Q J Med 2002; 95: 833-834
© 2002 Association of Physicians
Correspondence |
Heartlands Research Institute, Birmingham Heartlands Hospital, Birmingham e-mail: martin.wildman{at}lshtm.ac.uk
Sir,
We were interested to read the editorial by Raoult1 relating to our papers on chronic fatigue following Q fever. In the editorial, he suggested that the reported percentage of fatigue amongst controls in the Wildman et al. study2 was high in contrast to studies from England, which he cited as showing a fatigue prevalence of to 9.9–11.7%. It is unclear where the figures of 9.9–11.7% come from, as the paper referenced is entirely devoted to peer review.3 It is important to emphasize that a thorough understanding of the difficulties of fatigue definition was fundamental to the design of the Wildman paper, and that in fact the levels of fatigue found in the controls was exactly what would be expected in controls drawn from a UK general practice population. It is the use of well-defined instruments with large normal population reference data that makes the Wildman study of particular utility in gaining insight into patterns of fatigue following Q fever. Our study used the International Chronic Fatigue Syndrome study group's 1994 Centre for Disease Control definition, 4 operationalized with identical instruments to those used by Wessely. Wessely's primary care study is pivotal in any study of fatigue, because it provides the largest survey of fatigue prevalence in the ‘normal’ UK general practice population.5 This well-validated operationalization of the CDC definition of fatigue allows patients to be classified into three levels of increasing severity of fatigue: (i) fatigue, (ii) idiopathic chronic fatigue, and (iii) chronic fatigue syndrome. Wessely found the prevalence of fatigue to be 38% (95%CI 37.2–38.7) using questionnaires returned by post by individuals identified from a UK general practice population. This corresponds very well with the prevalence of fatigue of 36.3% (95%CI 25.8–47.8%) in controls returning the identical questionnaire in our study. Wessely found ‘idiopathic chronic fatigue’ in 18.3% (95%CI 17.7–18.9%) of his normal general practice population, in comparison to 15% (95%CI 8.0–24.7%) in the Wildman controls.
Painstaking attention to the quantification of fatigue using well-defined fatigue instruments is essential in any studies of this nature, and we were scrupulous in this regard. The editorial by Raoult did not cite the large Wessely series, but instead describes Raoult's experience of finding a prevalence of undefined asthenia of 5–10% in patients 6 months after Q fever onset, and asks why this 5–10% differs from the findings in the UK study. It is quite clear from the extensive literature on the difficulties in the measurement of fatigue that lack of explicit measurement instruments will make comparison of fatigue between studies impossible. In 1994, the Centre for Disease Control recommended that painstaking care should be taken in defining and measuring fatigue. 4 If workers are to make progress in the understanding of fatigue following Q fever, it is crucial as a first step that standard instruments are used. Our study applied these principles to the investigation of fatigue following Q fever, and the Wessely instruments used have performed exactly as expected in the controls, suggesting that they have utility in tackling this area. The Wessely instruments do have the potential to identify differences in fatigue between Q exposed individuals and controls; however, the methodological problems inherent in following up patients exposed to salient febrile illness are apparent. Our study makes it clear that the Wessely instruments would allow an important next step in understanding the prevalence of fatigue following Q fever to be made, if a large simultaneous questionnaire and serology study were carried out where respondents are blinded to their serology results until after the Wessely questionnaires are completed.
We reiterate that the increased fatigue scores in the Q exposed cohort have been measured with instruments that were standardized and well-validated, permitting other workers the possibility of replication in a way that will allow meaningful comparisons to be made. However, we readily acknowledge that the study design has the consequence that it is not possible to determine whether the elevated fatigue scores are due to bias, confounding, or the effects of persistent antigen.
References
1. Raoult D. Q fever: still a mysterious disease. Q J Med 2002; 95:491–2.
2. Wildman MJ, et al. Chronic fatigue following infection by Coxiella burnetii (Q fever): ten-year follow-up of the 1989 UK outbreak cohort. Q J Med2002; 95:527–38.
3. Wessely S. Peer review of grant applications: what do we know? Lancet1998; 352: 301–5.[Web of Science][Medline]
4. Fukuda K, et al. The chronic fatigue syndrome: a comprehensive
approach to its definition and summary. Ann Intern Med 1994; 121:953–9.
5. Wessely S, et al. Postinfectious fatigue: prospective cohort study in primary care. Lancet1995; 345:1333–8.[Web of Science][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||