The Implementation of Functional Family Therapy (FFT) as an Intervention for Youth Probationers in Singapore

Abstract

Family functioning is predictive of youth recidivism in Singapore. However, there is a lack of family based interventions for youth offenders on community probation. Evidence‐based family interventions developed in Western populations, such as Functional Family Therapy (FFT), have been found to be effective in mitigating subsequent youth criminal behavior. However, no study has examined whether such interventions can be implemented and adapted for use in Eastern cultures. Thus, this paper sought to detail the implementation of FFT in Singapore. Rationale for the adoption of FFT is discussed, and key activities undertaken during the first 18 months of implementation are described. Preliminary data suggest that initial implementation efforts were successful. Challenges encountered, and implications in relation to the broader literature are discussed.

There is growing recognition of the detrimental impact of poor family functioning on youth recidivism. This trend has underscored the need for evidence‐based family interventions in the context of youth offender rehabilitation. It has been widely acknowledged that sound implementation is critical to the effectiveness of evidence‐based programs (EBPs; Breitenstein et al., 2010; Dusenbury, Brannigan, Falco, & Hansen, 2003). To date, however, few studies have focused on the implementation of EBPs (Caron, Bérubé, & Paquet, 2017). Specifically, no study has discussed the cross‐cultural transportability of family based EBPs. To address this gap, this paper sought to detail the implementation of FFT in Singapore, and present preliminary evidence of implementation outcomes. Although not an evaluation of the effectiveness of FFT on youth criminal behavior, this article is one of the first to describe and discuss the implementation of a Western‐developed EBP in an Eastern culture. It is hoped that the information and insights provided may benefit future efforts in the cross‐cultural implementation of EBPs in community settings.

Family Functioning and Youth Offending

The role of the family on youth offending cannot be underestimated. Four decades of international research have demonstrated that family related influences can function both as risk and protective factors in relation to youth criminal behavior (see Farrington, 2010, for a review). Strengthening family functioning should therefore be a key objective of rehabilitative interventions for youth offenders.

Many empirical studies concur that family functioning is inversely related to youth offending. Low levels of parental supervision have been associated with higher rates of youth offending, and higher likelihoods of the persistence of criminal behaviors into adulthood (Farrington & Loeber, 1999). Poor parental disciplinary practices, such as harsh and inconsistent discipline, have been linked with violent behavior and criminal delinquency later in life (Herrenkohl et al., 2000). Inadequate and sporadic discipline has also been associated with higher rates of youth recidivism (Carr & Vandiver, 2001). Moreover, poor family communication and cohesion has been found in several longitudinal studies to be predictive of delinquent behavior (Gorman‐Smith, Tolan, Zelli, & Huesmann, 1996).

Conversely, healthy levels of family functioning appear to mitigate the occurrence of delinquent behavior. Good parental supervisory practices have been associated with lower levels of youth criminal behavior (Henneberger, Varga, Moudy, & Tolan, 2016). Similarly, appropriate and consistent use of discipline during childhood has been found to attenuate the occurrence of delinquent behavior in adolescence (Hoge, Andrews, & Leschied, 1996). Also, healthy parent‐child communication has been linked with lower rates of juvenile substance use (Bogenschneider, Wu, Raffaelli, & Tsay, 1998). Finally, meta‐analytic studies have consistently found family based interventions to be effective in reducing delinquent and criminal youth behavior (Baldwin, Christian, Berkeljon, & Shadish, 2012; Woolfenden, Williams, & Peat, 2002). Taken together, these findings show that improving the quality of family functioning may be effective in reducing the recurrence of youth criminal behavior.

The Need for Family Based Interventions for Youth Offenders in Singapore

A rising trend in risk severity profiles of local youth probationers has been observed in recent times. In 2015, youth arrests made up 16.1% of all arrests in Singapore. (Ministry of Home Affairs, 2016). Though this number has remained relatively stable over the past 3 years, youth offenders recently discharged from probation presented with higher levels of risk at the onset of probation, compared to earlier cohorts. The proportion of probationers aged 18 and below with a recidivism risk rating of Moderate, High, or Very High at intake—as measured on the Youth Level of Service/Case Management Inventory 2.0 (YLS/CMI 2.0; Hoge & Andrews, 2011)—increased from 66% in 2012 to 79% in 2015 (Ministry of Social and Family Development, 2016a). Given this increase in the severity of risk profiles, it is imperative that community rehabilitation efforts employ evidence‐based interventions to prevent an accompanying rise in recidivism rates.

Local research has established the relationship between family functioning and youth reoffending rates. An examination of the predictive validity of YLS/CMI 2.0 ratings on overall recidivism in a sample of 3,264 youth offenders who previously served probation orders revealed that higher scores on the Family Circumstances/Parenting subscale—indicative of poorer family functioning—were linked with higher rates of recidivism, regardless of gender (Chu et al., 2015). A follow‐up study aimed to further elucidate this relationship by studying youth offending in relation to composite family profiles, instead of individual family‐related variables (Chng, Chu, Zeng, Li, & Ting, 2016). A latent class analysis (LCA) of eight family variables yielded three distinct family profiles: (a) intact families, (b) families with criminality, and (c) poorly managed families (Chng et al., 2016). Youth offenders from families that were poorly managed or with criminal histories were found to be more likely to (a) offend earlier in life, (b) have higher rates of reoffending, and (c) reoffend more quickly, compared to their counterparts from intact families (Chng et al., 2016). Collectively, these findings underscore the need for evidence‐based family interventions for youth probationers in Singapore.

Addressing this need has been challenging for several reasons. First, there is a dearth in the number of evidence‐based family interventions for youth offenders. A review by Henggeler and Sheidow (2012) sought to identify evidence‐based family programs based on criteria developed by two established standards—the American Psychological Association Task Force on the Promotion and Dissemination of Psychological Procedures (Chambless et al., 1998), and the Office of Juvenile Justice and Delinquency Prevention's Blueprints for Violence Prevention (Elliott, 1998). This review found that only four programs fulfilled criteria to be considered “evidence‐based”: (a) Multi‐Dimensional Treatment Foster Care (MTFC; Chamberlain, 2003), (b) Multi‐Systemic Therapy (MST; Henggeler, Schoenwald, Borduin, Rowland, & Cunningham, 2009), (c) Brief Strategic Family Therapy (BSFT; Szapocznik, Hervis, & Schwartz, 2003), and (d) Functional Family Therapy (FFT; Alexander, Pugh, Parsons, & Sexton, 2000; Alexander, Waldron, Robbins, & Neeb, 2013; Sexton, 2016). Second, few studies have investigated whether these family based EBPs can be successfully implemented to meet the needs of families from different cultural environments, without compromising program integrity (Hoagwood, 2005). Third, the implementation of family based programs necessitates changes to existing work processes in order to facilitate effective program delivery. Specifically, practitioners will require support for coping with logistical challenges inherent in shifting from individual‐ to family based therapy, and professional challenges associated with competent delivery of a new program. Failure to support high‐quality program delivery may jeopardize program impact on client outcomes (Blasé et al., 2010).

This paper aims to describe the implementation of FFT within the Clinical and Forensic Psychology Service (CFPS)—a state‐run psychology service overseeing the rehabilitation of high‐risk youth offenders in Singapore. Despite the multitude of outcome studies on FFT, only two have discussed its implementation—one in the United States (Zazzali et al., 2008), and another in the Netherlands (Breuk et al., 2006). Moreover, of the ten countries where FFT has been implemented (Robbins, Alexander, Turner, & Hollimon, 2016), Singapore is the only site not predominantly influenced by Western culture. At a broader level, there is a paucity of literature examining cross‐cultural implementation of Western‐developed EBPs in Eastern cultures. This paper represents the first attempt to document the implementation of FFT in an Eastern culture. It is hoped that the present work yields useful insights that could inform future endeavors in the cross‐cultural implementation of EBPs.

Functional Family Therapy

Functional Family Therapy is an intensive, family based intervention which aims to reduce delinquent behavior in high‐risk children and youth aged 11–18 years through improving the quality of family functioning. The FFT clinical model is integrative, embodying the core elements of systems therapy and cognitive‐behavioral therapy (Alexander et al., 2000). It hypothesizes that problem behaviors exhibited by delinquent youth stem from dysfunctional patterns of interaction within the family (Alexander, 2011; Sexton, 2010). By helping family members to understand the functions served by these behaviors, and their impact on family relations, the therapist sets the stage for the introduction of behavioral changes aimed at improving the quality of family relationships (Alexander et al., 2013).

Functional Family Therapy comprises three phases, each with distinct but related goals. The engagement/motivation phase is concerned with building a positive therapeutic alliance, instilling hope, and encouraging motivation for change within the family. It also aims to address risk factors associated with treatment dropout such as blaming, feelings of negativity or hopelessness, and scheduling conflicts. In the relational assessment/behavior change phase, the therapist helps family members gain insight to the functions served by problem behaviors, in the context of family relationships. This facilitates the replacement of problem behaviors with more adaptive ones, which serve the same function without adversely affecting family functioning. In this phase, therapists also assist the family in cultivating healthy communication and problem‐solving skills. Finally, in the generalization phase, the focus is on maintaining the positive changes in behavior, and helping the family to extend these changes to other spheres of life.

Evidence in support of FFT's effectiveness has been amassed over four decades of empirical research (see Robbins et al., 2016 for a review). Recent evaluations of FFT have found it to be effective across a wide range of problem behaviors in at‐risk youth. These include truancy (Slesnick & Prestopnik, 2009), violence (White, Frick, Lawing, & Bauer, 2013), and substance use (Rhoades, Campbell, & Bumbarger, 2011). A meta‐analytic study of randomized‐controlled trial (RCT) and quasi‐experimental studies showed support for FFT's overall robustness in mitigating youth delinquent behavior (Hartnett, Carr, Hamilton, & O'Reilly, 2017). Findings from non‐US sites, however, yield mixed findings. Positive findings, in the form of reduced reoffending rates, have been reported in Norway (Thøgersen, 2012), and Sweden (Hansson, Cederblad, & Höök, 2000). Two studies in Ireland (Graham, Carr, Rooney, Sexton, & Wilson Satterfield, 2014; Hartnett, Carr, & Sexton, 2016) reported success in implementation, as well as significant, sustained reductions in youth behavioral problems. However, a recent RCT conducted in the United Kingdom did not find any differences in reoffending rates or self‐reported delinquency between youth who received FFT, relative to those in a comparison group (Humayun et al., 2017). More research may therefore be needed to conclusively establish the effectiveness of FFT outside the United States.

New FFT sites abide by a standardized implementation protocol provided by the program developers. This protocol consists of three components: (a) clinical training over a three‐year period, (b) ongoing expert consultation, and (c) use of an electronic data system for monitoring adherence to documentation processes and staff competence. The protocol has been used to guide the implementation of FFT in over 300 sites worldwide, providing evidence of its protocol's utility over a myriad of community settings (Zazzali et al., 2008).

Implementation: Concepts and Definitions

Implementation has been defined as an organized series of actions associated with the delivery of an intervention or service (Fixsen, Naoom, Blase, & Friedman, 2005). Interest in implementation research stemmed from a growing awareness that the use of EBPs, in and of itself, was insufficient for achieving positive client outcomes (Institute of Medicine—Committee on Quality of Health Care in America, 2001). It has been hypothesized that this absence of expected treatment effects is rooted in the complexities associated with human service interventions, where service delivery could be affected by an unpredictable combination of human and environmental influences (Fixsen, Blase, Naoom, & Wallace, 2009). Thus, in the context of human services interventions, the goal of implementation is to equip and empower practitioners to consistently deliver the new program with a high level of fidelity, regardless of external circumstances (Fixsen et al., 2009).

Program fidelity—the extent to which a program or service is delivered in the way it was prescribed to be—is a core indicator of implementation success (Breitenstein et al., 2010; Kramer & Burns, 2008). Systematic and meta‐analytic reviews consistently report high levels of fidelity to be linked with better client outcomes; conversely, low levels of fidelity were associated with an absence of positive treatment effects (Durlak & DuPre, 2008; Goense, Assink, Stams, Boendermaker, & Hoeve, 2016). Hence, without evidence of program fidelity, it is not possible to attribute improvements in client outcomes to program effects.

The implementation process consists of four main stages—(a) exploration, (b) installation, (c) initial implementation, and (d) full implementation—and typically spans 2–4 years (Bertram, Blase, & Fixsen, 2015). A comprehensive description can be found in work by Fixsen and colleagues (Bertram et al., 2015; Fixsen et al., 2005, 2009). As FFT has only recently been implemented in Singapore, this paper focuses on activities and challenges from the exploration to the initial implementation stages.

The Implementation of FFT in Singapore

Activities undertaken during the exploration stage serve to assess whether it is feasible to adopt a new program (Fixsen et al., 2009). Adoption—the decision by an organization to implement a new program (Proctor et al., 2011)—is a plausible outcome if the program of interest is assessed to be (a) well‐aligned with the organization's goals, (b) suitable for meeting the needs of the target population, and (c) operationally feasible in light of existing resources (Bertram et al., 2015; Fixsen et al., 2005).

Drivers of Adoption

Functional Family Therapy was perceived to be a good fit in relation to the Ministry of Social and Family Development's (MSF) organizational goals at various levels. The program's key objective of reducing delinquent behaviors through improving family functioning was well‐aligned with the Ministry's overarching mission (Ministry of Social and Family Development, 2016b). In addition, the strong evidence base for FFT's effectiveness resonated with CFPS’ vision of establishing a reputation as a leading provider of evidence‐based psychological services in Asia. Importantly, FFT addressed the gap in family based interventions for youth offenders on community probation. This was crucial, given the strong association between family functioning and youth reoffending in the local context (Chng et al., 2016; Chu et al., 2015).

Availability of professional expertise and IT infrastructure motivated the adoption of FFT. All CFPS psychologists held Masters‐ or Doctorate‐level qualifications, and were therefore suitably qualified to take on the role of a FFT clinician. Hence, there was no need to commit additional resources toward the recruitment and hiring of new staff. Furthermore, the use of IT systems for the storage, entry, and management of client data was already part of routine case management within MSF. This was important, because the use of an electronic client data management system was a key component of the FFT implementation protocol. The match between the organization's available resources and program's resource demands gave administrators confidence that FFT could be implemented readily.

Program cost‐effectiveness was an important concern, especially in anticipation of slowing regional economic growth (The Business Times, 2015). Given the tight fiscal climate, providing sound justification for the expenditure of public funds on the new program was essential. Economic evaluations of FFT provide strong evidence of the program's ability to save government dollars, in the form of averted crime costs. Cost‐benefit analysis (CBA) studies of FFT in the United States have reported estimated cost savings ranging from US$7.50 to US$28.81 per dollar spent on FFT (Aos et al., 2011; Jones, Bumbarger, Greenberg, Greenwood, & Kyler, 2008). Moreover, FFT has been found to be more cost‐effective than other interventions targeted at reducing youth recidivism (Aos et al., 2011). Thus, it was hoped that the FFT would lead to long‐term reductions in fiscal spending on youth crime.

Another key consideration in the local adoption of FFT was the program's potential to be transported and adapted for use in a markedly different culture. To date, no study has explored FFT's generalizability or effectiveness in a country predominantly influenced by Asian culture. However, there is some support of the program's utility in culturally diverse minorities. An examination of FFT completion rates in a nationwide sample of over 4,400 families in the US found no differences between Hispanic, African American, and Caucasian families (Alexander, Frietag, Hollimon, Turner, & Robbins, 2008). Also, there is evidence of FFT's effectiveness in reducing criminal behavior among non‐Caucasian youth (Darnell & Schuler, 2015). Collectively, these findings suggest that FFT could be both culturally acceptable and effective in Singapore. There was also a concern that cultural adaptation may involve the amendment of core program elements, which could compromise model fidelity (Breuk et al., 2006). However, an independent study found therapist cultural sensitivity to be a primary indicator of sound FFT delivery (Cohen, Doran, Hernandez, & Snowden, 2008).

Prior to the decision to adopt FFT, a team of psychologists embarked on a study trip to FFT LLC and several FFT sites. This trip allowed the team to learn more about the FFT model and its implementation protocol. Direct consultations with the developers were instrumental in anticipating and addressing potential site‐specific challenges before the actual roll‐out. Discussions with other FFT sites reaffirmed the view that adopting FFT would benefit youth offenders and families under MSF's care, and that implementation was feasible in the context of MSF's existing infrastructure and resources.

Implementation Activities

Functional Family Therapies implementation protocol is closely aligned with internationally established implementation frameworks, such as that developed by the National Implementation Research Network (NIRN; Bertram et al., 2015; Fixsen et al., 2005). The NIRN framework, as summarized in Figure 1, recognizes seven implementation drivers—also known as “core implementation components”—considered to be essential toward achieving and maintaining high‐fidelity program delivery (Fixsen et al., 2009). These drivers—namely: selection, training, coaching, performance assessment, decision support data systems, facilitative administration, and systems level intervention—were identified through a comprehensive review which identified elements that were commonly present in successful implementation efforts (Fixsen et al., 2005). Implementation drivers are hypothesized to have an interactive influence on each other; strengths of a particular driver may offset deficiencies in another driver, and vice versa (Fixsen et al., 2009). The following sections discuss the activities undertaken in relation to each implementation driver.

Figure 1.

Implementation Drivers. Reprinted from “Improving programs and outcomes: implementation frameworks and organization change” by R. M. Bertram, K. A. Blasé, and D. L. Fixsen, 2015, Developmental Psychology, 43, p. 479. Reprinted with permission.

Selection

Functional Family Therapy therapists were recruited among (a) CFPS psychologists and (b) social workers who were already collaborating with MSF in other interventions. Therapist selection was focused on criteria that were important for high‐fidelity delivery, but which could not be taught through training. Openness toward learning a new model, and a willingness to adhere to the new program were important selection criteria. Furthermore, practitioner beliefs that youth behavior could be changed through working with their families, relentlessness in achieving goals, and comfort with irregular working hours were also taken into account during recruitment. Such qualities were assessed based on therapists’ attitude and performance in their clinical work and other projects, as well as feedback from their peers, clinical supervisors, and managers.

The rationale for giving weight to these qualities stems from findings that treatment providers’ beliefs, attitudes, competence, and willingness to adhere to treatment models could be central determinants of successful implementation (Damschroder et al., 2009; Hill, Maucione, & Hood, 2007). Aside from their role as treatment providers, therapists also function as advocates who manage and direct the dissemination of EBPs in their organization (Southam‐Gerow, Rodríguez, Chorpita, & Daleiden, 2012). Thus, compatibility between practitioner beliefs and the therapeutic philosophies of the FFT model was paramount.

While the above criteria did not guarantee selection of the most suitable therapists, they were effective in ruling out individuals whose beliefs and preferences were incompatible with the FFT model and its demands (e.g., lack of appreciation of the value of family work, or unwillingness to work outside of regular office hours). Therapists who demonstrated willingness toward embracing the FFT model and a high level of motivation toward coping with the demands of delivering FFT were identified. The staff selection process culminated in the formation of a team of seven therapists—four psychologists and three social workers—out of a pool of approximately twenty social service practitioners.

Training

The newly assembled team underwent a two‐day introductory training conducted by a certified FFT trainer. Training consisted of an assortment of learning materials and modalities. Print materials and visual presentations provided information on the core theoretical concepts, principles, and key therapeutic techniques underlying the FFT clinical model. Videos of real‐life therapy situations were used to provide a clearer demonstration of the effective application of these skills. Experiential learning components allowed therapists to practice newly acquired skills in a safe environment, and the trainer to provide instant feedback in relation to difficult in‐session situations that may be encountered.

During the initial transition from individual‐focused Cognitive Behavioral Therapy (CBT) to family‐focused FFT, the trainer observed that therapists demonstrated a tendency to fall back on familiar treatment approaches (e.g., supplementing family sessions with individual sessions). To ease the transition, commonalities shared by both approaches were highlighted by the trainer. This helped therapists to use their grounding in CBT to better comprehend and appreciate the FFT model. First, CBT‐focused programs run by CFPS do acknowledge the role of the family on delinquent behavior (e.g., having parent sessions as part of group programs). This was consistent with the core principles of FFT, which identifies family roles, relationships, and interactional patterns as central to adolescents’ problem behaviors (Alexander et al., 2013). Second, skills taught in the Behavior Change phase of FFT such as positive communication, emotional regulation, and problem‐solving skills, were comparable with those employed in CBT‐based programs. These techniques were used by therapists to address both family and individual risk factors. Through appreciation of the common ground between the two treatment modalities, therapists were able to grasp the FFT model quickly.

Coaching

Although the skills required to deliver FFT are imparted during introductory training, regular coaching provides new therapists with a platform for obtaining advice on how to handle specific challenges encountered in practice. This builds their capabilities in delivering the program effectively over a wide range of real‐world scenarios (Schoenwald, Sheidow, & Letourneau, 2004). In this way, coaching ensures that high‐fidelity program delivery is the norm, rather than the exception (Fixsen et al., 2009).

The FFT protocol detailed a systematic coaching structure. This consisted of four additional coaching sessions after the introductory training in the first year, and one follow‐up session in the subsequent 2 years. There was also weekly Skype supervision with a US‐based consultant in the first year, and face‐to‐face supervision with a local supervisor from the second year onwards. While the additional sessions focused on strengthening clinical skills, the weekly sessions served as a platform for reviewing the progress of all cases handled by the team. The local supervisor also had fortnightly Skype supervision with the consultant on issues relating to the effective conducting of supervision sessions.

An initial challenge faced was the limited range of methods from which the overseas consultant could monitor therapist behaviors, owing to the constraints of long‐distance supervision. Nonetheless, the team found creative ways around this problem by using other methods such as audio recording, session transcripts, and role‐plays. These afforded the consultant several sources of information from which to appraise therapists’ capabilities, thereby allowing them to provide appropriate and constructive feedback during coaching.

Performance assessment

Monitoring adherence to, and competence in, program delivery is important for providing therapists with timely and targeted feedback on how they can further develop their proficiencies (Schoenwald et al., 2004). Information on therapist performance also helps administrators to decide whether changes to the operating environment are needed to better support program delivery (Fixsen & Blase, 1993).

In the present context, program fidelity was assessed using multiple sources of information. These included session notes completed by therapists, case review and session plans during group supervision, as well as additional audio recordings of some sessions. The team supervisor reviewed therapists’ performance periodically and provided a rating of overall therapist fidelity—known as the Global Therapist Rating—based on two criteria. The first criterion was dissemination adherence—the degree to which a therapist followed the core components outlined in the program protocol. The second was fidelity, which consisted of two facets: model adherence and practitioner competence. Therapists were evaluated based on (a) how closely they kept within prescribed guidelines of the implementation protocol (e.g., documentation, timelines for responding), (b) adherence to the FFT model during treatment, and (c) the degree to which they were capable of adapting intervention techniques to match the progress of in‐session interactions.

Decision support data systems

Information relating to staff competence and program characteristics are supplied by means of feedback structures known as decision support data systems (Fixsen et al., 2009). These systems are usually established prior to actual implementation, with the goal of providing prompt feedback on the quality of program implementation once operations commence (Bertram et al., 2015).

Client‐specific program information and therapy session notes were entered and stored in an electronic database—known as the Client Services System (CSS)—managed by FFT LLC. CSS data were routinely extracted and compiled every 4 months, in the form of a report known as the Tri‐Yearly Performance Evaluation (TYPE). Besides therapist fidelity, TYPE reports provided feedback on other process indicators, such as treatment dosage and duration, therapist utilization rates, note completion rates, and program completion rates.

The team utilized information from TYPE reports to aid decision‐making at both program and organizational levels. For instance, the site consultant and supervisor used feedback on therapist fidelity to identify areas of professional development that could be addressed during subsequent coaching sessions. These data would also be useful in helping administrators devise strategies to support the program's long‐term sustainability.

Facilitative administration

Administrators play a pivotal role in creating an organizational environment supportive toward program implementation. They achieve this by ensuring that practitioners are equipped with the resources they need to deliver the intervention competently (Sheldon et al., 2004). Where necessary, administrators may be required to alter work‐related processes that might interfere with program delivery.

Changes to therapists’ working conditions were an inevitable consequence of the adoption and installation of FFT. Interventions for youth offenders conducted by CFPS were usually held in the office, during conventional working hours. To adhere to the core components of the FFT model, therapists had to make significant work‐related adjustments. For example, more time had to be spent on work‐related travel because therapy was intended to be home‐based. Also, as families are usually available only after working hours (e.g., evenings, early mornings, and weekends), therapists would need to make compromises on their own work‐life balance. Failure to support therapists in coping with these stressors could adversely affect their well‐being, and consequently, their delivery of FFT.

To address these challenges, several changes were introduced in CFPS to increase the flexibility of working arrangements for all FFT therapists. “Time‐off” in‐lieu could be claimed as compensation for time spent working outside of regular office hours. Whenever possible, therapists were assigned cases near their homes, so as to reduce work‐related commute time. In addition, therapists were given the option to work from home if they had sessions scheduled with clients living near their residence. Finally, departmental events were organized in a way that accommodated therapists’ schedules. Ensuring the team's presence at these events was a way of reaffirming the importance of their work, and to avoid perceptions of workplace marginalization. Taken together, these organizational changes helped to ease the team's adjustment to the demands of running FFT.

Systems interventions

Systemic adjustments are key to the successful implementation of EBPs. Not only they are instrumental toward the conduciveness of the working environment for program delivery, but they also influence staff attitudes and behaviors toward the newly introduced EBP (Aarons & Sawitzky, 2006). These interventions typically involve collaborations with systems outside the program's immediate operating environment, with the aim of mobilizing monetary, professional, and organizational resources required for the installation and smooth operation of the new program (Klingner, Ahwee, Pilonieta, & Menendez, 2003). Such activities are also important toward demonstrating the organization's belief in, and support for, the new program (Mihalic & Irwin, 2003).

System‐level activities were vital throughout the implementation process. In particular, securing the buy‐in of external stakeholders—especially probation officers—was important for ensuring a steady stream of referrals. To this end, stakeholders across various agencies within MSF (e.g., probation officers, policy administrators, and researchers) were introduced to the clinical model through a variety of platforms, such as meetings and conferences. These provided stakeholders with opportunities to learn about the FFT model, and how the clients under their care might benefit from the program.

Changes at the systems level were also essential for the integration of FFT into existing work processes. Besides convincing stakeholders of the merits of FFT, referral procedures had to be discussed with administrators in the probation service. After a consensus on referral framework was reached, the team disseminated core information on FFT, eligibility criteria, and referral workflows to probation officers. These efforts fostered smooth collaboration between probation officers and therapists, well before FFT was rolled out. This minimized time spent on providing clarifications relating to the referral criteria and processes, allowing therapists to focus on effective program delivery.

Implications for Clinicians

While all drivers were important for successful implementation of a new model, the current implementation process highlighted three key messages for FFT clinicians. First, ongoing consultation and service monitoring is critical for maintaining the sustainability of the new model and should be strictly followed. This does not only improve program fidelity, but also helps clinicians gain expertise that boosts their sense of mastery, competency, and self‐efficacy. Failing that, the implementation of the new model may be perceived as just another change in administrative demands (Aarons, Sommerfeld, Hecht, Silovsky, & Chaffin, 2009).

Second, securing buy‐in from clinicians and stakeholders should not only be done at the initial implementation phase, but also be incorporated into the clinician's everyday practice. Clinicians are encouraged to provide timely feedback to the implementation team so that obstacles to their work can be proactively addressed. Although such obstacles (e.g., caseload, administrative hassles that diminish productivity) may be easily dismissed by clinicians, these are common barriers that could contribute to job stress and turnover. In addition, clinicians are also encouraged to actively collaborate with partners (e.g., caseworkers) to increase their understanding and receptiveness toward the new model.

Third, while adjustment and modification to the implementation process may be required when adopting a new clinical model, it is helpful to minimize such changes at the early phases of implementation. When changes are necessary, it is always advisable to seek inputs from the model developers first. In other words, it is crucial to focus on doing things right before doing things efficiently.

Preliminary Implementation Outcomes and Challenges

Evaluating implementation outcomes is important for several reasons. Before drawing causal attributions about a program's impact on its target population, it is first necessary to demonstrate that the program was implemented well (Breitenstein et al., 2010). If desired client outcomes are not observed, data on implementation outcomes can help evaluators determine whether the absence of expected program impact might be due to implementation failure—poor execution of program processes and activities—or theory failure—errors in the hypothesized association between program activities and client outcomes (Rogers & Weiss, 2007). Moreover, implementation outcomes serve as proxies of the quality of implementation (Proctor et al., 2011). Data on these indicators can therefore help to identify areas where subsequent implementation efforts can seek to improve upon.

Early Implementation Findings

Figure 2 displays mean dissemination adherence and fidelity ratings for the team over the first 18 months of implementation. Scoring of both ratings ranged from a minimum of 0 (low) to a maximum of 6 (high). The FFT developers had set a benchmark rating of 4 out of 6 for dissemination adherence, and 3 of 6 for fidelity. On the whole, the local FFT team displayed high levels of protocol adherence and competence in their delivery of FFT. Average dissemination ratings surpassed standards prescribed by the program developers, right from the first performance assessment. The steady increase in adherence scores reflects the team's consistent effort in familiarizing themselves with, and keeping to, protocol guidelines. Examples of guidelines include key program characteristics such as intervals between sessions, responsiveness toward clients, completion of measures used in tracking treatment progress, and conscientiousness in documentation of therapy progress. Adherence in these areas is critical in keeping with FFT's main treatment principles: such as swift and effective engagement, brief but intensive treatment, and using timely feedback to guide therapy progress (Alexander, 2011).

Figure 2.

Mean dissemination adherence and fidelity ratings of the Singapore FFT team in the first year of implementation.

Team fidelity ratings exceeded benchmarks at all time‐points. This indicated that the team was proficient in applying the FFT model. In general, therapists were reasonably successful in applying therapeutic techniques toward addressing the unique needs of each family. There was a slight decrease in fidelity ratings between the second and third evaluation reports, which could be due to the change in rater between these reports. Based on protocol, the local site supervisor was to assume responsibility for conducting performance assessment ratings around the end of the first year (third report). Caution by the site supervisor may have resulted in more conservative fidelity ratings.

Administrative data, collected as part of a larger evaluation project, provided complementary evidence of the success of early implementation efforts. Table 1 summarizes preliminary local data on entry rates, dropout rates, and treatment duration, together with process data from California, United States (California Institute for Behavioral Health Solutions, 2015), and New Zealand (Heywood & Fergusson, 2016) for comparison. Entry rates were slightly lower than in California. Among probationers who did not start treatment, failure to start was due to (a) the lack of an available therapist fluent in the family's mother tongue, or (b) the occurrence of another offence shortly after referral, resulting in institutional placement. These, however, were due to operational and systemic factors rather than therapists’ failure to engage the families. Over the first 18 months of implementation, a total of 31 probationers and their families (49 caregivers, 12 siblings) commenced and completed FFT. Dropout rates were comparable to California and New Zealand, suggesting that therapists were competent in engaging and motivating families to remain in treatment. Finally, mean number of sessions per family was also consistent with other sites, and within the prescribed range of 12–14 sessions (Alexander et al., 2013).

Table 1.

Program Characteristics of FFT in Singapore, Compared With Sites in USA and New Zealand

	Singapore	USA (California)	New Zealand
Entry rate (%)	87.3	93.6	Not reported
Dropout rate (%)	6.25	36.1	13.0
Mean total no. of sessions	10.6	14.1	10.4

Limitations

Taken together, preliminary data suggest that the initial implementation of FFT in Singapore has been relatively successful. Nevertheless, these findings should be carefully considered in light of three limitations. First, the data reported were collected over a relatively brief period of 18 months. More longitudinal data would be necessary for drawing conclusions about the sustainability of current implementation efforts. Second, it remains to be seen whether new challenges may emerge as implementation moves into full operational phase, and if so, what their impact on existing implementation strategies might be. Finally, it is unknown whether good implementation translates to positive client outcomes. However, at present, there is insufficient data on client outcomes. Future studies will incorporate such data in discussing implementation success, as recommended by Alexander (2011).

Cultural Challenges

Thus far, studies that have examined the implementation of FFT across diverse cultural and ethnic populations have been conducted in countries with a predominantly Western cultural influence. In the course of implementation in Singapore, two challenges characteristic of working with Asian cultures—namely, language barriers and emotion suppression—were commonly encountered when working with local families.

Language difficulties

Difficulties with communication regularly emerged as an issue of concern. This was similar to the experiences of non‐English speaking FFT sites (Breuk et al., 2006; Thøgersen, 2012). While youth probationers generally had little difficulty conversing in English, some caregivers (e.g., parents and grandparents) could only communicate in their mother tongue. In anticipation of this, therapists fluent in a variety of mother tongue languages were recruited, so as to increase the team's linguistic diversity. To facilitate accurate communication during sessions, therapists were assigned to families based on language compatibility. Nonetheless, there were several instances where a family member could speak only dialect (e.g. Hokkien, Teochew, or Cantonese) or a foreign language (e.g., Thai). In such cases, therapists would seek help from other family members with translation. As a result, language barriers were sometimes used as opportunities for developing rapport and a sense of mutual respect between the therapist and family members.

The loss of linguistic nuances during translation was another challenge posed by the language barrier. Even if therapists could be matched with the family's language preference, the meanings of certain words might be altered or lost when expressed in another language. For instance, one common theme hint used in the engagement/motivation phase was “pain”, or “emotional pain”. However, the Mandarin translation sounded stilted and unnatural in the local culture. To tackle this problem, therapists relied on group supervision and extensive preparation to refine phrasing before sessions. While these methods were not a perfect solution, they helped generate alternative terms or statements to elicit the same themes.

Emotion suppression

Functional Family Therapy was developed in an individualistic culture, where outward displays of emotions are encouraged to promote autonomy and a unique personal identity (Safdar et al., 2009). In contrast, maintenance of group cohesion and emotion regulation are highly valued by collectivistic cultures (Potter, 1988). Such cultures encourage emotion suppression to prevent conflict and preserve relationships (Butler, Lee, & Gross, 2007). Thus, Asian families have a higher tendency to dismiss strong negative emotions such as anger, shame, or contempt, despite having similar negative attributions and blaming behaviors as their Western counterparts (Safdar et al., 2009).

The first phase of FFT focuses on reducing blame and negativity among family members while increasing their hope for change. After helping family members identify problem behaviors, therapists are required to generate positive reframing for these behaviors for other family members. However, in the local context, therapists observed that many families were reluctant to disclose negativity during these initial sessions. Denial or downplaying of problems and negative emotions (e.g., “it's normal”; “all families have some arguments now and then”) was common. To circumvent this, local therapists had to first “elicit” these issues via sequencing (i.e., using a series of questions to examine repeated patterns of behavior that occur within the family; Alexander et al., 2013). Therapists were also encouraged to be more “respectfully presumptive” with family members. This was achieved by expressing feelings on behalf of clients, while simultaneously checking for client receptiveness through observing their verbal responses and nonverbal cues (e.g., facial expression or body language). These culture‐sensitive techniques helped families to be more emotionally connected and intrinsically motivated to change. They also helped therapists accurately assess family relations and youths’ problematic behavioral patterns, which were necessary for developing targeted intervention strategies for the next phase.

Notwithstanding the differences mentioned above, it is noteworthy that the main therapeutic needs of families eligible for FFT appear to be culturally universal. Through constant dialogues with FFT consultants, as well as the site supervisor's externship experiences with American families, it was apparent that the emotions and relational themes underlying family struggles in other FFT sites were also common in Singaporean families.

Conclusion

Effective implementation is critical for intervention effectiveness. Yet, research on program implementation is scant in the evidence‐based literature. While a few studies have described the implementation of FFT, these have been conducted in countries that are predominantly influenced by Western culture. The present study is the first to detail the implementation of FFT in an Eastern culture. Factors that facilitated program adoption were similar to those reported in earlier studies. Early implementation data suggest that the FFT model can be transported to a culture very different from the one in which it was developed, yet be delivered with high fidelity. Proactive organizational leadership, close partnership with implementation purveyors, and strong therapist commitment were crucial in surmounting the challenges encountered during first 18 months of the program's implementation.

This study extends the small but emerging body of evidence on the cross‐cultural utility of the FFT model. It also contributes to the wider evidence‐based literature, being one of the first to document the implementation of a Western‐developed EBP in an Eastern culture. By describing specific organizational efforts undertaken to facilitate the successful implementation of FFT, it provides a concrete example of the real‐life application of an established conceptual framework. Finally, this paper highlights several issues that could surface when transporting EBPs to culturally different environments. Future research should endeavor to evaluate implementation success by examining a wider range of process indicators, and exploring the link between implementation and client outcomes.

Biographies

Daniel Z. Q. Gan, MSocSci;

Yiwei Zhou, MPsych(Clinical);

Eric Hoo, DPsych(Clinical);

Dominic Chong, MAppliedPsych(Clinical);

Chi Meng Chu, DPsych(Clinical), Clinical and Forensic Psychology Service, Ministry of Social and Family Development.

Gan, D. Z. Q. , Zhou, Y. , Hoo, E. , Chong, D. , Chu, C. M. (2018). The Implementation of Functional Family Therapy (FFT) as an Intervention for Youth Probationers in Singapore. Journal of Marital and Family Therapy, 45, 684–698. 10.1111/jmft.12353