First of all, the authors did not actually do any meta-analysis (yet). The abstract does not make this very clear, but each of the reported outcomes is based on only 1 study (n_trials = 1).
www.bmj.com/content/387/... open-access
No drug, diet, or device intervention has supportive evidence for efficacy. CBT, rehab provided some relief of symptoms (moderate certainty).
We're still in desperate need for a validated treatment!
As ever, the devil is in the detail
TL; DR the authors' own risk-of-bias analyses show that their own conclusion is unwarranted
(Too bad they hid the crucial deets in an online supplement!)
cc #pwME #MECFS