Estimates of the number of people (prevalence) with ME/CFS are based largely on those who have the code G93.3 for “post-viral fatigue”, which includes “benign myalgic encephalomyelitis”, recorded in their medical notes.
Unfortunately, there are several limitations of using this code to estimate the prevalence of ME/CFS, including:
- The code may not accurately capture “ME/CFS” – or at the very least, the terminology used is inconsistent.
- Late, under- and non-diagnosis of this code may also be a problem meaning that many people who have the disease do not (yet) have a diagnosis.
According to the authors of a new study in Norway – including Anne Kielland who has recently published on employment in people with ME/CFS, and Professor Leonard Jason who developed the Depaul symptom Questionnaire (DSQ), these limitations are assumed to be common, and to disproportionately impact ‘underserved groups’ – populations that receive less than adequate access to resources, services, or representation, often due to factors like ethnicity, sexuality, geographic location, or social and economic position (typically measured using income, education, and occupation), leading to disparities in health, education, and other areas.
Importantly, this means that there may be many more people living with ME/CFS than existing estimates suggest – especially in underserved groups,
To try and obtain a more representative estimate of the number of people with ME/CFS in Norway, Kielland, Jason and their team developed a new approach – using questions from the DSQ, to identify those who met the Canadian Consensus Criteria for ME/CFS amongst a group of people with fatigue.
The approach also assessed who:
- Met the Fukuda criteria for CFS
- Met the Institute of Medicine (IOM), now the National Academy of Medicine (NAM), criteria for ME/CFS
- Met the International Consensus Criteria for ME
- Did not meet any ME/CFS diagnostic criteria.
The researchers hoped this approach would:
- Produce unbiased estimates of ME/CFS in a group of participants who have fatigue.
- Distinguish study participants meeting diagnostic criteria for ME/CFS – notably the CCC, from those who did not.
- Evaluate the extent to which people meeting the different diagnostic criteria had been given a diagnosis of G93.3 by a health professional.
The research team also investigated the impact of several different factors – including ‘gender’, previous Epstein Barr virus infection, rapid onset of ME/CFS, and previous psychiatric diagnosis, on length to G93.3 diagnosis.
What did the research team do?
An unusual system of recruitment – known as respondent driven sampling, was used to identify participants who were over 14 years of age, had a ‘fatigue type illness’, and were too sick to work or study full time.
Respondent driven sampling is an approach in which one participant is asked to recruit a fixed number of new participants from their own personal network. Importantly, the researchers stated that this method is best suited to populations who are hard to capture in research – including people with ME/CFS; especially those with severe or very severe ME/CFS, and those from underserved groups who may or may not have a diagnosis of the disease.
Once information was collected from participants, a set of well-defined computerised instructions (algorithms) were used to identify people with fatigue who met each of the ME/CFS diagnostic criteria.
What did the study find?
660 people with fatigue completed the whole questionnaire and consented to take part in the study; 75% (494 people) had a G93.3 diagnosis and 25% (166) did not.
Of these:
- 89% (584) people met the Fukuda criteria for CFS
- 85% (558) people met the IoM for ME/CFS
- 72% (474) met the CCC for ME/CFS
- 56% (394) met the ICC for ME
The researchers highlight that in this study, 26% of those with a G93.3 diagnosis did not meet the CCC for ME/CFS.
The study also identified that:
- Previously having a psychiatric diagnosis meant participants were less likely to have a G93.3 diagnosis.
- Previous EBV infection and rapid onset of ME/CFS symptoms meant participants were more likely to have a G93.3 diagnosis.
- Those with lower levels of education were less likely to have a G93.3 diagnosis.
- Younger age was associated with an increased chance of having G93.3 diagnosis.
- Interestingly, “gender” was not associated with chance of having G93.3 diagnosis.
What do the findings mean?
The researchers conclude that assuming the CCC and DSQ are accurate tools for identifying ME/CFS, their findings suggest considerable underdiagnosis of ME/CFS using G93.3 in Norway.
Notably, a quarter of those with a diagnosis of G93.3 did not meet the CCC in this study – this was similar to previous research findings which identified that 27% of G93.3 cases did not meet the CCC.
Of particular interest are the results which suggest that in this Norwegian population, there is no difference between men and women in the likelihood of having a G93.3 diagnosis. The researchers suggest that this finding indicates that the difference in disease prevalence between men and women reflects sex differences in the way the body functions – rather than differences in, and interactions between human societies (sociological differences).
The researchers also point out that a higher rate of G93.3 diagnosis in younger participants may be a reflection that the older participants faced a more difficult diagnostic process when they first presented with symptoms.
Overall, the researchers conclude that there are 5 key implications of the study:
- It is likely that assessing the prevalence of ME/CFS based on population health register data for G93.3 will produce significant underestimates as many people who meet the diagnostic criteria for the disease have not (yet) received a diagnosis for the disease – especially in underserved groups.
- The results highlight health care system inefficiency for people with ME/CFS. The researchers state that erroneous and late diagnosis may not only waste public resources, but also lead to misguided and potentially harmful recommendations.
- Clear evidence of the association between higher levels of education and an increased likelihood of having a G93.3. diagnosis.
- Recommendations for practice; a multidisciplinary approach could ensure “better patient assessment, lead to better treatment over time, and ultimately identify a biomarker”
- Recommendation that researchers must focus on underserved groups; the study team note that unorthodox sampling approaches such as respondent driven sampling may be especially useful here as traditional research methods are often less accessible to these groups.
ME Research UK notes that in this study, no information was provided on the ethnicity of participants. As research has shown that there may be “racial disparities in the diagnostic process for ME/CFS”, it is essential that different ethnic groups are also included in research. The impact of intersectionality –“a metaphor for understanding the ways that multiple forms of inequality or disadvantage sometimes compound themselves and create obstacles that often are not understood among conventional ways of thinking” should also be a central concept considered by ME/CFS researchers.
