The Qure study was one of the largest randomized treatment trials for patients suffering from a postinfectious fatigue syndrome. Although the study focused on Q-fever, a disease caused by the bacterium ‘Coxiella burnetii’, it might have implications for patients suffering from long-lasting fatigue after COVID-19 or other infections. Dutch physician Mark Vink has written a detailed analysis of the Qure study. This blog post tries to summarize the most interesting findings.
The Qure study took place in the Netherlands where one of the largest outbreaks of Q-fever occurred in the south-eastern part of the country between 2007 and 2010. Many people reported long-lasting fatigue and disability after a Q-fever infection, a condition that has been termed Q-fever fatigue syndrome or QFS in short.
The Qure study recruited 154 QFS patients and randomized them into three groups: one that received long-term antibiotics (doxycycline), one that received a placebo and, one that received cognitive behavioral therapy (CBT). The primary outcome was a fatigue questionnaire. Patients nor researchers knew who received the antibiotics and who got a placebo. I can already give away that there was no significant difference between these two groups at the end of treatment. What is of interest in this blog post, are the results of the CBT group.
Because of the nature of CBT, where active participation of patients is required, blinding of participants is impossible. While patients in the other two groups were uncertain whether they got an intervention or a placebo, patients in the CBT group (and their therapists) knew they were in one of the intervention groups.
The lack of blinding in the CBT group might have caused or inflated treatment effects. Patients who know they are receiving an active intervention rather than a passive control might be more optimistic about its effect on their health. They might also report symptoms according to what they think will please the therapists and investigators with whom they have worked closely for several weeks. As I have argued previously, this trial design is common in studies on rehabilitative interventions for myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) and it makes it quite difficult to interpret the results.
Usually, the findings of these trials adhere to the following pattern:
1) In the main publication, the behavioral intervention is said to be effective because, at the end of treatment, symptom questionnaires show a statistically significant difference compared to a control group. The difference, however, is usually quite small and sometimes not clinically significant.
2) At long-term follow-up the difference between the intervention and control group disappears or is no longer statistically significant. These results are usually published several years later in a less prestigious journal and receive less attention.
3) Objective outcomes that are less prone to bias such as activity levels or work resumption do not support the improvements seen on symptom questionnaires. These objective measurements are not published in the main publication because the authors prefer not to register them as a primary or secondary outcome or discuss the implication of their results.
The findings of the Qure study follow this pattern closely.
1) At first, a statistically significant difference in fatigue scores was highlighted between the CBT and placebo group. The difference, however, was not very large: a 6.2 point difference on a 56-point scale. The authors defined a clinically meaningful improvement as a reduction of 9 points on the scale.*
2) At follow-up, the improvements of the CBT group suddenly disappeared. Patients who received CBT experienced a relapse and the placebo group performed just as well at long-term follow-up.
3) The only objective measurement used in the trial (physical activity recorded with an actometer device) was not reported in the main publication but in a mediation analysis that came out two years later. Although the actigraphy results were not fully reported, the data suggest that there was no significant difference between the CBT and placebo group.
The authors of the study were a bit perplexed by the relapse in the CBT group at the follow-up assessment. They concluded that those patients didn’t receive enough CBT and that extra booster-sessions should be explored to maintain the treatment effect. They wrote:
“The beneficial effect of CBT on fatigue severity at EOT [end of treatment] was not maintained 1 year thereafter. Due to its initial beneficial effect […], we still recommend CBT as treatment for QFS.”
This is a bit perplexing. I think a more plausible explanation exists, one that doesn’t require the assumption that patients experienced a significant improvement after CBT followed by an equally large relapse shortly afterwards.
I think patients in the CBT group simply filled in the questionnaires a bit ‘better’ in the eyes of their therapists because the intervention encouraged them to do so. The protocol explained how patients receiving CBT “are encouraged to perceive feelings of fatigue as a normal part of an active and healthy life.” Another section explains how the “therapist elicits and challenges patients’ non-accepting and catastrophising beliefs with respect to fatigue. Additionally, patients are taught how to distract their attention from their fatigue.”
These encouragements might explain why patients in the CBT group reported having a bit less fatigue at the end of treatment compared to patients in the other groups who received no such instruction. When the same questionnaire was filled in a year later – a long time after they were told to distract their attention from their fatigue – there was no longer any difference between the two groups. In other words, the changes on the fatigue questionnaire might simply be the result of response bias, an explanation that is supported by a lack of improvement on activity levels, the only objective measurement in this trial.
I suspect that the same interpretation problem will arise in other trials on behavioral interventions for patients with postviral fatigue syndromes. Often these interventions actively try to manipulate how patients interpret their symptoms and the attention they give to them. Therefore, I think some of the implications of the Qure study might go beyond the QFS patient population. A more detailed analysis of this study is available in the paper by Vink & Vink-Niese published last week in the journal Healthcare:
“Vink, Vink-Niese. Could Cognitive Behavioural Therapy Be an Effective Treatment for Long COVID and Post COVID-19 Fatigue Syndrome? Lessons from the Qure Study for Q-Fever Fatigue Syndrome. Healthcare 2020, 8(4), 552; https://doi.org/10.3390/healthcare8040552
EDIT: Lou Corsius has also written extensively about the Qure study and reported how its results were presented rather misleadingly to an audience of occupational physicians.
*They also required a score lower than 35 points because this threshold was used as an inclusion criterion.
1 thought on “A closer look at the Qure Study”