Chronic fatigue syndrome (CFS) occurs in 0.4–2.5% of the population, depending on how it is defined.¹ CFS is characterized by post-exertional fatigue not relieved by rest, lasting more than six months' duration. The symptoms are not due to any alternative medical or psychiatric disorder, and can be accompanied by impairment of short-term memory and concentration, headache, tender lymph nodes, sore throat, post-exertional malaise lasting more than 24 hours, muscle pain, multi-joint pain, and unrefreshing sleep.² CFS causes a significant functional impairment. It is a clinical diagnosis made after excluding any explanatory medical or psychiatric disorder. There are no laboratory investigations to confirm the diagnosis of chronic fatigue syndrome.³

The NICE guideline for the assessment and management of CFS recommends that a diagnosis of CFS in adults should be made in primary care.⁴ Patients are then referred, if their illness is long-lasting or complex, to specialist CFS clinics for confirmation of the diagnosis and management of their illness. However, general practitioners (GPs) do not feel confident in making the diagnosis.⁵ To compound this, some specialist CFS clinics do not employ doctors, and rely on an accurate diagnosis being made in primary care.

Newton and colleagues have recently shown that misdiagnosis is common in patients seen in a specialist service.⁶ They found that 40% of referrals to a specialist CFS service had an alternative diagnosis, most commonly primary sleep disorders (including obstructive sleep apnoea) and psychiatric disorders (including depression, anxiety and post-traumatic stress disorder). This study aimed to explore the prevalence of alternative diagnoses in patients referred with a definite or provisional diagnosis of CFS.

These service evaluation surveys were conducted at the specialist CFS clinic at St Bartholomew's Hospital, London. This service accepts referrals from primary care for assessment and management of patients with CFS. All the referrals were routinely screened by both a consultant liaison psychiatrist and a consultant physician in infectious diseases. Once the referral had been accepted, the patient was offered an assessment that included a detailed history, physical and mental state examination and laboratory investigations. Management plans depended on the diagnosis following the assessment.

The first survey examined all the referrals which were rejected between March 2007 and September 2008. The survey was based on analyses of the letters to the referrer in all the cases not accepted. The second survey studied 250 consecutive new patients assessed in the same clinic, between April 2007 and November 2008, for the primary diagnosis following clinical assessment. This survey examined the case-notes for outcomes of assessments. In both surveys, the most recent case-notes and letters were examined in cases where the outcomes were not clear.

Almost half (49%) of all patients referred to a specialist CFS service did not have a diagnosis of CFS. Thirty-seven percent of referrals, screened by both a psychiatrist and a physician, were declined at the point of referral, sometimes for more than one reason, of which 61% were for likely alternative medical and psychiatric diagnoses. A further 46% of patients assessed did not receive a diagnosis of CFS. We expected a significant minority of referrals not to have a diagnosis of CFS, but did not expect that so many of the assessed patients would not have a diagnosis of CFS.

The strengths of this paper are that the two surveys studied referrals and assessments over the same 18 months period with a large number of cases. Both the surveys were carried out at similar times and the results are complementary. The limitations of this paper are that it is based on two surveys of only one service; we cannot be sure that the rejected referrals did not have a diagnosis of CFS; and the information from this cannot answer the primary question on effective identification and management of CFS in primary care.

These figures are reflected in a study by Euba and colleagues, which showed that 44% of patients referred to specialist clinic did not receive a diagnosis of CFS.⁷ A more recent study found that 40% of patients diagnosed with CFS in a specialist centre did not have CFS.⁶ Added to our findings, this would suggest that between 40–50% of patients referred to specialist services do not have a diagnosis of CFS.

Patients with chronic fatigue in general have strong diagnostic associations with both psychiatric and medical conditions.^8,9 Teasing these out to accurately identify patients with CFS, especially in the absence of confirmatory laboratory investigations, can be difficult. Harvey and Wessely suggest that looking for common medical and psychiatric illnesses in patients with chronic fatigue will improve their rate of detection.³ They add that a basic mental state examination is the most productive tool, as depression is the most common exclusionary and co-morbid condition with CFS. The most commonly applied research diagnostic criteria for CFS stresses the importance of a mental state examination.¹⁰ Even specialist doctors trained in CFS can make errors of psychiatric diagnoses, with non-psychiatrists missing psychiatric diagnoses more than psychiatrists.¹¹ A diagnosis of depressive illness is more likely than CFS when fatigue is relieved rather than exacerbated by exertion, when fatigue is not the primary symptom, and when low mood dominates the clinical picture.

Bowen and colleagues found that 48% of GPs were not confident of making a diagnosis of CFS, and 41% were not confident enough to provide treatment.⁵ Despite guidance now being available,⁴ a qualitative study suggested that GPs were not confident in making a diagnosis of CFS, with concerns that such a label might be harmful to the patient.¹² The current evidence suggests that a diagnosis of CFS probably is helpful and enabling, so long as it leads to a constructive treatment plan.¹³

The NICE guideline states that primary care doctors should diagnose patients with CFS and identify patients for referral onto specialist CFS services.⁴ Accurate diagnosis of the primary condition and identification of co-morbid illnesses will help in better management of the illness and improve quality of life. However, diagnosing CFS can be complicated, as described above, and specialist CFS services are well placed to provide such a service, especially if they have a mix of specialist clinicians who can accurately identify CFS and exclusionary medical and psychiatric illnesses. We argue that the NICE guideline's advice to diagnose within primary care may not be the best guidance, and more specialist CFS services (with a mix of specialist doctors) are needed to provide a diagnostic service that meets the requirements of patients. However, this has to be balanced against resources available, and commissioners of services will have to consider how best to provide a clinically effective service to patients while being cost-effective at the same time.

In conclusion, this paper replicates two previous studies finding high rates of misdiagnosis in patients accepted and assessed in a secondary care CFS service. The NICE guideline suggests that patients with CFS should be diagnosed and managed in primary care, and specialist CFS services used for patients with severe and disabling symptoms. Our results suggest that this may not be the best guidance. These results have implications for training, care pathways and service design, particularly in primary care and those specialist services without a doctor. Specialist services need doctors who are equally confident in making both medical and psychiatric diagnoses.