testing of adapted ABCT models in these integrated healthcare settings would be of value. Fourth, although O'Farrell's model provides explicitly for behavioral contracts to support use of alcohol treatment-specific medications (e.g., Antabuse), the integration of medications into ABCT is largely unexplored.

Summary and Conclusions

ABCT is a conjoint approach to alcohol treatment with a clear conceptual base and good empirical support for the efficacy of the treatment. Despite these strengths, the uptake of ABCT in clinical practice has been limited, and the development of dissemination strategies and more flexible models applicable to a broader range of populations are clear directions for the future.