Why and how should family therapists take part in family therapy research?
- Side: 332-344
- Publisert på Idunn: 2015-12-11
- Publisert: 2015-12-11
From the 1990s, the position of family therapy has been challenged by a culture inclined to define psychological problems as individual illnesses connected to specific diagnostic labels. Furthermore, a narrow definition of research, often called the gold standard designs, has influenced the psychotherapy field. To advocate a systemic perspective, we need clinical research with a broader scope. This article argues for systemic research that is designed closer to the clinical situation. Besides information about the context, there are three important sources of information about the process and outcome of clinical practice: Systematic and regular feedback from clients and therapists about the process and outcome, direct observations from therapy sessions, and reports about clients’ long-term development. If used alone, each of these gateways to knowledge may be very misleading. However, if used together they can help us to refine and develop the collaboration with families and couples. Furthermore, this approach can help us to improve our ability to communicate more meaningfully about the ideas for, and the outcome of, systemic interventions.Keywords: Systemic therapy; clinical research; process and outcome studies
Introduction1Revised manuscript after a presentation at the 10th Nordic Family Therapy Congress in Turku, August 2014.
Last year I heard Laurie Hetherington of Williams College, USA, speaking enthusiastically about systemic research at the 45th International Annual Meeting for the Society of Psychotherapy Research (Copenhagen, June 2014). Referring to more than 150 studies in couple- and family-therapy research, she concluded that systemically oriented therapy works. A chapter titled «The effectiveness of couple and family-based clinical interventions» in the last edition of the Handbook of Psychotherapy and Behavior Change supports the same message (Sexton, Datchi, Evans, LaFoliette, & Wright, 2013). Several therapeutic approaches appear on the American Psychological Association’s (APA) lists of empirically-supported and evidence-based treatments. «We no longer have to justify ourselves as family therapists», Hetherington said. In addition, she continued to point out that we have a large amount of specific knowledge about how to engage families in treatment, about reframing and relabeling the effects of psychological problems into a relational language, and about enactment in therapy – that is, the power of doing. Furthermore, the tradition has developed good tools relevant for both clients and researchers to monitor change, such as STICS® (Pinsof et al., 2009), alliance scales (Knobloch-Fedders, Pinsof & Mann, 2004), observational coding (Heyman, 2001) and feedback systems such as Client-Directed Outcome-Informed Treatment (CDOI; ORS and SRS) (Anker, Duncan & Sparks, 2009; Miller, 2010).
In spite of these promising conclusions, from the beginning of the 1990s until the present there have been many signs suggesting that family oriented and systemic thinking has lost the position as a preferred approach to psychological problems. In Norway, there have been multiple indications of this development. For instance, quite recently the Norwegian Health Department wanted to exclude family therapy as a relevant specialist education for psychologists in the health services. For a while, The Norwegian Psychological Association even supported the idea. Furthermore, the relative amount of psychologists with special competence in systemic thinking has declined, both among practitioners and in the scientific staff at the universities in Norway.
As several other colleagues within the Nordic countries, I have seen signals coming from the health authorities of a more and more individualized view in the understanding of psychological problems. The attention is directed to something within the person, whether this within is called psychiatric illness, cognitive problems, attachment and mentalization deficiencies or trauma, and these explanations have been stretched further on to neurological defects and genes. Ekeland (2001, 2012) once labeled it «the bio-medical architecture» for mental problems. Nielsen and Hertz (2013) have described how the practice of clinical exploration for diagnostic purposes can narrow, limit and restrict the therapist’s ability to be a helpful partner for the clients. Originally, a diagnosis was meant as a description of a momentary situation in life, but in practice it is seen as a trait or deficit that the person has to carry with him for the rest of his life. Thus, it easily functions as a self-fulfilling prophecy.
One reason why it has been so difficult to change this development is that the diagnostic labels are used for many other purposes than communication. For the clients they release economic support, admission to sick-leave and free medicine, access to certain treatment facilities, pedagogical support, and so on. For the State, influenced by neo-liberal ideas of New Public Management (NPM), they stimulate ambitions of efficiency control with the helping systems and to develop standards for where to turn with what problem (specialization), waiting-list guarantees, what kind of information therapists have to report (social control) and expected time-limits for diagnostic work and therapy (cost-efficiency control with «production»).
Furthermore, the labels can serve to free the individual and the relatives from guilt and responsibility attached to the problems. Recently, a near relative of a prominent Norwegian politician died. Previously, she had a heavy heroin problem. Both before and after her death the family and the press were strongly involved in a process of defining this problem as an «illness.» She was said to «give face to this illness,» as something very heroic and – implicitly – in opposition to moralizing attitudes. The reaction is easy to understand, but in my view, we need neither the label «illness» nor moralization in order to handle substance abuse.
Do systemic-oriented ideas have a chance to be heard in the present context of a medical understanding of human life problems – merged with ideas of New Public Management? Besides pointing out the unhealthy side-effects of the present systems, it is evident that we have to improve our answers to the questions about «does it work» and «how does it work». This has to be done in ways that are both acceptable for the scientific society and communicable to politicians. In recent years, several colleagues have written articles about ways to improve systemic research in this journal, for instance, Jensen (2006), Bargmann (2013) and Sundet (2015). With this article, I would like to support these initiatives. Based on projects I have been involved in, I will give examples and discuss how systemic-oriented research can contribute to influencing the discourses about the value of a context-oriented treatment, the value of inviting the individual, the family, the network and society to join forces in change processes. I will connect the discussion to some of the challenges that we face in clinical research.
Challenges regarding outcome and client feedback
There are a substantial number of studies showing that clients perceive the treatment they have received as helpful (Lambert, 2013). Even with different treatment programs, it is usual to find high satisfaction with the therapeutic process. 70–80% of the clients are content with the collaboration with the therapist and changes regarding the problems that brought them to therapy (Doherty & Simmons, 1996; Zhang, Gerstain & Friedman, 2008). One challenge here is to decide if the reported change is artificial and a response to a good relation to the therapist, while life is still in the same mess. How can we find more ways to document changes in the problems? And, furthermore, would it be relevant to include other sides of life in the evaluation of change, and how do we do that? (see also Dreier, 2008).
Research has documented that people are inclined to give more positive feedback when feedback is collected close to the end of treatment (Williams, Coyle & Healy, 1998). There have also been raised questions about what the feedback is related to, if it is an evaluation of the treatment or of something else (ibid). In one of our own studies of separating parents, some of the observations from the sessions did not match very well with the nice feedback from the couples; they could appear provoked, angry and upset and the therapist seemed helpless in the communication with them. Still, the couple evaluated the session as very helpful. Reportedly, if anyone was to blame for the lack of outcome it was the partner, not the therapist (Tjersland, 1992). We have also found that clients can tell very different stories when asked immediately after the therapy compared to what they say two years later (Gulbrandsen & Tjersland, 2013).
But an even more compelling question is whether the documented changes can be seen as an outcome of the treatment. To answer this question it is required to have both a clear description of the intervention and to show that the same intervention has been used across cases and by different therapists. Usually we do not have this kind of information. In the research literature, the interventions are mostly described in very general terms (as CBT, Psychodynamic, EFCT). From the articles, it is often difficult to know how well the practice corresponds to the theory and how the ideas of the treatment can be separated from the general alliance factors in the therapy (Goldfried & Davila, 2005; Wampold & Budge, 2012).
There has been a great deal of attention around the use of simple feedback systems such as SRS and ORS (Norwegian KOR). Research clearly points them out as promising tools in the therapeutic process (Anker, Duncan & Sparks, 2009; Bargman, 2013; Sundet, 2009). However, they are primarily therapeutic tools. What they tell us more specifically about the therapeutic process is an open question. We may be informed that something has changed, but what and how it is related to therapy can only be clarified when we talk with the clients. Furthermore, simple instruments of measurement are easily misused. Ekeland, Aurdal & Myklebust Skjelten (2013) have described how a therapeutic tool such as KOR (in the Family Guidance Offices in Norway) was turned into a control instrument for the government, and how such change of purpose for the use of an instrument can destroy its utility.
The usefulness of longitudinal studies of systemic interventions
It seems important to have more studies that follow clients for several years, and that we compare the development for similar clients who receive systemically oriented interventions with those who receive more traditional individualistically oriented approaches. For instance, such studies would allow a comparison of the development of children with an ADHD-diagnosis from a certain age receiving medical and individual treatment with a similar group of children receiving systemic interventions in the family, the school and the network (an idea mentioned by Nielsen & Hertz, 2013). Also longitudinal studies may enable a comparison of the development in a group of depressed clients in individual therapy with depressed clients receiving treatment within a family context, as performed by Jones and Asen (2000). Rather than simply comparing immediate effects we should look for descriptions of what happens over time, and as others have pointed out (for instance Dreier, 2008), the relevant factors to look for should not only be connected to how original problems are handled, but also to changes in the daily living, such as: Are the persons involved in regular education or work, or in other types of positive engagements in life; do they still use medicine for psychological problems; and how can the relationship to the family and the networks be described? In other words, we should promote studies that include a broader picture of the development after therapy. Whitaker’s (2010) critique of medical interventions in psychiatry is an example of how long-term studies can raise important questions about the prevailing practice. I have an expectation – yet to be documented – that this kind of studies will give more support to systemic interventions in the long run.
Focus on describing systemic processes, with specific client populations, and evaluate them
It is obvious that there can be vast relational and contextual differences around a child who is said to have an anxiety problem; still, it may be helpful to use the familiar individual diagnostic labels to communicate more effectively about alternative systemic ways to handle psychological problems. Systemic projects can be labeled working with families/couples with anxiety problems, abuse problems, depression problems, eating disturbances, self-harming behavior problems, and so on. Seikkula and his colleagues have been in the forefront here, describing their work with psychotic clients and their families (Seikkula, 2000, 2006).
Colleagues within the systemic fields in the Nordic countries have introduced many ideas for interventions in the work with specific groups of clients. To mention a few of the presenters at the Nordic family congress in Turku in 2014: Magnus Ringborg and Pravin Israel talked about attachment-based family therapy for youth with depression; Jon Middelborg and Dimitrij Samilow about how to talk about violence in couple sessions; Christina Ringborg and Vigdis Wie Torsteinsson about a multifamily group program for families with eating problems. Along with several others, they contributed with ideas about therapeutic practice that many find inspiring. Still, there is a need for a more solid empirical support as to the outcome of the approaches. In the last few years, there has been much research activity addressing individual therapy with both children and adults in Norway (Dittmann & Jensen, 2014; Nissen-Lie, Rønnestad, Høglend, Havik, Solbakken, Stiles & Monsen, 2015; Oddli & Rønnestad, 2012), while systemically informed projects have mostly ended up on the table for «interesting ideas». Designing empirical oriented projects to study interventions directed towards families and networks may seem an overwhelming task. Yet a couple of examples from the research I have been involved in can point out some possible directions.
A few years ago, I was involved in a project working systemically with cases where intrafamilial sexual abuse was suspected (Tjersland, Mossige, Gulbrandsen, Jensen & Reichelt, 2006; 2008; Jensen, Haavind, Gulbrandsen, Mossige, Reichelt & Tjersland, 2010). We identified symptoms among the children, attempted to establish close alliances with different parts of the family system, focused on breaking the silence and sharing information within the family, underlining the necessity to protect children – whether abuse was confirmed or not – and so on. With the help of process analyses based on video recordings from the treatment, and from the interviews with the parties 18 months later, we documented dramatic improvements concerning the children’s symptomatology and the dialogues within the families. Both the joint sessions with the children and the non-abusing family members and – when possible – the separate sessions with the suspected family members clearly contributed to this. The project raised a substantial professional and public debate about how to provide these families with the best support and help.
In a current project, we have followed 154 couples in the mediation process after divorce. Before the first session, we asked the parties about their ideas about the solutions for the children and their hope for reaching an agreement. We audiotaped the sessions, and collected feedback from the couples and the therapist after each session, based on a few questions. Telephone interviews followed 18 months later. The project has shown that with mandatory mediation, the agencies spend the bulk of their resources talking with separating couples who evidently could manage the situation themselves, leaving fewer resources available for work with the parents in high conflict. Most high-conflict couples terminated mediation early and without any agreement (Gulbrandsen & Tjersland, 2013). The analyses of the sessions with the high conflict cases contributed to identify several specific sources that seemed to enlarge the relational conflicts (Gulbrandsen, 2013; Kjøs, Madsen & Tjersland, 2015). Furthermore, we have identified mediation strategies that both contributed to stopping the process and helping the parties to continue their dialogues (Gulbrandsen, 2015; Kjøs, Oddli & Tjersland, 2015). Thus, the project has political implications relevant for a discussion about how we design services for separating parents, as well as practical implications for therapists working with conflict.
My point is that we need to initiate and improve the systemic-oriented research connected to different client groups in society. This may strengthen the arguments to trust interventions based on systemic thinking. Furthermore, it certainly can contribute to developing ideas about relevant paths to follow in order to establish a constructive collaboration with certain client-groups and their families. I follow Sundelin (2013), who describes systemic therapeutic manuals as more comparable to practical maps than to suffocating straitjackets. There has been a tendency within the systemic field to highlight the idea of a free co-constructive process between the therapist and the family, regardless of the problems the client and the family bring to the session, and regardless of the cultural expectancies to which the therapists are exposed. To me, these ideas seem rather naïve. The client is not always right, neither about explanations nor about solutions. I think it is important for a therapist to have ideas about directions to follow with different groups of clients, and that it is on these grounds that it is possible to be flexible and attentive to the client and the family, because the therapist knows more precisely what he or she brings into the process of co-construction. To me, this way of reasoning is quite in line with Torsteinsson’s (2014) reflections upon «finding wisdom in knowledge» at the recent Nordic family therapy congress.
Fourth, I will advocate more studies of processes with couples bringing specific problems to therapy
Compared to outcome studies of individual therapy, the studies of couple therapy have been few, and when performed, they have often been developed to support a certain general approach to couples, for instance, cognitive-behavioral therapy (CBT) (Fals-Stewart & Lam, 2008), emotion couple therapy (EFCT) (Halchuk, Makinen, & Johnson, 2010), or narrative-oriented therapy (Seikkula, Aaltonen, Kalla, Saarinend & Tolvanene, 2013). Studies that focus on how such approaches are practiced can contribute to develop couple therapy, but I do not think that efforts to compare outcomes will bring much new knowledge to the field. As already known from individual therapy research, comparing outcome from different traditions is likely to give very similar results. Today, the couple-therapy field is clearly turning towards integration of traditions (Lebow & Chambers, 2012; Gurman, 2013). In my opinion, what is needed is more empirical knowledge about how to enter a fruitful collaboration with couples bringing different types of problems to therapy, for instance, infidelity, loss of a child, conflict with in-law parents, sexual problems, violence, and so on. I think it would be helpful for both therapist and clients to put into practice specific ideas that have already been described within systemic literature in these areas, and to evaluate them.
Fifth, I would propose more observational studies of systemic practice
That is, to focus more on studies of the dialogues and the interaction that takes place between family members and between them and the therapists. In my view, this is the best way to develop ourselves as therapeutic «instruments». There is still a large gap between theoretical concepts of systemic therapy, say, for instance, reframing, co-construction, externalization, solution-, emotion- or narrative-orientation and how this is expressed, conveyed and brought to life in the sessions with the clients.
For several years, I have been especially engaged in the study of dialogues in therapy sessions. I think it belongs very naturally to the systemic tradition and its readiness to open up the therapy room for colleagues, and for observational and reflecting processes. At the same time, when we label this «research,» we have to be very attentive as to how therapists understand the project. Research means a sort of evaluation; therefore, it is not possible to avoid moments when the therapist taking part in an observational study may feel clumsy and unprofessional, or to be a failure as a therapist. Therefore, these studies require an open, safe and continuous dialogue between the researcher and the therapists. Indeed, failures are very valuable empirical data.
One advantage of observational studies is that we can get closer to the practice of the therapist and to the observable feedback coming from the clients. We can learn more about helpful ways for therapists to relate to different situations. For instance, as already mentioned, how to handle the suspicions of sexual abuse in families (Jensen et al., 2008; Tjersland et al. 2006) or to conflicts between separating parents (Gulbrandsen, 2014; Kjøs, Tjersland & Roen, 2014). Furthermore, observational studies enables a closer study of what clients brings to the therapeutic project at the beginning of therapy. Together with colleagues, I have completed such studies of group therapy with children who had been exposed to violence (Tjersland, Østvold Lindheim & Gudmudson, 2012). In another study, we transcribed the dialogues between therapist and children, when using a computer-based program to talk about matters of family loyalty (Egaas Bøhren, Stabrun, & Tjersland, 2014). Among other things, these studies highlighted how the context for the conversation is explained to the children and how we can invite them to share ideas about sensitive subjects in therapeutic processes.
To use Steve de Shazers’ (1985) metaphors, many of us who work with families do not always receive customers for therapy. Rather often the clients appear more as visitors, sent by others. Moreover, if the therapist is not wary of the ambivalent expectations woven into the language that the client brings to the session, it easily becomes a drop out case. Men who have been violent to their partners are often described as externalizing and declining responsibility for their behavior (Daly & Pelowski, 2000). Presently, I am engaged in a study of such men who have been referred to individual therapy. When we listened to the first sessions with these expectancies in mind, it was easy to find similar answers. However, when we were directing a magnifying glass to the nuances of what the men said in the first sessions, we found more details, and among these, there were invitations to enter into change projects together with the therapist (Lømø, Haavind & Tjersland, 2015).
My point is that such studies can contribute to more detailed descriptions of phenomena present in dialogues, and to develop a more differentiated language about expectancies for therapy: Thus helping the therapist to become more sensitive about what to look for in the co-creation of a therapeutic project.
I suggest that we need more longitudinal studies comparing systemically oriented work with other approaches directed towards specific groups of clients. Moreover, in this context we need to offer clear descriptions of the systemic paths that have been followed in work with families striving with specific problems.
We need more observational studies of couple and family therapy, preferably extended by open interviews with clients during and after the process, and to use the established research instruments more as supportive devices to the qualitative studies.
We need these types of empirical support, not only to improve our work as therapists, but also in order to bring attention to the importance of seeing psychological problems in the context of «the important others» and the dominating cultural expectancies.
|1||Revised manuscript after a presentation at the 10th Nordic Family Therapy Congress in Turku, August 2014.|