RE: https://disabled.social/@tomkindlon/116717085328668673
This study used a self-report for primary outcome, and found
Some secondary outcomes did have high posterior probabilities of efficacy, including 0.98 for symptom burden on day 120 and 0.96 for clinician diagnosis of long COVID on day 180.
So I think the real result from this study is "the study was underpowered" and not "metformin doesn't work for LC" because their results also cannot be distinguished from "metformin during acute COVID-19 reduces Long COVID rate by a quarter"
People who got Metformin:
- were 24% less likely to self-report Long COVID at 180 days
- were 22% less likely to be clinically diagnosed with Long COVID at 180 days
and consider their conclusion:
Metformin did not exceed the primary outcome threshold for declaring efficacy for reducing PASC. The posterior probability of efficacy for metformin was high for preventing a clinician diagnosis of long COVID and for symptom burden, consistent with prior clinical trial and observational data. No safety concerns occurred in this trial of metformin in low-risk adults. While a low event rate makes it challenging to evaluate therapeutics
https://dx.doi.org/10.1093/cid/ciag335
#COVID #COVID19 #SARSCoV2 #LongCOVID #Metformin #PASC