Identifying commonalities and differences between EHR representations of PASC and ME/CFS in the RECOVER EHR cohort, 2025, Powers et al

Abstract:
Background
Shared symptoms and biological abnormalities between post-acute sequelae of SARS-CoV-2 infection (PASC) and myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) could suggest common pathophysiological bases and would support coordinated treatment efforts. Empirical studies comparing these syndromes are needed to better understand their commonalities and differences.

Methods
We analyzed electronic health record data from 6.5 million adult patients from the National COVID Cohort Collaborative. PASC and ME/CFS diagnostic groups were defined based on recorded diagnoses, and other recorded conditions within the two groups were used to train separate machine learning-driven computable phenotypes (CPs). The most predictive conditions for each CP were examined and compared, and the overlap of patients labeled by each CP was examined. Condition records from the diagnostic groups were also used to statistically derive condition clusters. Rates of subphenotypes based on these clusters were compared between PASC and ME/CFS groups.

Results
Approximately half of patients labeled by one CP are also labeled by the other. Dyspnea, fatigue, and cognitive impairment are the most-predictive conditions shared by both CPs, whereas other most-predictive conditions are specific to one CP. Recorded conditions separate into cardiopulmonary, neurological, and comorbidity clusters, with the cardiopulmonary cluster showing partial specificity for the PASC groups.

Conclusions
Data-driven approaches indicate substantial overlap in the condition records associated with PASC and ME/CFS diagnoses. Nevertheless, cardiopulmonary conditions are somewhat more commonly associated with PASC diagnosis, whereas other conditions, such as pain and sleep disturbances, are more associated with ME/CFS diagnosis. These findings suggest that symptom management approaches to these illnesses could overlap.

Link | PDF (Nature Communications Medicine, April 2025, open access)

 

ME/CFS was defined like this:

Figure shows the number of ME/CFS and Long Covid cases over time in the health records:

 

The ME/CFS cases do not seem to have increased much or am I reading the figures incorrectly?

The rest of the paper focuses mostly on their machine learning algorithm that tries to model ME/CFS and compare it to Long Covid.

 

Data came from the National COVID Cohort Collaborative (N3C), a collection of EHR data from 83 health partners representing 21.7 million patients.

But there were only 5781 cases in this dataset. That's like a prevalence of 0.027% which seems very low. In comparison, the R53.82 group (unspecified chronic fatigue) had 146,424 cases.

 

I know that I’m labeled as unspecified chronic fatigue despite meeting ME/CFS criteria and have been since before the pandemic.

I suspect that the vast majority of doctors here are actually incorrectly coding ME/CFS as “chronic fatigue.” I actually don’t know anyone among my friends in the US who has ME/CFS as their actual diagnosis in their medical records (it’s come up in conversation before).

 

The angle of the slope of the prevalence rate (b) almost doubles in 2021 Q3 from ~10 to ~20 degrees.

 

It might be because the G93.32 code was introduced only a couple years ago. I think the n=146,424 is people who have both "R53.82 (chronic fatigue, unspecified in the ICD-10-CM vocabulary)" and "52702003 (chronic fatigue syndrome in the Systematized Nomenclature of Medicine-Clinical Terminology [SNOMED CT] vocabulary)". The chart seems to only include the G93.32 definition.

 

Thanks for pointing that out, was indeed misinterpreting the graphs.

Given that the PASC-ME/CFS cases seem to be similar to other ME/CFS cases, it indeed suggests a doubling of new cases in that period because of Long Covid. The PASC-ME/CFS cases, however, already seems to decrease from the third quarter of 2022. If the effect was mostly temporarily it might not be easy to estimate the effect it had on total ME/CFS prevalence.

 

My suspicion is that as “Long COVID” became more prevalent in the global lexicon, more doctors were just satisfied with only an LC label. There would be no need to add another diagnostic label that would describe the full range of the patient’s symptoms if, in their mind, LC already covered it.

added: there would be no additional treatments or insurance reasons to specify ME/CFS anyways over LC alone

Funny enough, I had a telehealth appointment with my doctor in late 2022 where she asked how I was managing my long COVID symptoms now that I was taking classes remotely again. I had to remind her that my illness started in 2018 and I never caught COVID.

 

There's also the added problem of ME/CFS being a 'label' that negatively impacts a person's experience with health care, even for unrelated problems. It's quite common for physicians to not diagnose ME/CFS because they don't believe in it or have been discouraged to, but also many who know that it's basically a "don't bother with this patient" mark that can actually harm them because of most of their colleagues' attitudes towards it.

And while we have this, we also have physicians who won't care, who understand that things won't change until reality is properly recorded, which is how it should always be, but there is no way to know how much of which adds up to the final numbers.

All problems that come with a system that has encouraged decades of mislabeling. It's not even possible to go back to records and unwind this. It's basically equivalent to straight up deleting all historical data, as far as making it unusable goes. In many circumstances this would amount to criminal sabotage, but health care doesn't really do accountability and this is standard operating procedure so here we are.