LogFAQs > #955843929

DarkRoast posted...

Given that the evidence for dexamethasone is much stronger, and the drug itself is much cheaper, I can't think of a rational reason to prescribe Ivermectin instead.

Like I said, I'm not saying there's good evidence to suggest clinical use here. Just that the fact that only two studies show a strong positive effect doesn't necessarily say anything important for the detection of any effect at all. All and all seems consistent with a weak positive effect that would require far more samples to really detect.

Never really been a fan of meta-analysis generally anyway. Seems too unreasonable to collapse data across a myriad of known and unknown differences. Sometimes they can be useful for documenting potential moderators that account for differences across the literature.

---

=E[(x-E[x])(y-E[y])]