Chronic fatigue syndrome (CFS) is characterised by severe and medically unexplained fatigue that lasts longer than 6 months and results in severe limitations in daily life (Fukuda et al., 1994; Reeves et al., 2003). Cognitive behaviour therapy (CBT) for CFS can effectively reduce fatigue and impairment (Castell, Kazantzis, & Moss-Morris, 2011; Chambers, Bagnall, Hempel, & Forbes, 2006; Price, Mitchell, Tidy, & Hunot, 2008). Even full recovery, defined as no longer being severely fatigued, being able to resume all activities, and having a perception of health and fatigue that is similar to healthy persons, is possible for a subgroup of patients (Knoop, Bleijenberg, Gielissen, van der Meer, & White, 2007). Several models underlying CBT for CFS assume that there are certain fatigue-related beliefs and behaviour that perpetuate symptoms (Fry & Martin, 1996; Gaab, 2004; Surawy, Hackmann, Hawton, & Sharpe, 1995; Vercoulen et al., 1998; Wessely, David, Butler, & Chalder, 1989). One of these models is that of Vercoulen et al. (1998). It is the only model that was statistically tested and it states that a high level of somatic attributions and focusing on symptoms and a low level of physical activity and low perceived control over symptoms lead to perpetuation of fatigue.
Recently, Wiborg, Knoop, Prins & Bleijenberg (2011); Wiborg, Knoop, Frank & Bleijenberg (2012); Wiborg, Knoop, stulemeijer et al. (2010) used mediation analyses of previous randomised controlled trials (RCTs) to test whether 1) CBT indeed leads to a change in these fatigue-perpetuating factors and 2) whether this change was related to the reduction in fatigue. Their studies showed that changes in focusing on symptoms and sense of control over fatigue indeed mediate the change in fatigue during CBT for CFS (Wiborg, Knoop, et al., 2012; Wiborg, et al. 2011). However, it is not the actual change in physical activity, but the perceived increase in activity that mediates the decrease in fatigue (Wiborg, Knoop, et al., 2012; Wiborg, Knoop, Stulemeijer, Prins, & Bleijenberg, 2010). Heins et al. furthered their work by studying how these fatigue-perpetuating factors change during CBT for CFS. Their study showed that these variables change in the same period in which fatigue decreases. Changes in fatigue-perpetuating factors explained up to half of the variance in fatigue during CBT for CFS (Heins et al. submitted for publication). So although these studies provided insight into the changes that lead to a decrease in fatigue, these changes could only partly explain the treatment outcome of CBT for CFS.
Besides specific treatment techniques, extant research suggests that the therapy relationship, i.e. “the feelings and attitudes that therapist and client have toward one another and how these are expressed” (Norcross, 2001), is also an important contributor to treatment outcome in a variety of psychotherapies. A recent task force on evidence-based therapy relationships even concluded that “the therapy relationship accounts for why clients improve (or fail to improve) at least as much as the particular treatment method” (Norcross & Wampold, 2011). Wampold and Budge recently proposed a model of the therapy relationship, which consists of the bond between patient and therapist, setting of goals and tasks and expectations with respect to treatment outcome (Wampold & Budge, 2012). This model thus includes outcome expectations and three elements of the working alliance as proposed by Bordin: task, goal and bond (Bordin, 1979). Tasks are what therapist and patient agree that needs to be done to reach the patient's goals. Goals are what the patient hopes to gain from therapy. Bond is the trust and confidence in the therapist.
Although both outcome expectations and working alliance are considered empirically supported relational attributes across studies of various therapies (Constantino, Arnkoff, Glass, Ametrano, & Smith, 2011; Fluckiger, Del Re, Wampold, Symonds, & Horvath, 2012; Horvath, Del Re, Fluckiger, & Symonds, 2011), their application is developed and maintained in different ways within different therapeutic models and protocols (Hatcher & Barends, 2006; Ulvenes et al., 2012; Wampold & Budge, 2012). We’ll therefore briefly discuss the application of outcome expectations and working alliance in CBT in general and CBT for CFS specifically.
In the protocol used for this study, recovery is considered the highest achievable outcome of CBT. Patients, however, may be hesitant to believe that full recovery is possible and may only expect to learn to cope with their chronic fatigue. Both the aetiology of CFS and its treatment are hotly debated (Action for ME/Association of Young People with ME, 2008; Kean, 2010; Working group on CFS/ME (Chronic Fatigue Syndrome/Myalgic Encephalomyelitis), 2002). Many CFS patients attribute their symptoms to a physical cause, such as a viral infection or immunological disorder (Deale, Chalder, & Wessely, 1998). Patients with strong somatic attributions will, understandably, not immediately have faith in a cognitive behavioural therapist and the rationale of CBT. Indeed, the placebo response, which is strongly related to outcome expectations (Di Blasi, Harkness, Ernst, Georgiou, & Kleijnen, 2001), is lower in psychological interventions for CFS than in somatic interventions (Cho, Hotopf, & Wessely, 2005). Furthermore, in a recent large RCT only 57% of the CFS patients allocated to CBT indicated that they were confident about the treatment (White et al., 2011). The fact that patients are often referred for CBT after they have had symptoms for several years (see for example the study of Prins et al. (2001) with a median symptom duration of 4.9 years) may further lower outcome expectations.
The working alliance in CBT is mainly reflected in collaborative empiricism, i.e. patient and therapist collaborating to uncover negative automatic thoughts through Socratic reasoning and homework assignments (Dattilio & Hanna, 2012). Although in the theory underlying CBT it is assumed that a good working alliance in itself is insufficient to bring about change (Clark, Beck, & Alford, 1999), a good alliance is considered necessary and has been related to positive treatment outcome (Horvath et al., 2011). The working alliance in CBT for CFS may not always be good as some CFS patients will have a different view than the CBT therapist about the cause of CFS (e.g. a somatic versus a cognitive behavioural model of CFS) and the goal of treatment (e.g. recovery versus coping with a chronic illness). The first objective of this paper is to determine whether outcome expectations and working alliance contribute to the reduction in fatigue during CBT for CFS. A second objective is to determine whether they have a direct effect on treatment outcome, independently of changes in fatigue-perpetuating factors or exert their effect (also) indirectly through changes in fatigue-perpetuating factors.