Moderators of the treatment response to guided self-instruction for chronic fatigue syndrome

[Dolphin]

(Not an exciting paper, IMHO)

Moderators of the treatment response to guided self-instruction for chronic fatigue syndrome.

J Psychosom Res. 2013 May;74(5):373-7. doi: 10.1016/j.jpsychores.2013.01.007. Epub 2013 Mar 5.

Tummers M1, Knoop H, van Dam A, Bleijenberg G.

Expert Centre for Chronic Fatigue, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands. m.tummers@nkcv.umcn.nl

Abstract*

OBJECTIVE:

The efficiency and efficacy of guided self-instruction for chronic fatigue syndrome (CFS) can be enhanced if it is known which patients will benefit from the intervention.

This study aimed to identify moderators of treatment response.

METHODS:

This study is a secondary analysis of two randomized controlled trials evaluating the efficacy of guided self-instruction for CFS.

A sample of 261 patients fulfilling US Center for Disease Control and Prevention criteria for CFS was randomly allocated to guided self-instruction or a wait list.

The following potential treatment moderators were selected from the literature: age, fatigue severity, level of physical functioning, pain, level of depressive symptoms, self-efficacy with respect to fatigue, somatic attributions, avoidance of activity, and focus on bodily symptoms.

Logistic and linear regression analyses were used with interaction term between treatment response and the potential moderator.

RESULTS:

Age, level of depression, and avoidance of activity moderated the response to guided self-instruction.

Patients who were young, had low levels of depressive symptoms, and who had a low tendency to avoid activity benefited more from the intervention than older patients and patients with high levels of depressive symptoms and a strong tendency to avoid activity.

CONCLUSION:

Guided self-instruction is exclusively aimed at cognitions and behaviours that perpetuate fatigue.

Patients with severe depressive symptom may need more specific interventions aimed at the reduction of depressive symptoms to profit from the intervention.

Therefore we suggest that patients with substantial depressive symptoms be directly referred to regular cognitive behaviour therapy.

*I gave each sentence its own paragraph

 

[Dolphin]

(these comments aren't that exciting but I thought I'd post them as I put myself through reading the paper)

The two sets of data they use are:

[9] Knoop H, van der Meer JW, Bleijenberg G. Guided self-instructions for people with
chronic fatigue syndrome: randomised controlled trial. Br J Psychiatry 2008;193:
340–1.
[10] Tummers M, Knoop H, van Dam A, Bleijenberg G. Implementing a minimal intervention
for chronic fatigue syndrome in a mental health centre: a randomized controlled trial. Psychol Med 2012:1–11 [Epub 2012/02/23].

This is from the Nijmegen group who could be described as huge fans of CBT for CFS. As with lots of their papers (and many other CBT papers from other countries) it would have been more interesting if objective measures had been used. Here neither the dependent variable (fatigue) nor the possible moderators are objective. Somebody could report improved fatigue after CBT and still be quite ill. With CBT (and presumably the guided self-instruction intervention),

"They are encouraged to perceive feelings of fatigue as a normal part of an active and healthy life and stop labelling themselves as a CFS patient."
(from: "The process of cognitive behaviour therapy for chronic fatigue syndrome: Which changes in perpetuating cognitions and behaviour are related to a reduction in fatigue?"

Adopting this sort of attitude could easily lower the quantity of fatigue one reports.

---
This is how "avoidance of activity" was measured:

Avoidance of activity. Five expressions of avoidance of activity were selected. The patient was asked if she or he (1) avoids symptoms by avoiding strenuous activity, (2) stops with activities in case of pain or when feeling fatigued, (3) restricts him or herself to simple activities, (4) takes a rest by sitting or lying down in case of pain or when feeling fatigued and (5) tries to take as much rest as possible to avoid symptoms. All items were scored on a 4-point Likert scale. A high sum was indicative of a high tendency to avoid activity [23,29].

 

[Dolphin]

They also investigated other potential moderators:

No significant interactions were found for the other potential moderators; fatigue severity, level of physical functioning, impact of pain, self-efficacy with respect to fatigue, somatic attributions, and focus on bodily symptoms (Table 2).

(This quote refers to liner regression, where fatigue is measured on a continuous scale.)

Similar results were found with regard to logistic regression. This involves dividing the outcome into two groups, "successful" and "unsuccessful" (my wording):

Patients were regarded significantly clinically improved with respect to fatigue if (1) the change in fatigue was statistically reliable (reliable change index >1.96) [22] and (2) the fatigue score at post-treatmentwas b35 on the CIS subscale fatigue severity. This latter score is within 2 standard deviations (S.D.) of the mean for healthy adults and below2 S.D. of the mean for CFS patients [31]. Logistic regression analyseswere performed for each potential moderator.

Similar results were found with regard to linear regression. However, physical functioning (SF-36 PF) was p=0.06 so I think that should be considered a borderline result so I don't think they should have tried to make the big distinction they did between SF-36 PF and SIP:

In a previous trial performed in a tertiary treatment centre, disabilities in different domains of functioning were associated with an unfavourable treatment response [9]. In this trial we chose to focus on disabilities in physical functioning, measured with the SF-36. The subscale physical functioning of the SF-36 measures disabilities in rather simple activities like carrying groceries, climbing stairs and walking [25]. The SIP measures disabilities in more complex activities such as work, social interactions, recreation and pastimes [24]. Physical functioning did not moderate treatment outcome. This is in contrast when disabilities are measured with the SIP. This finding implies that the SIP and the SF-36 not only measure different domains of functioning, but also that both questionnaires have a different relationship with the response to treatment, suggesting that it is important to operationalize the concept disabilities in different ways in CFS research.

Also, talking about it as "in a previous trial performed in a tertiary treatment centre" seems to me a bit weird: the current paper is looking at the data from paper 9, plus another study.

 

[Esther12]

Thanks D.

Interesting that those who were most depressed reported benefiting least from self-instruction -I could cheekily suggest that's depressive realism counter-acting response bias!

Also, those who reported avoiding activity reported benefiting least from self-instruction designed to discourage people from avoiding activity... shouldn't it be the other way around? Shouldn't these be the people most likely to have the cognitive problems being targeted by the intervention? Did they say what they'd predicted at the start of the trial?

 

[barbc56]

Sigh.

 

[Dolphin]

Thanks D.

Interesting that those who were most depressed reported benefiting least from self-instruction -I could cheekily suggest that's depressive realism counter-acting response bias!

Also, those who reported avoiding activity reported benefiting least from self-instruction designed to discourage people from avoiding activity... shouldn't it be the other way around? Shouldn't these be the people most likely to have the cognitive problems being targeted by the intervention? Did they say what they'd predicted at the start of the trial?

They described it as a post-hoc analysis and didn't mention any specific predictions, as I recall.

This is the introduction which lists previously findings:

Studies that investigated moderators and predictors of treatment outcome in face-to-face CBT for CFS were reviewed. These studies show that focusing on bodily symptoms and attributing symptoms to a physical cause are related to poor treatment outcomes [12,13]. However, evidence concerning the latter is contradictory [6,13,14]. Additionally, patients with a high sense of control with respect to fatigue gain greater benefit from CBT than those with a low sense of control [6] and patients with a low activity pattern tend to show less improvement following CBT compared to those with a high activity pattern [6]. After adapting the treatment manual of CBT for CFS, the relation between the level of physical activity and treatment outcome was no longer present [15]. Good CBT treatment outcomes are associated with a change in avoidance of activity and related beliefs [16]. The prognostic role of depression is still unclear. Some studies found that depression was negatively related to treatment outcomes, whereas others found no relation [17–19]. A recently published study found that baseline levels of depressive symptoms, measured with the HADS, significantly moderated fatigue at 1-year follow-up in an behavioural minimal intervention for CFS [20]. In contrast with these findings Prins et al. [18] found that patients with depression and psychological distress benefited from CBT as much as others. There is also evidence to suggest that high levels of pain are negatively correlated with treatment outcome [21]. In addition, treatment seems to be less successful when patients are older, are members of a self-help group, are involved in a legal procedure concerning disability related benefits, or received a disablement insurance benefit [6,7,17].

 

[A.B.]

Regarding the definition of "avoidance of activity":

Avoidance of activity. Five expressions of avoidance of activity were selected. The patient was asked if she or he (1) avoids symptoms by avoiding strenuous activity, (2) stops with activities in case of pain or when feeling fatigued, (3) restricts him or herself to simple activities, (4) takes a rest by sitting or lying down in case of pain or when feeling fatigued and (5) tries to take as much rest as possible to avoid symptoms. All items were scored on a 4-point Likert scale. A high sum was indicative of a high tendency to avoid activity [23,29].

Isn't this normal behavior for people suffering from fatigue? Patients with severe fatiguing illness such as cancer and MS will probably adopt the exact same behaviour, and anything else would be self harm. Who are the authors to decide that this is suddenly unhealthy pathological behavior?

Patients who were young, had low levels of depressive symptoms, and who had a low tendency to avoid activity benefited more from the intervention than older patients and patients with high levels of depressive symptoms and a strong tendency to avoid activity.

In other words, the more fatigued people were, the less this form of CBT worked.