**3. Results**

*3.1. Search Strategy*

3.1.1. Implementation

> The search strategy and its outcomes are summarised in Table 1 and Figure 1 below:

**Figure 1.** PRISMA Diagram.

At step 9, 143 papers were excluded, either because the focus was not primary care, or because they were not about ME/CFS, or because, although they did concern ME/CFS in primary care, they did not address knowledge or understanding of the condition. The papers identified were extremely heterogeneous with respect to the populations studied, research questions addressed, and methodologies followed, as to preclude any form of meta- synthesis or meta-analysis. Consequently, the synthesis without meta-analysis (SWiM) methodology, which was developed specifically to ensure an adequate standard of review in such circumstances, was utilised [12].

## 3.1.2. Papers Identified

The papers identified in the MEDLINE search were considered in detail within the categories identified in Table 2.


**Table 2.** Summary of papers identified.

\* Note that the total is greater than the number identified in the MEDLINE search, because some qualitative papers are included in more than one category.

#### *3.2. Quantitative Studies*

#### 3.2.1. Surveys of GPs

Seven papers were identified. Saidi and Haines (2006) distributed a postal questionnaire to GPs throughout the U.K., to assess the proportion of practices with children diagnosed with ME/CFS [13]. Of the 112 practices contacted, 62 (55%) had diagnosed children or adolescents with chronic fatigue.

For each of the other six studies, the outcome metric was the proportion of GP respondents to questionnaires who recognised ME/CFS as a genuine clinical entity, and these are summarised in Table 3. Three of these studied GPs were in different parts of the U.K., namely, South Wales [14], Scotland [15] and south-west England [16], while the other papers were from Australia [17], the Netherlands [18] and Ireland [19]. The Australian study reported that 31% of GPs surveyed did not accept ME/CFS as a distinct syndrome [17], but we lacked a full text of this paper.


**Table 3.** Acceptance in general practice of ME/CFS as a genuine clinical entity.

In the Dutch study [18] respondents were not specifically asked whether they accepted the existence of ME/CFS as a genuine clinical entity, and the proportion of GPs who reported that they did not accept ME/CFS as a genuine clinical entity was inferred from the number of those contacted who indicated, via a free text response, that this was their opinion. However, 73% of respondents reported that they had at least one patient with chronic fatigue syndrome, and 83% that they had at least one patient with post-viral fatigue syndrome.

The heterogeneous nature of populations studied, and the research methodologies utilised precluded a formal meta-analysis, but for comparison purposes we have calculated 95% confidence intervals for the British and Irish studies which specifically enquired about the acceptance of ME/CFS as a genuine diagnosis. The higher levels of acceptance of ME/CFS in Scotland and south-west England may demonstrate the impact of secondary referral facilities and active programmes of GP education in those areas. The results are itemised in Table 4.

There were additional findings of relevance in the studies examined. Bowen et al. [16] found that only 52% of respondents expressed confidence in their ability to diagnose the condition, and 59% in their ability to manage it. Sixty-eight percent of respondents to the study in South Wales had diagnosed the condition [14]. In the Irish study, 78% of respondents had patients with chronic debilitating fatigue in their practices [19].

These studies were published over a fourteen-year period, and are consistent in demonstrating that a substantial proportion of GPs, which changed little over that time, did not accept ME/CFS as a genuine clinical entity.


**Table 4.** Acceptance by GPs of ME/CFS—summary statistics.

#### 3.2.2. Surveys of ME/CFS Patients

Seven papers were identified in this section, but three could not be included in the overall comparative analysis, one for the lack of a full text, and the others for absence of relevant numerical information. The first of these, a Belgian study of 177 patients with different GPs, attending a tertiary clinic, found that only 35% of respondents thought that their GPs had experience of the condition, and only 23% felt their GP had sufficient knowledge to treat it [20]. Another Belgian study of 155 patients with ME/CFS recruited via primary care practitioners reported that 43% of subjects self-assessed as having interpersonal problems with their GPs. A disparity with physician assessments was asserted, and the authors concluded that this disparity had to be seen in the context of previous research, demonstrating that patients with ME/CFS tended to feel misunderstood and disrespected. However, this disparity was not reported numerically [21]. Finally, a French report on 231 participants in a clinical trial undertaken in general practice found a tendency in primary care to attribute fatigue to somatic causes in cases with more reported symptoms. They attributed this to a predilection not to entertain somatic explanations of mild or moderate fatigue, but this could not be quantified from the information presented [22].

The remaining four papers are summarised in Table 5. Three of them, from Norway, are interrelated [23–25], and it can be noted that, although the outcome measures in these studies were not precisely the same as that in an American study by Jason et al. [26], and the populations studied and the modes of selection of participants were different, the proportions of respondents expressing reservations about aspects of the quality of primary care were similar in magnitude.

**Table 5.** Patients' opinions about GP care of people with ME/CFS.



**Table 5.** *Cont.*

3.2.3. Other Quantitative Studies

Other quantitative studies identified included two database studies [27,28] a review of medical records [29], and an evaluation of a training programme [30].

Gallagher et al., [27] in an analysis of data from the U.K. General Practice Research Database (now the Clinical Practice Research Datalink), found that, between 1990 and 2001, there was a marked decline in diagnoses of post-viral fatigue syndrome, paralleled by increases in diagnoses of ME/CFS and fibromyalgia, suggesting that diagnostic fashion has a significant part to play in the allocation of diagnostic labels by GPs. A study based on the Norwegian Patient Register found that there were substantial delays in the primary care diagnosis of ME/CFS in children and adolescents. Three-quarters of those patients identified were initially diagnosed with weakness/general tiredness, and for nearly half of them the interval between this initial diagnosis and the definitive diagnosis of ME/CFS was over a year. A comparison with diagnoses of type 1 diabetes mellitus found that only 3.5% of patients were initially diagnosed with weakness/general tiredness, and there was no comparable diagnostic delay [28].

A comparative study of the primary care prevalence of ME/CFS in Sao Paolo and London was carried out by means of a review of medical records [29]. The overall prevalence of chronic fatigue syndrome plus unexplained chronic fatigue was similar in both countries. However, a slightly higher prevalence of chronic fatigue syndrome was apparent among the U.K. patients. The authors attributed this to a cultural factor, namely, a relative lack of recognition of chronic fatigue syndrome among Brazilian doctors, but in fact the difference in prevalence of CFS between the Brazilian and English samples was not statistically significant (prevalence: Brazil 1.6%; U.K. 2.1%. *p* = 0.09).

An American study evaluated a series of five two-day "Train-the-Trainer" workshop training programmes directed towards increasing ME/CFS understanding in primary care [30]. There were marked improvements in both knowledge and self-efficacy, leading to increased confidence in making the diagnosis, but the point was made that the participants were self-selected.

#### *3.3. Qualitative Studies*

#### 3.3.1. Studies of GPs

We identified six papers reporting qualitative studies involving GPs dating from 1993 to 2016. The earliest was from New Zealand [31], and the others were all from the U.K., the most recent four coming from the same team based in north-west England [32–36]. The papers are summarised in Table 6:

All the papers reviewed were consistent in concluding that there were substantial gaps in levels of knowledge and understanding of ME/CFS.

#### 3.3.2. Studies of Patients

Nine papers were identified in this category. Our detailed analysis is summarised in Table 7.

It will be noted that the methodologies followed were extremely heterogeneous, precluding any sort of meta-synthesis, but the overall conclusions in all cases were very similar. Concern was expressed in most cases about the lack of legitimation of the condition, and many GPs were seen as being unsympathetic and lacking in knowledge of the condition, and therefore not a good source of advice. By contrast, a good rapport with the doctor was seen to be very positive, though frequently missing.

#### *3.4. Overview Papers*

The final category identified in this analysis was of a small number of publications which made reference to problems of GP knowledge and understanding of the condition, but presented no empirical research. Bansal wrote a wide-ranging paper centred on the use of a simplified scoring system for the diagnosis of ME/CFS in general practice, in which he described ME/CFS as poorly understood, and refers to disagreements concerning investigation and managemen<sup>t</sup> [37]. Wearden and Chew-Graham reviewed the evidence on the primary care treatment of ME/CFS. They acknowledged that some primary care physicians find ME/CFS hard to diagnose, but argued that early diagnosis and coherent explanation of symptoms would be of benefit [38]. Murdoch produced a straightforward, easy-to-follow guide to the diagnosis and care of patients with ME/CFS, via an illustrative clinical scenario, and asserted that ME/CFS is best managed by the patient's GP in a primary care setting [39]. Campion, in a letter to the British Journal of General Practice, stated that the biopsychosocial model of ME/CFS had caused disagreement between doctors and patients, and that doctors should respect patients, and, given our ignorance of the precise causes of the condition, show humility [40].


**Table 6.** Papers reporting qualitative studies of GPs' knowledge and understanding of ME/CFS.


dissatisfaction with GPs' perceived failure to meet patients' needs.


