The UK CFS/ME National Outcomes Database (NOD) collects assessment and follow-up data on patients attending or visited by NHS specialist CFS/ME services. The aim of the NOD is to enable benchmarking and evaluation of these services. All contributing clinical teams are members of the British Association for CFS/ME (BACME). BACME comprises more than 40 adult and 12 paediatric teams, together assessing more than 6,000 adults and children each year. We included in our study all patients ≥ 18 years and ≤ 64 years old who had attended specialist CFS/ME services provided by clinical teams in Bristol (Frenchay), Wells (Somerset), Leeds (Leeds & West Yorkshire), Barts and The London, and Epsom and St Helier (South West London and Surrey) during the period 01/04/2006 to 31/03/2010. These teams had assessed and treated a constant number of patients each year during the study period. For each team, patients were referred with a diagnosis of CFS/ME and this was confirmed at assessment according to CDC diagnostic criteria [1, 2]. Other teams that contributed data to the NOD during this period were excluded either because they contributed data for a small number of patients or because their CFS/ME service had been established relatively recently.
We collected the following data at the time of assessment by the CFS/ME service: age, sex, ethnicity, diagnostic criteria, and self-reported duration of illness (time elapsed, in months, between onset of symptoms and clinical assessment). Current employment status was recorded in 4 categories: i) "Currently employed full- or part-time"; ii) "Temporarily discontinued because of fatigue-related symptoms"; iii) "Indefinitely discontinued because of fatigue-related symptoms"; iv) "Other". In addition, patients completed the following inventories prior to assessment: 11-item Chalder Fatigue Scale ; 10-item SF-36 physical function subscale , 14-item Hospital Anxiety and Depression Scale (HADS)  and a Visual Analogue Pain Rating Scale (score of 0 for "no pain" and 100 for "pain as bad as possible"). The Chalder Fatigue Scale was scored using the 0-3 method for scoring each question (0 "Less than usual", 1 "No more than usual", 2 "More than usual", 3 "Much more than usual"). On the physical function sub-scale of the SF-36, patients scored 0 ("Yes, limited a lot") 5 ("Yes, limited a little") or 10 ("No, not limited at all") for each question, so that the most disabled patients scored 0 while those with unaffected physical function scored 100. Inventory total scores (and each HADS sub-scale score) were coded as missing if > 1 question was unanswered; if only one item was missing, an adjusted total score was calculated.
Factors associated with discontinuation of employment
We used Chi-squared tests and Student's t-test to compare the characteristics (age, sex, duration of illness, fatigue, disability, anxiety, depression and pain) of patients who were in employment at time of assessment with the characteristics of patients who had temporarily or permanently discontinued employment (before assessment) due to their illness. These characteristics were included in a logistic regression model to identify factors independently associated with discontinuation of employment among patients for whom complete data on all characteristics were available. So that regression coefficients for the different inventories were comparable, they were rescaled so that the range for each was approximately 0 - 10.
Discontinuation of employment
In May 2010, the NOD forms and questionnaires were amended to include more detailed questions about the effect of CFS/ME on employment, specifically, patients were asked to give the "Date when reduced hours/sick leave/unemployment began". We used responses to this question to estimate the duration of unemployment as a fraction of the duration of illness in an external sample of patients ≥ 18 years old assessed by any team contributing data to the NOD during the period 01/06/2010 to 30/11/2010. We used the mean value of this fraction in our calculations, stratified by sex but not by age (due to insufficient numbers in some of the age strata).
Per patient productivity costs due to discontinuation of employment
Productivity costs were estimated using average annual earnings data by sex and age group obtained from the Office for National Statistics . These were used to estimate age- and sex-specific cumulative loss of earnings due to discontinuation of employment prior to assessment by a NHS CFS/ME specialist clinical team. The total productivity costs in the patient sample were estimated as the product of: the number of patients assessed who had a new or confirmed diagnosis of CFS/ME; the fraction who had discontinued their employment; the median duration of illness; the proportional duration of unemployment (mean 'duration of unemployment'/mean 'duration of illness') from the external sample; and mean annual income. This product was divided by the number of patients to yield an estimated total loss per person.
Nationwide productivity costs
Access to CFS/ME specialist assessment throughout the UK is variable [16, 17]. The five CFS/ME teams contributing data to this study represent areas of the country that currently have higher levels of access. We used data from these five centres to extrapolate and estimate the potential nationwide productivity costs due to CFS/ME in patients prior to specialist assessment. The clinical teams in our study were asked to identify the Primary Care Trusts (PCTs) from which the majority of patients were referred. The annual proportion of the population who access specialist assessment was estimated by dividing the number of patients in the sample by the number of study years (four) and then by the estimated population in the PCTs served by the teams in our study in each age and sex stratum . These proportions were multiplied by UK population estimates  to estimate the total population of people with CFS/ME in the UK who might receive specialist assessment each year. The nationwide productivity costs due to CFS/ME prior to specialist assessment were estimated for this population using age and sex stratified earnings data, as previously described. In two one-way sensitivity analyses, we used the lower and upper bounds of the 95% confidence intervals for the fractions of those who had discontinued their employment and the fraction 'duration of unemployment'/'duration of illness' to estimate upper and lower bounds for the total UK loss.
The North Somerset & South Bristol Research Ethics Committee decided that the collection and analysis of CFS/ME patient data were part of service evaluation and as such did not require ethical review by a NHS Research Ethics Committee or approval by NHS Research and Development offices (REC reference number 07/Q2006/48).