Introduction

In psychotherapy generally and in couple, marital, and family therapy specifically, numerous studies and meta-analyses have pointed to the critical role of the therapist in change processes; the therapist is a central figure in positive (and negative) therapeutic change (Blow et al. 2007; Wampold 2001). This statement implies that evidence based practice is not solely comprised of well validated, theoretically coherent treatments, found in lifeless books/manuals. Rather, effective practice occurs through competent clinicians who draw from the best evidence available as well as lean on their clinical wisdom. A growing body of literature indicates that there is no doubt that effective treatment flows through wise, skilled, and flexible clinicians (Wampold 2001; Wampold and Brown 2005; Wampold and Imel 2015). These clinicians are able to engage client systems, navigate the therapy process, and work with openness to feedback from their clients about progress. Ideally, these clinicians deliver treatments informed by the best available evidence, and which fit well with the cultural and contextual components of the clients with whom they work (APA Presidential Task Force on Evidence-Based Practice 2006). These clinicians require therapeutic wisdom and elasticity, knowing when to shift direction and try something new when therapy is not going well. In short, effective therapists are able to work in evidence informed ways, adhering to evidence-based treatment approaches where indicated, and departing from these approaches when suggested by the client system, treatment progress, or the therapeutic context.

This becomes more complicated as layers are added to the therapy process. When it comes to working with complex cases and systems, such as those often faced by couple, marital, and family therapists, increased clinician competencies, talent, and in some (but not all) cases, experience, are required to achieve success (Blow et al. 2007). We argue that the role of the therapist in working with these complex systems is even more crucial than in individual-only work. There are likely personal characteristics and qualities that are more beneficial for couple, marital, and family therapists to possess as compared to more individually focused practitioners. For example, they need to be directive and assertive in order to break up dysfunctional couple and family cycles, an essential quality in effective family therapy practice. Individual therapists, on the other hand, may do better by remaining more “hands off” as they deal with only one client at a time. In spite of the complexities, only a small body of literature addresses the required skills and competencies of therapists working with these complex systems. This paper will discuss the centrality of the therapist in the effective delivery of interventions delivered to couples and families and implications of this central role for training, practice, and policy.

The Key and Crucial Role of the Therapist in Delivering Couple, Marital, and Family Therapy Interventions

Wampold suggests that who the therapist is and how he/she acts in therapy comprises a large part of therapeutic change variance (Wampold 2001). His review of the state of outcomes in psychotherapy as a whole convincingly argues that the therapist plays a central and key role in positive and effective therapy (Wampold 2001; Wampold and Brown 2005; Wampold and Imel 2015). He estimates that therapist effects account for between 3 and 7 percent of treatment outcomes. These are large estimates when compared to other therapy related change ingredients. Other scholars support these estimates of importance. For example, Lutz et al. (2007), using a real world data set, concluded that there are substantial effects attributable to therapists with approximately 8 % of the total variance attributed to the therapists, and 17 % of the variance in rates of improvement attributed to therapists. In another study by Saxon and Barkham (2012), therapist effectiveness was found to be similar (6.6 %), and even greater when it came to more difficult cases. Baldwin and Imel (2013) argue that the differences among therapists in their effectiveness can substantially affect the mental health of the public as a whole, and as a result, more focused attention is needed on therapists and what makes one therapist more effective than another. In addition, they urge researchers to consider the issue of therapist variability and effectiveness with different client types. This is especially important for our contention that those who work with couples and families require a different set of skills and talents.

In addition to the positive impact of therapists, when therapy does not go well, the therapist shares some blame (Wampold and Imel 2015). While Wampold and Imel (2015) describe therapy generally and not marriage and family therapy specifically, we contend that greater skills are needed when delivering interventions to couples and families. This is because in working with these cases, the therapist manages multiple (often feuding) members of a family/couple in the therapy room at one time, at different ages/stages of development, while at the same time keeping alliances intact with these individuals, people who have varied views of the presenting problem and its underlying causes. These therapists are required to work within these larger systems and deliver the best evidence available to treat the presenting issue. Therapy can easily get off track or mired down in irrelevant content. These therapists require skill and flexibility and need to be able to rapidly deliver interventions, assess the responses from members of the system, balance the system if needed, all while under a moderate to high amount of emotional pressure. Family therapists also are required to manage a great deal of information in the midst of family conflict, and intervene with an appropriate intervention at the right time, targeting the family member/s most likely to effect change, all while at the same time controlling his/her own reactivity (e.g., own values, own family experiences) to the family problem presented.

While there is a growing body of research indicating the effectiveness of family-based interventions for a wide range of presenting issues (Sprenkle 2012) including adolescent conduct disorder (Baldwin et al. 2012), alcohol abuse in a couple context (O’Farrell and Clements 2012), affective disorders (Beach and Whisman 2012), and child behavior problems (Kaslow et al. 2012), to name a few, there is a glaring lack of evidence about what makes a therapist effective in working with these systems, separate from the treatment used (Blow et al. 2007). This is in spite of the notion that the couple, marital, and family therapist is intertwined with change processes, and is an essential figure, deciding when to intervene and facilitating change mechanisms as they play out in the therapy (Blow et al. 2007).

Even though there is a long list of effective family and marital/couple interventions, little is known about the qualities, skills, personality styles, training, or other pertinent variables of the therapists who deliver these interventions (Blow et al. 2007). There are surprisingly few studies of what makes one family therapist more effective than another, even though it is intuitive that there are differences in outcomes between these therapists in the same way that we would expect different coaches of basketball or football teams to have skills and qualities that differentiate them (and the results of their teams) from one another.

While the therapist appears to be central and critical to change in therapy involving couples and families, this core role of the therapist is not acknowledged in research or policies about best treatments involving couples and families, and often not in reimbursement for services delivered. These conclusions lead us to several important recommendations for training, dissemination of treatments (transportability), how we conceptualize knowledge of what works in therapy (policies for funders), and reimbursement for effective services. While we target our comments to our area—couple, marital, and family therapy—we believe that these considerations apply across the board to all helping professionals in the psychotherapy arena.

Evidence-Based Therapists

If indeed it is true that the therapist is a crucial part of the change process, what does this mean for therapist selection into the field? What considerations do agencies need to take into account when they hire a new therapist? “How should therapist development and evaluation be conducted?” We believe that the central importance of the therapist (as opposed to treatment manuals) has vital implication for who therapists are and how they learn and improve as therapists. A focus on therapists, instead of manuals, has large implications for the transportation of evidence based treatments into community settings. We believe that the time is overdue for the psychotherapy field as a whole to research and develop the idea of evidence based therapists (Blow et al. 2007). Studies are beginning to emerge on this topic and one recent study explored how clients are likely to benefit more or less depending on the therapist they are working with, and that clinical outcomes are improved when clients work with therapists who are a good fit for their specific condition and strengths (Kraus et al. in press).

Evidence-based therapists are those providers who consistently are able to achieve positive results with their clients. It is of note that therapists can get trained, supervised, licensed, and reimbursed with little-to-no evaluation of their actual work in the therapy room (Baldwin and Imel 2013). We advocate for therapist outcomes with clients to be a core metric in determining if a novice therapist (or any therapist for that matter) is a good fit for the practice of therapy over time. Therapists need to be held accountable by the outcomes of their clients (Baldwin and Imel 2013). Baldwin and Imel capture this sentiment when they state: “Provided they do not violate ethical mandates and follow the law, mental health professionals are left alone to practice as they see fit” (p. 258). In other words, there are few mechanisms that hold therapists accountable to effective work. While all therapists have clients who experience little to no change or even deteriorate, we contend that excellent therapists reflect a preponderance of excellent outcomes across the majority of their clients. For therapist accountability to occur in reality, data need to be consistently collected, independent of the therapist, on all clinical cases, and databases will need to aggregate and report these results in fair ways. Clients should have a way to provide this input without a breach of their confidentiality. These data would ideally be published online and subject to the scrutiny of the consumer. Therapists not only should publicize their client outcome data overall, but also for specific client types/presenting problems. If a therapist works with couples for example, he/she should publish couple therapy outcomes, across all cases, in aggregate form. Training programs should collect data, not only on student academic performance or supervisor subjective evaluations, but also on client performance, i.e. how are therapists doing in the practical field of their study. Evidence based therapists ideally have data connected to their work that provides an indication of reliability of the therapist’s work across cases. Therapists who receive consistently bad scores should be terminated from training programs or counseled into different careers. Public availability of data will also have a “free market” effect on therapist credibility and earning potential by extension. Clients would have complete transparency about the success rates of the particular therapist they selected to see. In addition, therapist scrutiny of outcomes would make the therapist more aware of his/her effectiveness and increase accountability for improvement. Therapists who received consistently low scores would be motivated to improve or find a different profession. We agree with Baldwin and Imel (2013) who suggest that the time has come for more serious discussion about what needs to happen with therapists who are not meeting a standard of effectiveness across cases. While it is suggested that these therapists could receive more training or counseling, some may also need to be steered completely away from the helping profession.

Training Effective Marital, Couple, and Family Therapists: Skill Based versus Model Based, Finding a Balance

A focus on therapists as central to change, places the emphasis where it needs to be—on therapist skill and talent as opposed to theories written in books or manuals. It is not to say that these theories are not important; however, they are worthless if not delivered by someone with the required skillset. This would be the same for any activity. For example, an athlete could look like a soccer player, study the manual of soccer, and receive good coaching, but still fall far short of been effective on the soccer field. In this case, a soccer player would be relegated to the bench or ultimately kicked off the team. There are certainly soccer players who can improve considerably with coaching and study of soccer skills manuals, but the results should be demonstrated in the game. In the same way, therapists who work with complex systems, need basic talent, and they also need exposure to a wide array of evidence based theories and other humanistic knowledge. There is some evidence that excellent therapists are voracious learners who devour knowledge, indicating that they do not only possess talent, but they are also open to learn a lot and continue to learn (Skovholt et al. 2004). But this knowledge needs to be applied in the therapy room in a way that leads to improved outcomes.

In the case of therapist induction into the field, training programs choose therapists based on a number of factors, but traditionally, therapists are brought into the field with consideration of their undergraduate academic scores, GRE scores, letters of recommendation, notable life experiences, and personality, as a few examples. Some programs interview their applicants while others do not. No matter how one looks at it, therapist selection into the field is often comparable to speed dating. Of note, it is notoriously difficult to counsel therapist trainees out of programs once they have enrolled. Usually, they have to be considerably incompetent before programs have reason for dismissal. In the same way, once a therapist is hired at an agency, they are not easily let go if they are ineffective with clients. In training programs, students study theories in books, and usually undergo some type of supervised practical training experience. However, as far as we know, it is rare for therapist training competencies to be assessed by systematically evaluating their success with clients. As stated earlier, we advocate for much stricter levels of accountability in therapist training. In addition, we encourage our field to study the differences between therapists, more than they have already been studied, and for there to be a clearer delineation between what therapist effectiveness skills can be taught versus what skills happen to be innate. Some therapists are born with strong interpersonal skills, including verbal fluency, warmth, acceptance, and empathy, essential skills for good therapy. However, these may not translate into effective therapy for a number of reasons, perhaps if they are unable to apply their knowledge at the right time or in the right way. Successful family therapists are influential, persuasive, and convincing; they provide an acceptable and adaptive explanation for the client’s distress and frame the problem skillfully for all the members of the system to buy into. Our questions are “what happens to therapists who consistently are unable to operate at a minimal level of effectiveness with their clients?” How do training programs assess these competencies and what happens when these therapists fall short in terms of helping their clientele?

The question about what to do with incompetent therapists is up for debate. The logical answer is to provide them with more training. However, as Beutler et al. (2004) point out in their extensive review of therapist qualities, this research is mixed and the relationship between training and improved outcomes is not clear. It might be the case that therapists who work with couples and families can achieve better outcomes through training (and we would expect this to be the case in some cases with added supervision or additional reading); however, we would also expect that it would be almost impossible for some therapists to work effectively with these populations. Even though they may be ineffective with couples and families, this does not mean that they cannot be effective with individual clients. The installation of feedback mechanisms into this work, and increased supervision may help therapists form more effective alliances, but more research is clearly needed.

Training of new therapists in the marriage and family therapy field is important given the premise that the therapist plays a large role in change (Karam et al. 2015). Ideally therapist training would include to a large degree the skills needed to be an effective therapist. This focus would include specific immersion in building and sustaining alliances with couples and families, understanding change mechanisms, and training therapists to continually monitor client progress in an authentic way using feedback and outcome informed technology. In addition, therapists should have training in evidence based models (not just one model), current research on human development, and the best research evidence related to particular presenting problems and client characteristics (Blow et al. 2007). In short, therapists should be immersed in as many theoretical and evidence informed knowledge bases as is feasible. While we do not advocate for the unreasonable idea that therapists are adept at every model, it would be a good thing if therapists had more than a cursory understanding of these approaches, and a deeper understanding of the change mechanisms behind these approaches.

There are important reasons for increasing the emphasis on therapist factors in training programs. First, learning about the skills and qualities that produce successful therapists is important in the same way that a beginning athlete needs to know about factors that lead to success. Focusing on core, straightforward strategies to build on inherent personal therapeutic abilities may reduce anxiety for students struggling with mastering the complexity of a specific model or a challenging family system. If young therapists are confident in their ability to build upon existing, innate personal qualities, then self-confidence issues around youthful appearance or lack of professional experience may be minimized.

Second, it is unrealistic to believe that the therapists will stay with one pure model throughout the duration of their careers, especially as they work with a wide range of clientele and presenting problems. As they grow professionally, so also will their confidence grow in expressing their unique therapeutic strengths. Many of these qualities are not expressly mentioned or prioritized in manualized treatment or pure therapy models. The majority of seasoned therapists do not utilize only one or two approaches in their work, a notion supported by comprehensive studies which conclude that clinicians generally do not professionally practice this way (Northey 2002; Orlinsky and Rønnestad 2005). Therapists-in-training, especially, must experiment with ideas and techniques on their own so that they can discover for themselves what works and makes sense, given their own particular personality and caseload. Further, some therapists pick approaches for the wrong reasons, often having little to do with aspects that are most helpful to clients. These reasons are related to their worldviews, spiritual leanings, or own life issues, but may miss the unique characteristics and needs of their clients and their contexts (Blow et al. 2007; Simon 2006).

Successful therapists share a number of qualities and personality traits. While many of these therapist level variables may be innate (e.g., warmth, empathy, compassion) others can be practiced and enhanced throughout time in training programs. We recommend the following general therapist characteristics be focused on in training programs in addition to standard training content. Adaptability/flexibility, projecting hope and confidence, patience and pacing in clinical work, self-reflection and insight, curiosity, humor, and instilling in therapists a sensitivity to diverse cultures and contexts.

Transportation of Evidence Based Models to Communities

Transportation of evidence-based approaches to community settings has proved to be a difficult problem. It is critical that this process be improved upon if clients in communities are going to have access to the best evidence available in their treatments. We contend that an evidence based therapist focus (away from models) does a great deal to help with the issue of dissemination of effective treatments. Part of the problem of dissemination is that therapists are not quick to value evidence based approaches, for many reasons. They may not want to adopt an approach that seems to clash with their current favored method. In addition, they may be underpaid or burnt out, and adoption of a new approach may simply seem like too much work. Or, some may feel as if their agency is imposing a top down approach on the way they practice. A focus on therapist outcomes would go a long way to resolve this issue.

For example, if an agency or county was considering bringing in an evidence based approach, they could consider the following strategy. Therapists who are already obtaining exceptional outcomes would not need to adopt and adhere to a new treatment given that they are already successful. These therapists would be encouraged to learn the new approach to add to their already stellar repertoire, but it would not be mandatory given that they are already highly productive. However, therapists who are underperforming can view adoption and adherence to an evidence based model as a career saver. Adoption of a new model would give him/her the chance to retool and improve outcomes with clients. Finally, the chance to improve outcomes would be incentive and motivation for therapists to try something new. They may not have this motivation in current contexts if a new approach is seen to add to their burden. A fidelity focus would also not need to be the highest priority if therapists were already attaining consistently good outcomes. An outcome focus allows for therapist idiosyncratic styles (they could be encouraged to modify approaches to their own styles as long as they were obtaining good outcomes). Currently, many therapists view their work as good, with no consistent outcomes to back up these claims.

It is important to note that therapists need to be delivering the best available treatments available with the best evidence. As Whiteford et al. argue (Harris et al. 2015; Whiteford et al. 2014), there need to be strategies to increase the use of evidence based treatments, and that the emphasis should not only be on individuals receiving treatment, but also the correct type of treatment for the specific problem, the appropriate dosage of treatment, and treatments with the best available evidence. We go one step further and suggest that the best evidence be delivered through clinicians who are competent and effective.

What We Study (Policy for Funders): A consideration for How Therapy Works

A focus on therapists (as opposed to therapies) should change what funders prioritize as the focus of studies. Traditionally, funders want investigators to come up with interventions or approaches that can be tested in randomized controlled trials, with an emphasis on manualization, therapist fidelity to the manual, and adherence to the approach. These interventions are then widely disseminated through trainings or online. There is one big problem with this approach: therapists who are effective are left out of this focus of study, and therapist variance is controlled for instead of explored (Wampold and Imel 2015). A shift in focus to funding studies of therapists such as the tracking of therapist outcomes, therapist’s skills and competencies, and therapist training would lead to important results. These may lead to a different emphasis on the famous quote by Paul (1967); instead of saying “What treatment, by whom, is most effective for this individual with that specific problem, and under which set of circumstances?” a shift would occur to “Which therapist, with what level of expertise, is able to relate to and help, this specific client, from this specific background, culture, or context, and with this specific presenting problem?” The shift is a subtle but an important point of emphasis. The recent study by Krause et al. seems to support this view (Kraus et al. in press). Funders have not privileged this type of research in spite of the strong evidence that therapists account more for variance in therapy than do models (Wampold and Imel 2015). Why is this the case? We posit that the medical model has dominated mental health research (a focus on specific ingredients), and it is time for a different emphasis in the study of what works in therapy. In addition, qualitative and process oriented studies would also shed more light on the role of the therapist in change. As an intermediary step in this culture change, more studies of differential therapist efficacy could be embedded in current research designs that are more fundable in today’s funding climate. Progress Research (Process + Outcome), first proposed by Pinsof and Wynne (2000) would also result in studies that directly impact the moves of the therapist in therapy. We need to know much more about what differentiates effective and non-effective therapists from each other (Baldwin and Imel 2013), and therapists who are more effective than others with specific populations such as couples and families.

Reimbursement for Therapy Services

There is no doubt that reimbursement for therapy services is based on politics more than results. While in the United States, there are five Federally recognized core mental health professions (marriage and family therapists, psychiatrists, clinical psychologists, social workers and psychiatric nurse specialist), there is not parity in the amount of reimbursement given to each practicing specialty (Health Resources and Services Administration 2016). Rather, reimbursement is based upon lobbying activities of mental health professions, and at times the whims of insurance companies. The 2014 implementation of the Affordable Care Act in the United States will lead to the increased utilization of mental health services. These required treatment protocols have steadily moved to evidence-based treatments (EBTs). Reimbursement for EBTs is largely based on reimbursing therapists who have been trained in these approaches and who claim to be using them. This recent shift for mental health care treatment to be evidence-based is gaining momentum is illustrated by the state of Oregon, which has mandated 75 % of treatment be EBT (Lewis and Simons 2011). That means that potentially therapists can be reimbursed at a higher rate if they are trained in a model that is considered evidence based, no matter the quality of the training, the rate of adoption, or the effectiveness of the therapist. In short, training in evidence based practice or belonging to a specific mental health discipline (having a certain license) leads to higher reimbursement, regardless of outcome.

However, if mental health reimbursement was linked to client outcomes, a large shift would occur in the delivery of mental health services. More therapists would seek to improve their services. Clients would have more transparent options. A member of any mental health discipline can claim to have good training and do good therapeutic work, whether he/she is a social worker, psychologist, psychiatrist, or marriage and family therapist. These claims can currently be made in the face of no evidence to support these contentions. It is time for a shift to occur.

We acknowledge that there are numerous controversies involved in attaching any kind of reimbursement to client reported satisfaction and effectiveness. However, when compared to how therapists are currently reimbursed, we believe a time has come for more rigorous debate about how to reimburse for the quality of care. We question whether any type of accountability system would be effective if it did not have financial implications for those who practice. One suggestion is that those who reimburse for services should reimburse at the highest rate, clinicians who consistently get good results, and not reimburse at the same rate those who see clients long term with minimal results. As Baldwin and Imel (2013) suggest, the time has arrived to consistently monitor therapists outcomes over time in order to see which therapists are effective and which are not. This would require a change that would ensure that clients’ outcomes were independently tracked and used to evaluate therapist effectiveness. In an Australian report, Duckett and Breadon (2014) describe ways to save costs in the medical system, particularly hospital settings. They argue that healthcare should consider paying for care that works, not just any care, and that there should be penalties in cases where healthcare makes things worse or leads to higher levels of health difficulties. One important suggestion they make is to connect reimbursements to patient reported outcomes. In their model, patients would complete questionnaires on key outcomes before and after treatment, with follow up assessments. The authors caution that these types of reimbursement systems need to be “managed through good design, careful evaluation, and adjustment” (p. 40). This is important, in that such a system could lead to unintended consequences such as gaming the system, or overinflating of effectiveness numbers. We are not convinced that patient reports are the best indicator of therapist effectiveness, especially when they are not collected independently. There are a number of problems with a system of reimbursement based upon performance. In particular, this is a challenge in that clients may feel their own pressure or subtle pressures from a therapist to inflate the scores in order to ensure reimbursement. Or, a client may give negative evaluations based upon issues separate from treatment effectiveness, such as a personality clash with the therapist. In the case of clients who present with complex presenting issues, or issues with a poor prognosis at the outset, therapists may avoid working with these clients altogether, so as to not interfere with their excellent ratings. These would be examples of the unintended consequences of reimbursement based on performance or simply a monitoring system of therapist effectiveness. Another consideration for reimbursement is that even the best therapists are not going to be effective with all clients. Clients do play a substantial part in their treatment outcomes (Tallman and Bohart 2002). For example, a couple who is more motivated to change will make greater therapy progress when compared to a couple where one or both parties are demoralized and demotivated, no matter who the therapist is who is doing the work. Some clients have a far better prognosis than do others.

As stated, there are several challenges to the implementation of an effective therapist monitoring system, especially when such a system is linked to reimbursement and credibility. These are challenging topics to consider. We suggest that more debate is needed on these topics. One thing we believe in strongly is that some monitoring of therapist accountability is required and that this is especially needed for those therapists who are ineffective with the majority of their cases. Any monitoring system should be implemented incrementally, with careful consideration, and much debate, until it provides a reliable indicator of therapist effectiveness. However, we would expect effective therapists to do well with a large number of their cases if they are monitored over time, and ineffective therapists to overall do poorly across a number of their cases. Consideration should also be given to types of clients and types of presenting problems. Some therapists are likely more effective with couples, some more effective with families, some more effective with individuals, and there are multiple possible variations on this effectiveness depending on the presenting problem and client characteristics. Saxon and Barkham (2012) suggest that careful consideration of matching clients with therapists should occur, especially when it comes to more severe cases. They also suggest monitoring of therapist outcomes and an intervention of some kind for therapists who consistently do not produce effective client outcomes over time to get them to a point where they are more effective.

Conclusion

We have argued in this paper that if the therapist is so important in change, and evidence supports this view, that this importance should be reflected in which components of psychotherapy are privileged and studied. We suggest that this is important also for therapists working with different complexities of cases such as those working with couples and families. We would hope that future research would highlight which personal therapist qualities are the most adaptive, and for which client configurations, and in what settings. We would hope that policies related to research and reimbursement would reflect these realities as well.