Correspondence in Lancet psychiatry, 18 January 2016, by James Charles Coyne, Keith R Laws: Results of the PACE follow-up study are uninterpretable

The PACE follow-up study[1] is something of a curate’s egg, admirable in ambition, but interpretatively indigestible. Although the PACE programme of cognitive behavioural therapy (CBT) or graded exercise therapy (GET) led patients to report less fatigue or greater physical function than patients in the adaptive pacing therapy and specialist medical care groups in the short term, evidence in the long-term follow-up is unconvincing.

The lack of between-group differences at follow-up takes precedence over within-group differences, which are inflated by attribution of any change associated with non-specific factors to the specific interventions. Re-assignment to new treatment also makes between-group comparisons uninterpretable, as we will see.

Although the investigators made heroic efforts to correct for attrition using linear mixed-effects regression models, such efforts depend on their unjustified assumptions of random missing data and of treatment received at follow-up not distorting any signal of effects of the earlier randomisation. Various covariates are introduced for statistical control, but without adequate rationale and documentation of diagnostics. It is therefore doubtful that the complexly adjusted results are reliable.

Several other unfortunate decisions further undermine the findings. Foremost, the unregulated crossover between treatments during follow-up. The follow-up outcome data [1] in table 3 and figure 2 are uninterpretable because they refer to initial randomisation, without reflecting the—quite different treatments—received during follow-up.

The authors argued: “In so far as the need to seek additional treatment is a marker of continuing illness, these findings support the superiority of CBT and GET as treatments for chronic fatigue syndrome”. [1] More participants in the specialist medical care alone and the adaptive pacing therapy (APT) groups received additional treatments during follow-up than did those in the CBT and GET groups, but this finding is hardly surprising when half of participants in the specialist medical care group rated it as not being a logical treatment for them and only 41% were confident about being helped by specialist medical care.[1]

This lack of equipoise was facilitated by the initial description of treatments investigators offered to patients and compounded by a mid-trial Newsletter for patients praising the PACE interventions. It is easy to see how specialist medical care effectively became a waitlist control of frustrated sufferers, who naturally awaited re-assignment. This limitation casts doubts not only on the validity of the follow-up data, but on the integrity of the trial itself.

Putting doubts about the validity of the PACE follow-up aside, let’s take a close look at the unadjusted physical functioning follow-up data in figure 2.1 There are no group differences, and the overall mean short-form 36 (SF-63) physical functioning score is less than 60. It is useful to put this number in context. 77% of the PACE trial participants were women, and the mean age of the trial population was 38 years, with no other disabling medical conditions.

Patients with lupus have a mean physical functioning score of 63,2 patients with class II congestive heart failure have a mean score lower than 60,3 and normal controls with no long-term health problems have a mean score of 93.4.

Lastly, the PACE investigators have previously complained in The Lancet Psychiatry of “the apparent campaign to bring the robust findings of the trial into question”. [5] We think the further scrutiny that the follow-up study has brought casts further doubt on whether there ever were “robust findings”. The investigators should get more accustomed to rigorous post-publication peer review, which is not a campaign, but a reality of the 21st century.

We declare no competing interests.

References
1. Sharpe, M, Goldsmith, KA, Johnson, AL, Chalder, T, Walker, J, and White, PD. Rehabilitative treatments for chronic fatigue syndrome: long-term follow-up from the PACE trial. Lancet Psychiatry. 2015; 2: 1067–1074
Summary | Full Text | Full Text PDF

2. Tench, CM, McCurdie, I, White, PD, and D’Cruz, DP. The prevalence and associations of fatigue in systemic lupus erythematosus. Rheumatology. 2000; 39: 1249–1254
CrossRef | PubMed

3. Juenger, J, Schellberg, D, Kraemer, S et al. Health related quality of life in patients with congestive heart failure: comparison with other chronic diseases and relation to functional variables. Heart. 2002; 87: 235–241
CrossRef | PubMed

4. Bowling, A, Bond, M, Jenkinson, C, and Lamping, DL. Short Form 36 (SF-36) Health Survey questionnaire: which normative data should be used? Comparisons between the norms provided by the Omnibus Survey in Britain, the Health Survey for England and the Oxford Healthy Life Survey. J Pub Health. 1999; 21: 255–270
CrossRef | Scopus (165)

5. Chalder, T, Goldsmith, KA, White, PD, Sharpe, M, and Pickles, AR. Methods and outcome reporting in the PACE trial–Author’s reply. Lancet Psychiatry. 2015; 2: e10–e11
Summary | Full Text | Full Text PDF