Patient-reported treatment outcomes in ME/CFS and long COVID.
Eckey, Martha, Li, Peng, Morrison, Braxton et al. · Proceedings of the National Academy of Sciences of the United States of America · 2025 · DOI
Quick Summary
This study asked nearly 4,000 ME/CFS and long COVID patients about their experiences with over 150 different treatments to see which ones helped them feel better. The researchers found that ME/CFS and long COVID are very similar illnesses with overlapping symptoms and that patients with each condition responded similarly to treatments. The study also discovered that different patient subgroups—like those with heart rate and blood pressure problems, or those with thinking and memory problems—benefited most from specific types of treatments.
Why It Matters
With no FDA-approved treatments for ME/CFS or long COVID, this real-world patient data provides crucial evidence for identifying which management strategies may help specific patient subgroups and informs the design of future clinical trials. Understanding that different symptom patterns respond to different treatments could help clinicians provide more personalized care and researchers prioritize which therapies warrant rigorous testing.
Observed Findings
89.7% of ME/CFS patients and 79.4% of long COVID patients reported postexertional malaise (PEM).
Treatment response patterns between ME/CFS and long COVID patients were significantly correlated (R² = 0.68).
Patients with POTS-dominant symptom profiles showed preferential response to autonomic modulators.
Patients with cognitive dysfunction-dominant profiles showed preferential response to CNS stimulants.
Over 150 different treatments were evaluated across diverse patient demographics and comorbidity profiles.
Inferred Conclusions
ME/CFS and long COVID share substantial symptomatic and therapeutic similarities despite different origins, suggesting common underlying pathophysiological mechanisms.
Patient heterogeneity in symptom presentation predicts differential treatment responses, supporting a precision medicine approach to management.
In the absence of approved pharmacological treatments, patient-reported outcome data can identify promising therapeutic targets for future clinical trials.
Remaining Questions
Which of the treatments identified as beneficial by patients would demonstrate efficacy in rigorous, placebo-controlled clinical trials?
What are the underlying biological mechanisms linking specific symptom clusters (e.g., POTS-dominant, cognitive dysfunction) to treatment responses?
What This Study Does Not Prove
This study does not prove that any specific treatment is effective—it only documents what patients perceived as helpful. The findings are based on patient memory and subjective reporting without clinical measurements, placebo controls, or verification of whether treatments truly caused improvements versus coincidental symptom changes. The correlation between ME/CFS and long COVID treatment responses does not establish whether they share identical underlying mechanisms.
How do long-term outcomes compare between patients who used treatments perceived as most beneficial versus standard care approaches?
Can the identified patient subtypes be validated using objective biomarkers, and do they represent truly distinct disease endotypes or a spectrum of manifestations?