The dissemination and implementation of psychological treatments: Problems and solutions
Treatment researchers expend their efforts identifying effective treatments, and for whom and how they work, but there are matters over and above these that are of concern when it comes to dissemination and implementation. These include the clinical range of the interventions concerned, the ease with which they can be learned, and their mode of delivery. It is these three topics, as they apply to the psychological treatment of eating disorders, that form the focus of this article. Alongside these considerations, we discuss how modern technology has the potential to transform both treatment and training. © 2013 by Wiley Periodicals, Inc. (Int J Eat Disord 2013; 46:516–521)
There has been an assumption among treatment researchers that evidence-based treatments will simply be absorbed into routine clinical practice, a passive process sometimes termed “diffusion.”1 This view is mistaken. If we take the example of the eating disorders, a specific form of cognitive behavior therapy (CBT) has been the leading evidence-based treatment for bulimia nervosa for at least two decades, yet there is good evidence that it is not widely practiced.2, 3 Similar findings exist for other mental health problems and, indeed, across the whole of medicine. It is for this reason there is a burgeoning interest in how to improve the dissemination and implementation of evidence-based treatments. In general, dissemination refers to the transmission of information about a treatment (e.g., its nature, effectiveness, indications, and characteristics), whereas implementation refers to its adoption and use in clinical settings.1
In this article, we highlight three topics of relevance to both dissemination and implementation. They are the clinical range of a treatment; the ease with which it can be learned; and its mode of delivery. The focus here is on how they apply to the psychological treatment of eating disorders. Alongside these considerations, we discuss how modern technology has the potential to transform both treatment and training.
Clinical range refers to the breadth of the indications for a given treatment. Some interventions are suitable for more patients than others. Even within a specific disorder, certain treatments work with a wider range of patients than other equally effective interventions. In short, some treatments have fewer moderators of outcome than others. What is needed are effective treatments that can be used with the broadest possible range of patients.
Within the eating disorders, the leading evidence-based treatment for both bulimia nervosa and binge eating disorder is CBT.4, 5 However, these two eating disorders (as defined in DSM-IV) together only account for about 40%–50% of the eating disorder cases seen in most clinical settings.6 If there were an evidence-based treatment that could be used with a broader spectrum of patients than this there would be a strong case for its dissemination.
“Transdiagnostic” treatments are of particular interest in this context. In this regard, the eating disorder field is ahead of most others. Ten years ago a transdiagnostic view on the processes that maintain eating disorder psychopathology was proposed.7 It was argued that these processes are likely to be largely the same across the eating disorders and that therefore, if they can be disrupted in bulimia nervosa, it should be possible to disrupt them in the other eating disorders. Hence, CBT for bulimia nervosa was reconceptualized as a treatment for the eating disorder psychopathology seen in bulimia nervosa and, on this basis, it was suggested that this treatment should work across the eating disorders, if suitably adapted. The result was the development of a new cognitive behavioral treatment, so-called “enhanced CBT” (CBT-E), designed to treat all forms of eating disorder.8
Subsequent research indicates that CBT-E can indeed be used across the eating disorders. Three case series of patients with anorexia nervosa, two composed of adults and one of adolescents, all large for this comparatively rare disorder (total N = 145), indicate that CBT-E has a strong and lasting effect in those who complete the treatment.9, 10 As regards the remaining eating disorder cases, those with bulimia nervosa or eating disorder not otherwise specified (eating disorder NOS), CBT-E has been shown in two separate studies (total N = 219) to have substantial effects, again well-maintained over 60-week closed follow-up periods, that appear to be superior to those obtained with the earlier version of the treatment.11 Together, the findings of these studies indicate that CBT-E is an evidence-based treatment for all forms of eating disorder seen in adults. Moreover, in common with a related CBT protocol,12 it can be used with adolescent patients as well.9 Therefore, with respect to its clinical range and effectiveness, CBT-E is ideal for dissemination and implementation. It overcomes the problem of therapists having to learn multiple different treatments, one for each eating disorder.13
What is the case for disseminating other evidence-based psychological treatments for the eating disorders? In contrast with CBT-E, none has been demonstrated to be fully “transdiagnostic” in scope. In part, this is because they have been designed to be disorder-specific, and in part it is a consequence of the relative absence of transdiagnostic treatment research, which, in turn, reflects the conservative policies of many funding agencies.
Interpersonal psychotherapy (IPT) is a candidate for dissemination and implementation as it seems comparable in effectiveness to CBT when treating BED.5, 14 Its clinical range appears to be more limited, however. It has not been tested as a treatment for eating disorder NOS; it is slower to work than CBT in bulimia nervosa15, 16; and it is of uncertain value in the treatment of anorexia nervosa.17, 18 Other potential candidates are the various alternative psychological treatments for adults with anorexia nervosa, but as yet none has good evidence to support it and all are anorexia nervosa-specific and so have a narrow clinical range. Turning to adolescents, a specific form of family therapy, so-called family-based treatment,19 has good evidence to support it as a treatment for anorexia nervosa and there is evidence that it can be used in bulimia nervosa.20, 21 Its status as a treatment for eating disorder NOS is not clear.
Overall, there is a clear case for disseminating and implementing family-based treatment for adolescents with anorexia nervosa or bulimia nervosa, and CBT-E for adults with any form of eating disorder.
Ease of Learning
One of the disadvantages of psychological treatments is that many are difficult to implement well and this certainly applies to treatments such as CBT-E and family-based treatment, and probably to many other treatments as well. Therefore, a factor that has a bearing on the dissemination of psychological treatments is the ease with which they can be learned. Treatments that require intensive training from expert therapists are difficult to disseminate because of the lack of qualified trainers.22, 23
The topic of therapist training is a neglected one. A two-stage approach is generally advocated, assuming that the therapist already has the relevant background experience.24–26 In the first stage the treatment is described (usually in the context of a “workshop” from an expert, typically lasting between 90 minutes and 2 days), and, in the second, the trainee employs the treatment under the supervision of someone who is proficient at it. The supervisor also ensures that the trainee becomes familiar with any written guide or “manual.” All three components are thought to be essential as the clinical and research consensus is that workshops and treatment manuals on their own are inadequate in the absence of supervision.27
These considerations apply to the eating disorder field as much as to any other, yet all too often they are neglected or sidestepped. Even in controlled treatment trials there are examples of there being little or no attempt to ensure that the therapists were trained and supervised by professionals competent, let alone expert, in the treatments concerned. Under such circumstances it is impossible to have confidence in the findings. If training and supervision are less than satisfactory in research settings, it is likely that there are even greater shortcomings in routine clinical practice.
The daunting task is how to train all those therapists who require training—not just those who work in a particular organization, state or country, but worldwide. The current method of training is not suited to this task. Workshops from experts are costly and difficult to arrange, and few of those who attend have the benefit of subsequent case supervision, either because there is no one available to provide it or because it is too expensive. New approaches to training are therefore required, ones that are far more scalable than the current method.
One option that has been promoted is the “train-the-trainer” model. This involves one staff member of a mental health organization being trained in the treatment concerned and then this staff member becoming a trainer of the remaining members of staff.27 This strategy has been shown to be effective in training therapists to deliver CBT for panic disorder,28 and there is also evidence that it can be used to train therapists to use guided cognitive behavioral self-help for binge eating problems.29 The strategy remains to be evaluated in larger controlled studies. However, even if it proves effective, it is not suited to the training of therapists on a regional or country-wide scale. It would take far too long. Instead, a more radical rethinking of psychological treatment training is required.
It has recently been argued that a new form of training, so-called “Internet-enhanced training,” might be both scalable and effective.30 This would involve trainees having access to a specially designed, clinically rich, training website that would provide detailed but accessible information on the treatment's indications and use, as well as its strategies and procedures. Thus, it would duplicate and extend the information provided in face-to-face training workshops.
Websites of this type are currently under development at the Centre for Research on Dissemination in Oxford (CREDO), one for CBT-E and another for behavioral activation for depression.31 Each website will include an extensive “library” of acted illustrations of the treatment, something that is impossible to provide in conventional workshops. In addition, they will include formative tests of knowledge, including tests of clinical problem-solving, together with guidance for further learning. Trainees in the treatment concerned will have several weeks to work their way through the core material on the website including both the factual material and the main clinical demonstrations. Once this has been done, trainees can either receive conventional case supervision, assuming it is available or, if engaging in Internet-enhanced training, they can move on to receive a modified form of case supervision in which the supervisor makes extensive and repeated reference to the website and, in particular, to the library of acted clinical illustrations. This type of supervisor would not need to be an expert in the treatment concerned: rather, he or she would need to be good at providing “web-centered supervision.” This requires that the supervisor has good generic supervisory skills, be familiar with the disorder being treated and the general class of psychotherapy in question, and have intimate knowledge of the website. Thus, the supervisor's role would differ from that in conventional supervision. Instead of providing expert guidance themselves, supervisors would help trainees solve problems and enhance their practice by referring to the information, guidance, and illustrations on the website. Their role would be somewhat equivalent to that of a “facilitator” (as against that of a therapist) in guided self-help.32 The supervision could also be remote rather than face-to-face, and provided via Skype or the equivalent. Indeed, supervision centers could serve entire regions or countries. CREDO is in the process of setting up such a center to serve Ireland.
Internet-enhanced training should be less expensive and more scalable than current methods as there is no need for an initial workshop and the supervisor does not need to be a specialist in the treatment concerned. Whether Internet-enhanced training will prove to be as effective as conventional training needs to be determined. It is possible that it might be more effective, given the access to the clinically rich website throughout the period of supervision.
Internet-enhanced training might also address another matter of concern, that of “sustainability.” This term refers to the extent to which an intervention (in this case training) has a lasting effect. It would be possible for trainees to return to the website at intervals to engage in courses designed to counter so-called “therapist drift,”33 and this could be combined with the completion of validated, Internet-based, measures of applied knowledge of the treatment concerned.30 Internet-based refresher courses of this type could become a scalable part of continuing professional education.
Mode of Treatment Delivery
How psychological treatments are delivered also has a major bearing on their suitability for dissemination. Treatments that are lengthy and require a great deal of professional input and expertise are difficult to disseminate, whereas those that can be delivered with limited external input and by therapists with minimal training are more scalable and potentially as effective. Less can be more, and if no external input is needed, this is the ideal.
It has recently been noted that face-to-face psychotherapy sessions, the “dominant model of treatment delivery” in the United States, is a form of treatment that is inherently not scalable.34 It is too labor intensive. This criticism applies not only to long-term psychotherapy but also to treatments such as CBT and IPT that are shorter and time-limited. Innovative ways of delivering treatment are needed. There are a number of interrelated options.34 There is the “task-shifting” of therapy to other health care professionals, ones who are less expensive to employ (see Patel and colleagues35); there is the modification of treatments to render them wholly or partly self-administered; and there is the use of technology to deliver treatment, be it via the phone, electronic media, or the Internet. And there can be combinations of all three.
With regard to task-shifting, it is the authors' experience that therapists from a wide range of backgrounds can implement CBT-E to a high standard so long as they have been provided with appropriate training. CBT-E is certainly not the exclusive province of clinical psychologists or psychiatrists, and the same is likely to be true of many other psychological treatments.
It is in the area of self-help that there have been particularly important developments and especially so in the eating disorder field. Perhaps the best example is the development of guided cognitive behavioral self-help.36 This is a brief adaptation of full-scale CBT for those eating disorders characterized by binge eating. The intervention typically consists of eight to ten 20-minute sessions in which the “therapist” or provider oversees the patient's implementation of a self-help program (usually provided in the form of a self-help book). Therapists from a variety of backgrounds can deliver the intervention as it requires more generic skills than disorder-specific ones.37 The active ingredient of the approach is the self-help program: the therapist's job is to help the patient make the best possible use of it. There is mounting evidence that guided cognitive behavioral self-help delivered by nonexperts can be as effective as full-scale treatment delivered by highly trained therapists, at least in the treatment of binge eating disorder.37 Even if it were somewhat less effective, it would have many advantages in view of its greater scalability.
Pure self-help has been less well studied. Its importance lies in the fact that sufferers do not need to approach a professional to obtain help. In this way it circumvents a number of barriers to treatment include stigma, poor case detection by professionals, and scarcity of resources and expertise. Unfortunately, at least with regard to the eating disorders, pure self-help is unlikely to be fully transdiagnostic in its scope. This is because people with certain eating disorders, most notably anorexia nervosa, do not view themselves as having a “problem,” and therefore are unlikely to engage in self-help. Pure self-help is also not suitable for those who are “at risk” psychologically or physically.
The ideal intervention from the perspective of scalability and ease of access is pure self-help delivered via the Internet. This has not been tested. The existing web-based programs have been designed to be used with external support.38–40 These programs show promise but ones that are suitable for solo use also need to be developed and evaluated.
Developing an intervention that is scalable is not a goal in its own right: the intervention also needs to be effective. Over the past 30 years much has been learned about eating disorder psychopathology and the processes that maintain it. This information needs to inform the strategies and procedures utilized by any intervention, be it face-to face therapy or a self-help program. Overcoming Binge Eating,41 the best tested self-help program for binge eating problems, was relatively unsophisticated in this respect. It used tried and tested procedures to curb binge eating and address two of the maintaining processes—rigid dietary restraint and the tendency to binge in response to day-to-day difficulties—but it did not address perhaps the most potent set of maintaining processes, those that stem from the overevaluation of shape and weight (viz., body checking, body avoidance, and feeling fat). It is therefore not surprising that this program, and those like it, has been found to be more effective in binge eating disorder than bulimia nervosa,37 where the overevaluation of shape and weight is a prominent feature. New self-help programs are needed that are capable of addressing the full range of the user's eating disorder psychopathology, including the body image disturbance and problems with emotion regulation. Ideally such programs should be capable of being personalized: indeed, this will be essential as otherwise they will be unwieldy for the user and will include many redundant elements.
Many questions need to be answered regarding self-help interventions and Internet-based ones in particular. These include their effectiveness when presented in written form or as an interactive web-based program, the value of external guidance, and whether the guidance can be provide remotely. To address such questions rigorous research is needed.42 Any tendency to employ minimal methodological rigor to evaluate minimal programs must be resisted.43 These interventions need to be evaluated to the same standard as any other treatment. Until such research has been conducted, it will be impossible to calibrate these programs against the effects of face-to-face treatments and to distinguish one program from another.
CGF holds a Principal Research Fellowship from the Wellcome Trust (046386). The Centre for Research on Dissemination at Oxford (CREDO) is supported by a strategic award from the Wellcome Trust (094585). Details of the research may be found at credo-oxford.com.