Brief Strategic Family Therapy: An Intervention to Reduce Adolescent Risk Behavior

Abstract

This article reviews the brief strategic family therapy (BSFT; J. Szapocznik, M. A. Scopetta, & O. E. King, 1978, The effect and degree of treatment comprehensiveness with a Latino drug abusing population. In D. E. Smith, S. M. Anderson, M. Burton, N. Gotlieb, W. Harvey, & T. Chung, Eds, A multicultural view of drug abuse, pp. 563–573, Cambridge, MA: G. K. Hall & J. Szapocznik, M. A. Scopetta, & O. E. King, 1978, Theory and practice in matching treatment to the special characteristics and problems of Cuban immigrants, Journal of Community Psychology, 6, 112–122.) approach to treating adolescent drug abuse and related problem behaviors. The treatment intervention is reviewed, including specialized features such as engagement of difficult families. Empirical evidence supporting the BSFT approach is presented. We then illustrate ways in which clinicians can use the model with troubled families whose adolescents may be at risk for drug use and HIV. Finally, future directions for BSFT research are described.

In this article, we describe the development of, and research findings testing brief strategic family therapy (BSFT; Szapocznik, Scopetta, & King, 1978a, 1978b) over the last four decades, along with the continuing evolution of our program of research based on lessons learned. We present a brief overview of the BSFT model; research on BSFT’s clinical interior, treatment outcomes, and the effects of therapist behaviors on adolescent and family outcomes. We conclude with a review of lessons learned in moving research findings into practice and for future research on implementation of the BSFT approach in community settings.

The BSFT Model

BSFT is a short-term (approximately 12 sessions), family-treatment model developed for youth with behavior problems such as drug use, sexual risk behaviors, and delinquent behaviors. Developed over nearly 40 years of research at the University of Miami’s Center for Family Studies, the BSFT approach operates based on the premise that families are the strongest and most enduring force in the development of children and adolescents (Gorman-Smith, Tolan, & Henry, 2000; Steinberg, 2001; Szapocznik & Coatsworth, 1999). BSFT targets families in which youth engage in clusters of risk-taking or problematic behaviors, including drug and alcohol use, delinquency, affiliation with antisocial peers, and unsafe sexual activity (Jessor & Jessor, 1977; Willoughby, Chalmers, & Busseri, 2004). Families of behavior-problem youth tend to interact in ways that permit or promote these problems (Véronneau & Dishion, 2010). The goal of BSFT, therefore, is to change the patterns of family interactions that allow or encourage problematic adolescent behavior. By working with families, BSFT not only decreases youth problems, but also creates better functioning families (Santisteban et al., 2003). Because changes are brought about in family patterns of interactions, these changes in family functioning are more likely to last after treatment has ended, because multiple family members have changed the way they behave with each other.

In most cases, drug abusing and delinquent adolescents are referred to treatment by the juvenile justice system. On occasion, adolescents may be referred by schools or social service agencies. Our research indicates that, before entering treatment, families with troubled youth are often hopeless and blaming in their view of the problem, and in family members’ relationships with each other (Coatsworth, Santisteban, McBride, & Szapocznik, 2001; Santisteban et al., 1996; Szapocznik et al., 1988). Moreover, the same family interactional problems that help to maintain the adolescent’s symptoms often also prevent the family from working together to get into treatment. Getting the family into treatment is often as challenging as treating the adolescent’s behavior problems and the family processes that maintain these problems. As a result, the BSFT model uses the same types of intervention strategies to engage and retain families in treatment as it uses to reduce the adolescent’s presenting problems.

Our early formative research (Szapocznik, Scopetta, & King, 1978a, 1978b; Szapocznik, Scopetta, Kurtines, & Aranalde, 1978) indicated that Cuban families in Miami, for whom the BSFT approach was developed, tended to value family connectedness over individual autonomy, and that they tended to focus on the present rather than on the past. As a result, we sought to develop a treatment model that would align with this value structure. Family connectedness is emblematic of the critical role that families play in the Cuban immigrant population. The present orientation required that we quickly address the family’s presenting concerns.

The BSFT intervention was therefore formulated as an integrative model that combines structural and strategic family therapy techniques to address systemic/relational (primarily family) interactions that are associated with adolescent problem behaviors. The structural components of the BSFT treatment draw on the work of Minuchin (Minuchin, 1974; Minuchin & Fishman, 1981). The strategic aspect of the BSFT approach was influenced by Haley (1976) and Madanes (1981). The integration of structural and strategic approaches to family therapy led us to develop a treatment that is problem-focused, planful, and practical—focusing primarily on identifying and enacting the changes necessary to ameliorate the adolescent’s presenting problems. Other family issues, such as problems between the parent figures, are not addressed unless they are directly related to the adolescent’s problem behaviors, such as drug use or risky sexual behaviors.

Not surprisingly, the BSFT approach shares a number of characteristics, such as a systems orientation, in common with other family-based therapies, such as multidimensional family therapy (Liddle & Hogue, 2001), functional family therapy (Alexander & Robbins, 2010), and multisystemic therapy (Henggeler, Schoenwald, Borduin, Rowland, & Cunningham, 1998). However, the BSFT approach is unique in that it focuses on diagnosing family interactional patterns and restructuring (i.e., changing) the family interactions associated with the adolescent’s problem behaviors. One of the major innovations of the BSFT approach has been the notion that challenges in engaging families into treatment are derived from the same interactional problems that are maintaining the adolescent’s symptoms. The specialized engagement procedures developed to address these challenges (Szapocznik, Muir, & Schwartz, in press) have revolutionized the field of family therapy.

BSFT is a manualized intervention (Szapocznik, Hervis, & Schwartz, 2003) that targets structural, interactional patterns in the adolescent’s family environment, and that creates changes in these patterns by strategically intervening to disrupt or alter these interactional patterns. There are three core principles on which BSFT is built. The first is that BSFT is a family-systems approach. “Family systems” means that family members are interdependent. The experiences and behavior of each family member affect the experiences and behavior of other family members. According to family-systems theory, for example, the troubled adolescent is a family member who displays risk-taking behaviors such as drug use and unsafe sexual activity that reflect, at least in part, what else is going on in the family system (Szapocznik & Kurtines, 1989). As such, the adolescent’s behavior can be said to reflect maladaptive family interactions. We define maladaptive interactions as those exchanges in which the family repeatedly engages in the intent to achieve a certain outcome (e.g., eliminate adolescent drug use), but that continue to be used, despite clear evidence that these interactions do not work.

Hence, the second BSFT principle is that the family’s habitual or repetitive patterns of interaction influence the behavior of each family member. Patterns of interaction are defined as the sequential behaviors among family members that become habitual and repeat over time. An example is an adolescent who disrupts fights between her two caregivers (e.g., her mother and grandmother) by attracting attention to herself, thereby distracting the two caregivers from their conflict and redirecting their attention to the adolescent. In extreme cases, the adolescent may suffer a drug overdose, engage in high-risk sexual behavior with multiple partners, or get arrested as a way of distracting her mother and grandmother when they are engaged in a severe conflict. This kind of adolescent behavior is known as triangulation (Bowen, 1978), because the adolescent (a third party) is inserting herself (or is inserted) into the conflict between her two caregivers. The role of the BSFT counselor is to identify the patterns of family interactions that are associated with the adolescent’s behavior problems. For example, a mother and grandmother who are arguing about rules and consequences for a problem adolescent never reach an agreement because the adolescent disrupts their arguments with self-destructive attempts at attracting attention.

The third principle of BSFT is to plan interventions that are problem focused and targeted—that is, that target these repetitive maladaptive patterns of family interactions, while strengthening adaptive patterns of interaction (e.g., caregivers sharing their concerns about the daughter) that will achieve the caregivers’ goal of reducing the adolescent’s problematic and risky behavior. BSFT interventions may attempt to change, for example, the way in which mother and grandmother attempt to establish rules and consequences for the adolescent, but fail because the adolescent disrupts the mother–grandmother discussion. Interactions become the target for intervention when they are directly linked to the adolescent’s problem behaviors.

BSFT interventions are organized into four theoretically and empirically supported domains (Robbins et al., 2011a; Szapocznik & Kurtines, 1989). Each of these domains of intervention is used throughout the treatment process, although some are used more often than others in specific phases of treatment. Early sessions are characterized by joining interventions intended to establish a therapeutic alliance with each family member and with the family as a whole. Joining requires that the therapist demonstrate acceptance of and respect toward each individual family member, as well as acceptance of and respect toward the way in which the family as a whole is organized. Early sessions also emphasize tracking and diagnostic enactment interventions that are designed to systematically identify adaptive and maladaptive family patterns of interactions, and to use these patterns of interactions to build a treatment plan. A core feature of tracking and diagnostic enactment techniques is that the therapist encourages family members to behave as they would if the counselor were not present. This means encouraging family members to speak with each other about the concerns they raise in therapy, rather than directing comments to the therapist. Indeed, when family members do address the therapist, the therapist asks the family member to redirect the statement or question to the person referenced in the statement. For example, if a father says to therapist, “You know, my wife is all wrapped up in our son and has no time for me,” the therapist will ask the father to direct this concern to his wife. Once this happens and the wife responds, an overlearned family pattern of interaction is likely to be enacted in the present in front of the therapist. As noted, although therapists are most likely to encourage family interactions and diagnose interactional patterns in early sessions, these techniques are used throughout the course of therapy.

Considerable work has gone into defining the structural diagnostic classifications on which the treatment plan is built; we refer the reader to our work on family structural (i.e., repetitive patterns of interactions) diagnosis (Szapocznik et al., 1991). Briefly, diagnoses are made on the dimensions of organization (e.g., hierarchy, patterns of alliances between/among family members), resonance (extent of emotional closeness or distance between specific family members), developmental stage (age-appropriateness of family roles), life context (conditions affecting the lives of the family or its members, such as divorces, deaths, crime-ridden neighborhoods, etc.), identified patienthood (the extent to which a single family member is “blamed” for all of the family’s problems), and conflict-resolution style.

Reframing interventions are utilized to reduce negative affect in family interactions while creating a motivational context for change. Over the course of treatment, therapists are expected to maintain an effective working relationship with each family member (joining), to facilitate within-family interactions (tracking and diagnostic enactment), and to transform negative affect (often reflective of overly strong family bonds) into constructive interactions that establish a motivational context for change. For example, consider a case in which a father is angry at his daughter for getting pregnant. The daughter withdraws emotionally as her father vents his anger at her. The therapist reframes the father’s anger into caring by stating, “I can see how concerned you are for your daughter. You had so many dreams for her and you are worried that they will not be possible now. You must have a great deal of love for your daughter for her missteps to make you so angry.” The father might then respond sadly, “You are damned right. I am afraid that she has ruined her future, and she could have HIV—she won’t tell me if she has been tested.” The therapist would then turn to the daughter and say, “Did you know that your dad is worried about you?”

Because reframing by promoting constructive interactions creates a motivational context for change, it serves as a natural springboard for restructuring interventions that transform family relations from problematic to effective and mutually supportive. Such restructuring interventions include: (a) Directing, redirecting, or blocking communication, (b) changing family alliances, (c) helping families to develop conflict resolution skills, (d) developing effective behavior management and conflict resolution skills, and (e) fostering positive parenting and parental leadership skills. All of these interventions involve assigning in-session tasks, followed by out-of-session “homework” tasks once the in-session tasks are proceeding well. For example, parent figures might be asked to engage in a conversation about managing the adolescent’s behavior, and the therapist will block the adolescent from interfering with the conversation. For another example, an adolescent and a disengaged father figure might be asked to engage in collaborative tasks together, as a way of building a positive relationship. If successful within therapy, these activities would then be assigned as homework tasks.

Engagement

When families are not able to agree on (or even successfully discuss) ways to manage an adolescent’s negative behavior, it is unlikely that they will be able to negotiate coming to therapy together. Further, if family members believe that the adolescent is “the problem,” they may think that only she or he needs to be in therapy. Indeed, the same interactional problems that maintain the adolescent’s symptoms are also associated with the family’s inability to come to treatment. Within the BSFT model, specialized engagement techniques have been developed and evaluated (Coatsworth et al., 2001; Santisteban et al., 1996; Szapocznik et al., 1988). The same intervention domains used in BSFT treatment—joining, tracking and diagnostic enactment, and reframing—are also utilized to engage families into therapy. Often one essential family member, a powerful problem youth or an alienated father, may not want to come to treatment. With the approval of the person (usually the mother) who called the therapist for help, the therapist will reach out to, and join with, the family member who is unwilling to attend therapy in an effort to assure that family member that she or he has something to gain from coming to treatment. From speaking with the family member who called for help, it is often not difficult for a therapist to identify the interactional challenges for a family to come into treatment. The therapist begins to explore the family interactions in a first call by giving the caller a task: “Bring all the members of the family into the first session.” The organization of the family will become apparent when the caller either responds that, “My son won’t come to treatment,” or “My husband won’t come to treatment,” or “It is best if just my son and I come—it is not necessary to bring my husband.” In the first and second cases, the caller believes that she lacks the influence needed to bring that family member into treatment. In the third case, the caller either prefers not to bring her spouse, or is at best ambivalent about bringing him. In each case, and with the caller’s approval, the therapist will insert him- or herself into the family process by reaching directly to the family member who either does not want to come to treatment, or whom the caller is not eager to bring to treatment, as a way of getting around the interactional patterns that interfere with bringing all family members into treatment.

BSFT is a flexible approach that can be utilized with a broad range of family situations (e.g., two-parent families, single-parent families, stepfamilies, multigenerational families), in a variety of service settings (e.g., mental health clinics, drug-abuse treatment programs, and other social-service settings), and in a variety of treatment modalities (e.g., as a primary outpatient intervention, in combination with residential or day treatment, as an aftercare/ continuing-care service to residential treatment, and for family preservation or reunification). Moreover, the BSFT approach is applicable across a range of ethnic/cultural groups.

Goals of Brief Strategic Family Therapy

In BSFT, whenever possible, preserving the family is desirable. That is, wherever possible, the focus should be on changing family dynamics rather than removing the adolescent from the family or prompting family members to leave the home. Within this approach to family preservation, two goals must be set: (a) To eliminate or reduce the adolescent’s problem behaviors, such as drug use and other risk-taking behaviors, known as the “strategic or symptom focus,” and (b) to change the family interactions that are associated with the adolescent’s problem behaviors, known as “system focus.” An example of system focus occurs when a parent directs his anger toward the youth who is exhibiting the problematic behavior. The parent’s negativity toward the adolescent serves only to increase the youth’s problematic behaviors, and the adolescent’s problematic behaviors increase the parents’ negativity (Koh & Rueter, 2011). At the family systems level, the counselor intervenes to change the way family members behave toward each other—and therefore to interrupt the cycle of negativity between family interactions and adolescent problem behavior. This will prompt family members to speak and act in ways that promote more supportive family interactions, which, in turn, will make it possible for the adolescent to reduce his or her problem behaviors.

BSFT Outcome Studies

BSFT has been found to be efficacious in treating adolescent drug abuse, conduct problems, associations with antisocial peers, and impaired family functioning. All of these outcomes are important risk factors for unsafe sexual behavior (e.g., Bersamin et al., 2008; Guo et al., 2005). The BSFT model has been evaluated in a number of randomized clinical trials evaluating the efficacy and effectiveness of the model, and identifying specific therapist behaviors that are associated with the most favorable adolescent and family outcomes. These studies have led the United States Department of Health and Human Services to label the BSFT approach as one of its “model programs,” and to be included in the National Registry of Evidence-Based Programs and Practices (NREPP; http://nrepp.samhsa.gov/viewintervention.aspx?id=151). We should note that the majority of the earlier studies on BSFT were conducted with Hispanic families (Coatsworth et al., 2001; Santisteban et al., 1996, 2003; Szapocznik et al., 1988, 1989). The model was originally developed to address acculturation discrepancies between Cuban adolescents and their parents in Miami (Szapocznik, Scopetta, & King, 1978a, 1978b). Indeed, at the time when BSFT was developed, Szapocznik et al. (1978) found that nearly all of the drug-abusing and delinquent adolescents referred for treatment evidenced both cultural and normative developmental conflicts with their parents. However, BSFT effectiveness research has suggested that the model is equally applicable to African Americans, Hispanic Americans, and White Americans (Robbins et al., 2011b), and the model is currently being used broadly with a variety of populations in the United States and several countries in Europe.

BSFT efficacy

The efficacy of the BSFT model in reducing behavior problems and drug abuse has been tested in two randomized, controlled, clinical trials. In the first trial, Szapocznik and colleagues (1989) randomized behavior-problem and emotional-problem 6–11-year-old Cuban boys to BSFT, individual psychodynamic child therapy, or a recreational placebo/control condition. The two treatment conditions, implemented by highly experienced therapists, were found to be equally efficacious, and more efficacious than recreational control, in reducing children’s behavioral and emotional problems and in maintaining these reductions at 1-year posttermination. However, at 1-year follow-up, the BSFT condition was associated with a significant improvement in independently rated family functioning, whereas individual psychodynamic child therapy was associated with a significant deterioration in family functioning.

In a second study, Santisteban and colleagues (2003) randomly assigned Hispanic (half Cuban and half from other Hispanic countries) behavior-problem and drug-abusing adolescents to receive either BSFT or adolescent-group counseling. The adolescent-group counseling condition was modeled after a widely used program in our community. The BSFT condition was significantly more efficacious than group counseling in reducing conduct problems, associations with antisocial peers, and marijuana use, and in improving observer ratings of family functioning. Baseline family functioning emerged as a moderator of treatment effects. For families entering the study with comparatively good family functioning, family functioning remained high in the BSFT condition, whereas it deteriorated in the families of adolescents in group therapy. For families entering the study with comparatively poor family functioning, the BSFT condition significantly improved family functioning, whereas family functioning did not improve in families assigned to adolescent-group therapy. Moreover, adolescent-group counseling was associated with clinically significant increases in marijuana use.

We have also explored the extent to which the BSFT model can be used with African American as well as Hispanic adolescents with behavior problems. In an uncontrolled study examining the suitability of the BSFT approach for adolescents from both ethnic groups, Santisteban and colleagues (1997) assessed conduct problems, delinquency in the company of peers, and observer-rated family functioning before and after BSFT treatment. Although BSFT significantly reduced association with antisocial peers and improved family functioning for both Hispanics and African Americans, BSFT treatment was significantly more efficacious in reducing association with antisocial peers among African Americans than among Hispanics. Conversely, the BSFT treatment was significantly more efficacious in improving family functioning among Hispanics than among African Americans. These early findings suggest that BSFT may benefit ethnic groups through different mediational pathways.

BSFT engagement

The efficacy of BSFT engagement was tested in three separate studies with Hispanic adolescents with behavior problems and their families. In the first study (Szapocznik et al., 1988), Hispanic (mostly Cuban) families with drug-abusing adolescents were randomly assigned to BSFT + engagement as usual (the control condition) or to BSFT + BSFT engagement (the experimental condition). The engagement-as-usual condition was modeled after community-based adolescent outpatient programs’ approaches to engagement in the Miami area. The results of the study revealed that 93% of the families in the BSFT engagement condition, compared with only 42% of the families in the engagement-as-usual condition, were engaged in treatment (defined as all family members in the household attending an admission session). Moreover, 75% of families in the BSFT engagement condition completed treatment (defined as reaching a mutual decision with the therapist that treatment should be terminated), compared with only 25% of families in the treatment-as-usual (TAU) condition.

In the second study (Santisteban et al., 1996), families were randomly assigned to a BSFT engagement or engagement control (no specialized engagement) condition. In the BSFT engagement condition, 81% of families were successfully engaged (defined as attending an intake and a first therapy session), compared with 60% of the families in the engagement control condition (defined as attending the admission session plus one family therapy session). A major finding of this study was that the effectiveness of BSFT-engagement procedures was moderated by Hispanic nationality. Among the non-Cuban Hispanics (composed primarily of Nicaraguan, Colombian, and Puerto Rican families) assigned to the BSFT engagement condition, the rate of engagement was high (93%) compared with the lower rate for Cubans assigned to this same condition (64%). Most of the Cuban families had United States-born adolescents, whereas the majority of adolescents from other national backgrounds were foreign-born. Hence, the families of United States-born Cuban adolescents had spent more time in the United States than the families of non-Cuban, foreign-born adolescents. Evidence suggests that United States-born Hispanic adolescents tend to be more Americanized than adolescents born outside the United States (Schwartz, Pantin, Sullivan, Prado, & Szapocznik, 2006). There is evidence that, in Hispanic families, acculturation to American values and behaviors is associated with decreased orientation toward family (Sabogal, Marin, Otero-Sabogal, Marin, & Perez-Stable, 1987). As a result, it is possible that the lower engagement rate found for Cubans was due to higher rates of Americanization in the Cuban families. It is possible that more Americanized families perceive less need for family involvement in adolescent drug-abuse treatment. Given this finding, specific family reconnection strategies, focusing on reorientation toward the importance of family, have been incorporated into the current version of BSFT engagement.

A third study (Coatsworth et al., 2001) tested the ability of BSFT + BSFT engagement to engage and retain adolescents and their families in comparison with a community control condition. An important aspect of this study was that the control condition was implemented by a community treatment agency and, as such, was less subject to the influence of the investigators. The Hispanic adolescents and families in this study were primarily Cuban or Nicaraguan. Findings in this study indicated that BSFT engagement successfully engaged 81% of families into treatment—significantly higher than the 61% rate in the community control condition. Likewise, among families who were successfully engaged, 71% of BSFT cases, compared with 42% in the community control condition, were retained to treatment completion.

BSFT effectiveness

An effectiveness trial (Robbins et al., 2011b) of the BSFT approach was conducted in the context of the National Institute on Drug Abuse’s National Drug Abuse Treatment Clinical Trials Network. In this study, both therapists and families were randomized within clinics. As discussed below under lessons learned and future directions, this design did not represent the implementation approach used by evidence-based family treatment programs with troubled adolescents. The study compared BSFT and TAU (which was allowed to vary based on whatever treatment the agency typically provided for drug-using adolescents) by randomizing 480 families of adolescents (213 Hispanic, 148 White, and 110 Black; 377 male, 103 female) referred to drug-abuse treatment at eight community treatment agencies located around the United States. Seventy-two percent of these adolescents were referred for treatment by the juvenile justice system, and most of the remaining cases were referred from residential treatment. Services in both conditions were delivered by therapists in community agencies. These therapists were randomized within agency to deliver either the BSFT or TAU modalities.

Engagement and retention

Families in TAU were 2.33 times (11.4% BSFT; 26.8% TAU) more likely to fail to engage (defined as not completing at least two sessions) than families in the BSFT condition. Families in TAU were 1.41 times (40.0% BSFT; 56.6% TAU) more likely to fail to retain (defined in this study as completing fewer than eight sessions) than families in BSFT. These differences were statistically significant and were consistent across the three racial/ethnic groups in the study: African Americans, Hispanic Americans, and White Americans.

Treatment duration

Therapy took much longer to administer than expected. The usual expectation is that BSFT therapy should last approximately four months, which is consistent with our implementation experience. However, the median length of treatment for those participants who were retained in treatment across both conditions was approximately 8 months for both conditions. As discussed later, this difference between prior and current experiences in delivering BSFT may have occurred because BSFT was implemented by therapists who had additional caseloads, often involving other treatment approaches, in addition to their BSFT caseload for the study.

Effects on adolescent drug use

Drug use was operationalized as the number of self-reported drug-using days within each 28-day period. There were no significant differences by treatment condition in terms of the number of drug-using days per 28-day period at 1-year postrandomization. However, using nonparametric analyses, the median number of self-reported drug-use days per month at the 12-month follow-up was significantly higher in the TAU condition (3.5 days) than in the BSFT condition (2 days). It should be noted that the median number of drug-use days was low and restricted, with an interquartile range between 1 and 3 days of self-reported use per month. Such a restricted range made it difficult to detect statistically significant or clinically meaningful differences in substance use trajectories. The overwhelming majority of adolescents in the study were referred from residential treatment or from juvenile justice, both of which involved surveillance (and limited opportunities to engage in drug use). These referral sources may have been responsible for the relatively low baseline rates of drug use, and in the case of the juvenile justice referrals, continued surveillance may have been responsible for the low levels of drug use over time.

Family functioning

Patterns of findings for family functioning differed between adolescent and parent reports. The BSFT condition produced significantly greater improvements in parent-reported family functioning (defined as positive parenting, parental monitoring, effectiveness of parental discipline, parental willingness to discipline adolescents when necessary, family cohesion, and absence of family conflict) than the TAU condition. Adolescents in both conditions, however, reported significant improvements in family functioning, with no statistically significant differences by treatment condition.

Parental functioning

Post hoc analyses demonstrated that BSFT was more effective than TAU in reducing alcohol use in parents, and that this effect was mediated by parental reports of family functioning. In addition, BSFT as compared with TAU, had its strongest effect in reducing adolescent drug use among youth whose parents used drugs at baseline (Horigian et al., submitted).

BSFT Therapist Behavior, Therapy Process, and Their Relationship to Outcomes

Research has demonstrated that negativity in family interactions in the first session leads to failure to retain families in treatment past the first session (Fernandez & Eyberg, 2009); that families are more likely to engage in treatment if negativity is reduced (Robbins, Alexander, & Turner, 2000); that reframing is an effective method of reducing negativity (Moran, Diamond, & Diamond, 2005); and that reframing is the technique that is least likely to damage therapists’ rapport (alliance, bond) with family members (Robbins et al., 2006). Research also shows that early engagement requires the therapist to maintain a balanced bond with the parent (often the father figure) and the problem youth. Research on BSFT has shown that if, in the first session, the therapist does not develop a balanced set of bonds with the parent and the youth, this imbalance leads to early dropout from treatment (Robbins et al., 2000). These findings have been incorporated into BSFT treatment as conducted today.

Effects of BSFT therapist adherence and behaviors on outcomes

Using data from the effectiveness study, Robbins et al. (2011a) examined the extent to which BSFT therapists implemented the treatment protocol properly. To do this, adherence items were rated in terms of the four theoretically and clinically relevant expected/prescribed therapist behaviors: joining, tracking and eliciting enactments, reframing, and restructuring. These items were completed by trained independent raters who watched videos of therapy sessions. These items demonstrated adequate factorial validity and converged well with clinical supervisor ratings. Mean levels of adherence varied over time in theoretically and clinical relevant ways. Therapist adherence to BSFT was associated with:

(1) Engagement

Using adherence ratings for the first session, with engagement defined as whether or not the family attended a second treatment session. Results revealed that higher levels of restructuring and reframing (reducing negativity) significantly increased the likelihood of families being engaged into treatment. Because joining and tracking and diagnosis were high across most cases, what distinguished cases that came to a second session from those that did not was reframing and restructuring, the technique domains that therapists found most challenging.

(2) Retention

The impact of adherence on retention was evaluated using adherence ratings for Sessions 2–7, with retention defined as a family attending at least eight sessions. Results indicated that higher levels of all four technique domains—therapist joining, tracking and enactment, reframing, and restructuring—predicted significantly higher rates of retention. A 1-SD increase in reframing predicted a 19% increase in the likelihood of retention; a 1-SD increase in joining predicted a 22% increase in the likelihood of retention; a 1-SD increase in restructuring predicted a 59% increase in the likelihood of retention; and a 1-SD increase in tracking and eliciting enactment predicted a 62% increase in the likelihood of retention.

(3) Family functioning

Overall joining levels predicted improvements in observer-reported family functioning.

(4) Adolescent drug use

The effect of prescribed therapist behaviors on adolescent drug use was complex. Across time, as would be expected, joining decreased, and restructuring increased. Smaller declines in joining and larger increases in restructuring predicted significantly less adolescent drug use at the 12-month follow-up. That is, therapists who were high in joining in early sessions and remained so throughout treatment were associated with “better” adolescent drug-use outcomes. Therapists whose attempts to restructure maladaptive family interactions increased the most during the course of treatment were also associated with “better” adolescent drug-use outcomes. Thus, therapists who failed to implement sufficient numbers of restructuring interventions were less able to affect the youths’ drug use.

These results indicate that, within a sample of therapists from community agencies, therapists’ clinical interventions follow a pattern that is consistent with the theory behind the BSFT model. Indeed, the specific therapist behaviors prescribed by the BSFT approach are needed to engage families into treatment, retain them, improve family functioning, and reduce adolescent drug use. However, when therapists did not engage sufficiently in these behaviors, adolescent outcomes tended to suffer. The authors concluded that adherence ratings were affected by a number of systemic factors, including overburdened therapists and therapists’ lack of embeddedness within dedicated BSFT units.

Future Directions: Implementing BSFT in Community Practice

Now that the BSFT model has been found to be efficacious (in controlled clinical trials) and effective (in a community-based trial), the next step is to conduct more rigorous research on implementing the model in community practice (see Henggeler, 2011, for a review of the stages of treatment evaluation and dissemination). Work in this direction is currently underway, and we are proceeding using the lessons that we have learned in the Clinical Trials Network (CTN)-effectiveness trial. Our BSFT Institute (Miami, FL) has been engaged in a Stage I study of the implementation of BSFT in community agencies across the United States and some European countries. Engagement, retention, and recidivism-prevention rates have been excellent.

In real-world implementation of BSFT and other evidenced-based family-therapy models, such as multisystemic therapy (Henggeler & Sheidow, 2012; Letourneau et al., 2009), multidimensional family therapy (Henderson, Dakof, Greenbaum, & Liddle, 2010), and functional family therapy (Sexton & Turner, 2010), groups of therapists are assigned to administer only the evidence-based family-therapy intervention (i.e., they have no other caseload), and units are created that are fully committed to the intervention model with appropriate support from the agency leadership. This support is essential to ensure adherence to various aspects of the model, including availability of therapists when families are available (e.g., evenings and Saturdays). Similarly, in BSFT real-world implementation, a certain number of active families (typically 12) are assigned to each therapist in a BSFT unit, and the therapist is evaluated based on her or his treatment outcome with these families. Moreover, when therapists are able to practice only BSFT and to work with a team of therapists who also practice only BSFT, fidelity to the model is fortified. Fidelity is essential because, as we demonstrated in our effectiveness study, delivery of prescribed therapist behaviors were directly related to improvements in all target outcomes—engagement, retention, family functioning, and drug use.

However, because of the conventional clinical trials format used with the CTN consortium, the BSFT effectiveness trial was conducted as a traditional individual-level randomized clinical trial (i.e., randomizing therapists and participants within each site). Therapists assigned to the BSFT condition were expected to conduct BSFT with study families in addition to their other caseloads using other intervention approaches. Our experience in the effectiveness trial clearly indicates (a) that many therapists felt overwhelmed with their caseloads, (b) that therapists did not have the time flexibility needed to conduct the “outside-of-session” work required by BSFT (e.g., engaging family members, being available on evenings and weekends when family members were free), (c) that some therapists felt burdened while others felt stimulated by the supervision requirements of an evidence-based practice, and (d) that therapist fidelity to the model faltered when a therapist did not have adequate organizational support. A better approach to evaluating BSFT in community settings is to establish matched clinic units (each with 4–5 therapists) and to randomize each pair of matched units to BSFT versus TAU. Implementation requires support from throughout the agency and its funders (Henggeler, 2011), and a systemic organizational intervention is likely needed to ensure successful implementation. We are currently documenting our experiences with this implementation strategy and are designing a randomized trial to ascertain the impact of these implementation methods.

Conclusion

In conclusion, in this article we have described the BSFT model and supportive empirical evidence garnered over four decades of research and practice. The model was initially designed to match the cultural values and preferences of Cuban immigrant families in Miami and brought together key principles from structural and strategic family therapy. BSFT shares common elements with other evidence-based family therapies for adolescent problem behavior, but the focus on diagnosing and restructuring present-time family interactions, as well as its specialized engagement techniques, are largely unique to BSFT. Further, although most of the earlier work testing the BSFT approach was conducted with Hispanic families in Miami, effectiveness research has suggested that the model can be used across racial/ethnic groups in the United States. Implementing the model in community settings will provide therapists with an effective tool to increase family involvement in therapy, increase retention, reduce adolescent drug use and related risk-taking behaviors, and reconfigure family interactions to support healthy development.