Specialist treatment of chronic fatigue syndrome/ME: a cohort study among adult patients in England

Background NHS specialist chronic fatigue syndrome (CFS/ME) services in England treat approximately 8000 adult patients each year. Variation in therapy programmes and treatment outcomes across services has not been described. Methods We described treatments provided by 11 CFS/ME specialist services and we measured changes in patient-reported fatigue (Chalder, Checklist Individual Strength), function (SF-36 physical subscale, Work & Social Adjustment Scale), anxiety and depression (Hospital Anxiety & Depression Scale), pain (visual analogue rating), sleep (Epworth, Jenkins), and overall health (Clinical Global Impression) 1 year after the start of treatment, plus questions about impact of CFS/ME on employment, education/training and domestic tasks/unpaid work. A subset of these outcome measures was collected from former patients 2–5 years after assessment at 7 of the 11 specialist services. Results Baseline data at clinical assessment were available for 952 patients, of whom 440 (46.2%) provided 1-year follow-up data. Treatment data were available for 435/440 (98.9%) of these patients, of whom 175 (40.2%) had been discharged at time of follow-up. Therapy programmes varied substantially in mode of delivery (individual or group) and number of sessions. Overall change in health 1 year after first attending specialist services was ‘very much’ or ‘much better’ for 27.5% (115/418) of patients, ‘a little better’ for 36.6% (153/418), ‘no change’ for 15.8% (66/418), ‘a little worse’ for 12.2% (51/418), and ‘worse’ or ‘very much worse’ for 7.9% (33/418). Among former patients who provided 2- to 5-year follow-up (30.4% (385/1265)), these proportions were 30.4% (117/385), 27.5% (106/385), 11.4% (44/385), 13.5% (52/385), and 17.1% (66/385), respectively. 85.4% (327/383) of former patients responded “Yes” to “Do you think that you are still suffering from CFS/ME?” 8.9% (34/383) were “Uncertain”, and 5.7% (22/383) responded “No”. Conclusions This multi-centre NHS study has shown that, although one third of patients reported substantial overall improvement in their health, CFS/ME is a long term condition that persists for the majority of adult patients even after receiving specialist treatment. Electronic supplementary material The online version of this article (doi:10.1186/s12913-017-2437-3) contains supplementary material, which is available to authorized users.


Background
Chronic Fatigue Syndrome (CFS), also known as 'ME' , is a long-term disabling condition characterised by debilitating fatigue of unknown cause, post-exertional malaise, cognitive dysfunction and disturbed/unrefreshing sleep, plus other symptoms including muscle and joint pain, headaches, and dizziness [1]. CFS/ME imposes an immense burden on patients, carers and families [2,3]. In the UK, adults who attend NHS specialist CFS/ME services have been ill for a median duration of 3 years, and half of those employed at the onset of their illness cease working [4]. A meta-analysis of CFS/ME prevalence studies based on clinically-confirmed cases in several countries gave a prevalence of 0.8% (95% CI 0.2% to 1.3%) [5].
Around 8000 patients are treated annually by NHS specialist CFS/ME services in England [6]. There are approximately 50 such services in England, many of which were established under the CFS/ME Service Investment Programme (2004)(2005)(2006) [7]. These services follow guidance provided by the National Institute for Health & Care Excellence (NICE), including specific guidelines for diagnosis, specialist care, and ongoing management, with an overall patient-centred approach to treatment [8].
In this study we used initial assessment and patientreported outcome data from newly referred and former patients who attended specialist CFS/ME services in England to investigate outcomes 1-5 years after initial assessment by the service. We also collected patientlevel treatment data to describe variation in treatments across specialist services.

Study cohort -newly referred patients
Newly referred patients were recruited from 11 specialist CFS/ME services across England (10 NHS services, 1 registered independent provider) during the period 01/ 06/2014 to 30/09/2016. Patients were eligible if they were ≥18 years old and had a CFS/ME diagnosis made or confirmed at an initial clinical assessment appointment in accordance with NICE guidelines [8].

Patient-level data -newly referred patients
At the time or their initial assessment, patients complete standard questionnaires to obtain quantitative measures of fatigue (Chalder Fatigue Scale [9] and Checklist Individual Strength (CIS20-R) [10]), physical function (RAND SF-36 [11]), general function (Work & Social Adjustment Scale [12]), mood (Hospital Anxiety & Depression Scale (HADS) [13]), pain (visual analogue pain rating scale), daytime sleepiness (Epworth Sleepiness Scale [14]), and sleep problems (Jenkins Sleep Scale [15]). Patients were also asked about the impact of ill health on employment, education/training, and unpaid work (e.g. housework, child care, voluntary work). The same set of questionnaires was sent to patients by post or via email (with a link to online versions of the questionnaires) approximately 12 months after their initial clinical assessment. At follow-up, patients were also asked to rate changes in overall health and CFS/ME (Clinical Global Impression scale), and were asked "Do you think that you are still suffering from CFS/ME?" Patients who didn't respond were contacted by the clinical team via phone or email on up to 2 further occasions to elicit a response. Data on treatment received by each patient were recorded by clinical teams or extracted from hospital administration databases at the end of the study. Treatment data included the date, duration and type of session (group or individual), and the qualification of the health care professional delivering the session.

Study cohort -former patients
For 7 of the 11 services, random samples of pseudonymous patient identifiers (50-60 per year for the period 2010-2013) were obtained from the CFS/ME National Outcomes Database (NOD). The CFS/ME NOD is a centralized repository of clinical assessment and patientreported outcome data which were routinely collected by NHS specialist CFS/ME services across England from 2006 to 2013 for the purpose of service evaluation. Lists of pseudonymous patient identifiers were sent to the clinical teams, who then sent out information sheets, consent forms and questionnaires to patients' home addresses.

Patient-level data -former patients
Patients who were treated by specialist services during the period 2010-2013 had already completed questionnaires to provide measures of fatigue (Chalder Fatigue Scale [9]) and physical function (RAND SF-36 [11]) when they were first assessed. At follow-up, former patients were sent these two questionnaires and were asked about changes in overall health and CFS/ME (Clinical Global Impression scale), and whether there had been any changes in employment, education/training, and their ability to do unpaid work/domestic tasks and social/leisure activities. They were also asked "Do you think that you are still suffering from CFS/ME?" Patient-level treatment data for former patients were not extracted from medical records.

Statistical analysis
Characteristics of patients who provided follow-up data were compared with patients who did not respond using the Chi-squared test for proportions and Kruskal-Wallis test for continuous measures (both α = 0.05). Changes (mean differences) in patient-reported measures between baseline (initial clinical assessment) and 1-year follow-up, and mean patient-reported measures at baseline and 1-to 5-years follow-up were plotted with 95% confidence intervals. Evidence for associations between baseline comorbidities and overall change in health at 2-to 5-year was assessed using the Chi-squared test (α = 0.05).

Results
One-year follow-up of newly referred patients Of the 1067 newly referred patients recruited by the 11 services during the study period, 952 (89.2%) had baseline data and were available for follow-up (Fig. 1   hours, p = 0.06). There were no differences between responders and non-responders in any other characteristics or measures.
Treatment data were available for 77.9% (742/952) of patients with baseline data, and for 98.9% (435/440) of patients with baseline and follow-up data (Table 1). Of patients who provided follow-up data, 40.2% (175/435) did so after they had been discharged. The remaining 59.8% (260/435) had not yet completed treatment or had an open appointment or future follow-up appointment. Four services (D, E, F, and K) achieved ≥60% follow-up. Follow-up at the other services ranged from 23% to 52%. One service (G) had insufficient data (n = 7) for analysis of outcomes.
At 1 year after initial assessment, there was substantial variation in total duration of treatment sessions across services (Fig. 2, Table 1) and in the mix of individual and group sessions, type of health care professional delivery the session and/or content of the session (Additional file 1: Table S2). Median total duration of treatment sessions ranged from 19 to 20 h for services C and K, to 4 h for services D and J. One service (D) provided only individual therapy sessions, two services (E and G) provided mainly (>90%) individual sessions, two services (C and I) mainly group therapy, and the remainder had ratios of individual: group therapy between 50:50 and 60:40. The median duration of therapy sessions was longer among patients with vs without follow-up data (8 (IQR 5-17) vs 6 (3-12), p < 0.001) and was shorter among patients who had been discharged vs patients who were still under treatment at the time of follow-up (6 (3-13) vs 9 (6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19), p < 0.001).
Changes (mean difference) in patient-reported measures between baseline and follow-up are shown in Fig. 3 and Table 2. There was evidence (p < 0.05) of reduction in fatigue on the Chalder Fatigue Scale across all services, with some variation in size of effect. Changes in other outcome measures tended to indicate beneficial effects of treatment, particularly in general function (Work and Social Adjustment Scale), depression, sleep (daytime sleepiness and sleep problems) and concentration and motivation (CIS20R). Evidence for improvements in physical function (SF36), pain, anxiety, and activity (CIS20R) was weaker and less consistent across the services.
Overall, 18% (71/394) of patients reported returning to work or increasing working hours since first attending a specialist service, whilst 30% (118/394) reported having ceased working or reduced hours because of CFS/ME and 47% (186/394) reported no change (Table 3). Increased ability to perform unpaid work and domestic tasks was reported by 35% (129/372) of patients, reduced ability by 34% (127/372), and no change by 29% (109/ 372). Similarly, 35% (139/397) of patients reported being able to do more social and leisure activities, 36% (143/ 397) reported being less able, and 26% (102/397) reported no change. Only 5% (11/240) reported returning to, or increasing hours of education/training, whilst 13% (31/240) ceased or reduced their hours and 79% (189/ 240) reported no change.        Overall changes in health were reasonably consistent across the specialist services, with 28% (115/418) reporting their health as much better or very much better, 65% (270/418) reporting little or no change, and 8% (33/418) reporting much worse or very much worse health (Table 4, Fig. 4). Of those reporting little or no change, the majority said that they were 'a little better' (36.6% (153/418)), with 15.8% (66/418) indicating 'no change' and 12.2% (51/418) a slight deterioration. Similar proportions were observed in response to the question "Overall, how much do you feel your CFS/ME has changed since you first came to the service?" In response to the question "Do you think that you are still suffering from CFS/ME?" 87% (341/391) responded "Yes", 3% (11/391) "No" and 10% (39/391) were "Uncertain". Of those who responded "Yes", 23% (78/339) also said that their overall health was much or very much better, compared with 82% (9/11) of those who responded "No" and 58% (21/36) of those who were "Uncertain". Changes (mean difference) in patient-reported measures between baseline and follow-up for each level of overall improvement in health are provided as supplementary information (Additional file 2: Figure S1 and Additional file 1: Table S3).

Long-term (2-to 5-year) follow-up of former patients
Seven of the specialist services attempted to contact random samples of former patients who were assessed and treated in 2010, 2011, 2012 or 2013 (total N = 1265), and long-term follow-up questionnaires were returned by 30% (385/1265) of these patients (Fig. 5). Baseline characteristics (at time of assessment) were the same in patients who did vs did not respond, except that patients who responded tended to be older (43 (33-51) vs 38 (30-47) years, p < 0.001) (Additional file 1: Table S4), and less likely to have presented with comorbid depression (26.6% (89/335) vs 34.8% (247/709), p = 0.008). Otherwise, there were no differences between responders and non-responders in any other baseline characteristics or measures. There was no difference between the two groups in their responses to the Clinical Global Impression question "Overall, how much do you feel your health has changed since you first came to the CFS/ME service?" at 1 year follow-up, with 32% (84/263) of non-responders vs 38% (84/223) of responders indicating that they were very much or much better, 58% (153/263) vs 56% (124/223) reporting little or no change, and 10% (26/263) vs 7% (15/223) indicating that they were very much or much worse (p = 0.26).
Overall change in health at 2-to 5-year follow-up was associated with comorbid fibromyalgia (p = 0.005) and depression (p = 0.02) recorded at the patient's baseline clinical assessment, but was not associated with comorbid anxiety, migraine, or irritable bowel syndrome. Very much or much better health was reported by 15

Discussion
This study has described substantial variation in total therapy time for adult patients attending NHS specialist CFS/ME services in England. Changes in patientreported outcome measures approximately one year after patients' initial assessments demonstrated clear improvement, particularly in fatigue, general function, depression, sleep, concentration, and motivation. At one year follow-up, one quarter of patients rated their overall health as very much or much better, the majority (two thirds) reported little or no change (of whom half had improved slightly), and <10% reported much or very much worse health. Just under 90% of patients said that they were still suffering from CFS/ME, although one quarter of these also said that their overall health was very much or much better. Among former patients who returned long term (2-to 5-year) follow-up questionnaires, one third rated their overall health as very much or much better, half reported little or no change, 17% reported substantial deterioration, and 85% said that they were still suffering from CFS/ME.

Strengths and limitations
The main strength of our study is that a large cohort of patients was recruited from NHS specialist services in England, all of which follow NICE guidelines for the diagnosis and management of CFS/ME [8]. This should ensure that the patients in our study have been diagnosed with CFS/ME according to the same criteria, namely persistent or recurrent debilitating fatigue of ≥4 months' duration which is not lifelong, or the result of ongoing exertion, or alleviated by rest, or explained by other conditions, and which results in a substantial reduction in activity. We measured patient-reported outcomes using standardised questionnaires which are used routinely in clinical practice, and which have also been used widely in epidemiological research [16,17].
The main limitation of this study is the substantial loss to follow-up, both among the newly referred patients and among the former patients. Although 12-month follow-up of newly referred patients was 74% at one service and 60-65% at three services, follow-up at the other seven services ranged from 23 to 52%. Long-term follow-up of former patients was 30%. These losses to follow-up mean that all of our analyses must be interpreted with the caveat that bias could be introduced in either direction if patients who were more (or less) likely to recover were more (or less) likely to return their questionnaires. Another limitation is that we measured outcomes in newly-referred patients at a fixed time after their initial assessment, with 60% of patients still being under treatment at time of follow-up. Also, measurement of outcomes at a single time point is not ideal for an illness which can fluctuate over time.

Comparison with previous literature
The best comparators for our real world results are the results reported by the PACE trial of cognitive behaviour therapy (CBT) and graded exercise therapy (GET) for adults with CFS/ME. The PACE trial showed a reduction Fig. 4 Clinical Global Impression responses at 1-year follow-up across CFS/ME specialist services (in the CBT and GET arms) in mean fatigue (on the Chalder Scale) from 28 points at randomisation to 20 points at 12 months [18], followed by a slight (1-2 point) further reduction at 31 (IQR 30-32) months follow-up [19]. This is very similar to the trend observed in our data. PACE reported larger improvements in physical function (from 37 to 39 points at baseline to 58 points at 12 months followed by slight or no further change at long-term follow-up) than observed in our study, a difference between trial and real world outcomes which we found previously in an analysis of routinelycollected data [20]. The proportions of PACE trial participants (in the CBT and GET arms) who rated their overall health as very much or much better at 12 and 31 months were 41% and 42-48%, respectively, compared with 28% and 30% in our study (at 12 months and 2-5 years).
We cannot say whether these differences represent better long term outcomes attributable to the treatment programmes in the trial, the effect of the majority of patients in our study still being under treatment, or bias in our study if patients who improved were more likely to be lost to follow up. Participants in the CBT and GET arms of the PACE trial received (median (IQR)) 14 (12)(13)(14)(15) individual CBT sessions (86% face-to-face, 14% by telephone) and 13 (12)(13)(14) individual GET sessions (94% faceto-face, 6% by telephone) respectively, and one fifth went on to receive further therapy sessions after the trial ended. Only four services in our study (C, F, G, and K) provided a similar number of therapy sessions, with considerable variation in follow-up and outcomes. One of these services (F) provided 12 (6-13) therapy sessions and follow-up was obtained from 74% (58/78) of patients of whom 40% (23/57) rated their overall health as very much or much better. In this service, the ratio of individual: group therapy was 50:50, which suggests scope for even better cost-effectiveness than reported in the PACE trial [21].

Implications for clinical practice and future research
Pointers for future research are perhaps given by some of the between-service differences in outcomes. For example, patients who were treated by service D (with 60% follow-up) reported particularly good improvements in physical function despite receiving a relatively small number (median 7 (IQR 3-8)) of (mostly GET) sessions -this service also delivered 28% of therapy sessions via telephone. Service K (with 65% follow-up) obtained similar improvements in physical function by means of a larger number of sessions (12 (9-13), mostly with an occupational therapist) of which 65% were group sessions. These similarities in outcomes and differences in modes of delivery merit further exploration, and could form the basis of future randomised controlled trials. Similarly, treatments in our study were delivered by multidisciplinary teams variously comprising specialist clinicians, clinical psychologists, occupational therapists, and physiotherapists, which highlights the lack of an evidence base regarding the optimal mix of health care professionals required to deliver the most cost-effective service. Factorial trials are particularly suited to addressing multiple questions about content and delivery of treatments [22],  Table 5 Median (IQR) fatigue and physical function at baseline (assessment) and at 1-to 5-year follow-up among patients treated by CFS/ME specialist services  (Chalder) and physical function (SF36) from initial assessment up to 5 years follow-up among patients treated by CFS/ME specialist services and these could also be designed (by randomising predefined subgroups of patients with specific comorbidities and/or symptom profiles) to investigate the big question in CFS/ME treatment, namely why some patients experience substantial improvement whilst others report little or no improvement. Qualitative research has shown that specialist CFS/ME services play an important role in patients' journeys towards improved quality of life [23,24]. Anecdotally, services report overall high patient satisfaction, which may appear to be at odds with a minority of patients experiencing substantial improvement. This apparent paradox may be explained in part by the difficulty of measuring long-term outcomes in a complex chronic illness [16,25], a problem which could perhaps be addressed by using objective rather than subjective measures [26]. We might also consider how outcomes in our study compare with other chronic illnesses, such as chronic widespread pain, fibromyalgia, arthritis and multiple sclerosis [22,27,28], and with the poor prognosis for CFS/ME in adults if no specialist treatment is received. Our qualitative study was set in three of the eleven services in the present study [23]. It showed that referral to a CFS/ME specialist service was typically the first positive step in coming to terms with a life changing contested illness, with improvement coming about through a process which included validation of patients' experiences, acceptance of change, practical advice and support, and therapeutic outcomes. Professional support was an important theme, and we recognize that some of the between-service differences in outcomes in the present study may be attributable to therapist effects [29][30][31].

Conclusions
This multi-centre study in the NHS has shown that CFS/ME is a long term condition that persists for the majority of adult patients even after receiving specialist treatment. Whilst 50-65% experienced little or no change in their condition 1-5 years after accessing a specialist service and 10-20% reported a deterioration, up to 30% of patients reported very much or much better health (and the majority of those who experienced little or no change had improved slightly). Given the adverse impact of CFS/ME on patients and their families, substantial improvement in 20-30% of the approximately 8000 patients treated each year by specialist CFS/ ME services in England represents a large individual and societal benefit, and supports the argument that services need to be sufficiently resourced to treat promptly all newly diagnosed patients according to individual need.

Additional files
Additional file 1: Table S1. Characteristics of newly-referred patients with and without 12-month follow-up data across CFS/ME specialist services. Table S2. Treatments received by newly-referred patients across CFS/ME specialist services. Table S3. Mean change (95% CI) in patient-reported outcome measures between assessment and 1-year follow-up, categorised by overall improvement in health. Table S4. Baseline characteristics of former patients with and without 2-to 5-year follow-up data across CFS/ME specialist services. (DOCX 27 kb) Additional file 2: Figure S1. Changes (mean difference with 95% CI) in patient reported measures between initial assessment and 1-year followup by overall improvement in health among patients treated by CFS/ME specialist services. (TIFF 12271 kb)