Thread by @leticiakawano

1/ Multicentric chinese randomized open label trial on HCQ + SoC vs SoC for mild-mod #COVID19
1:1 randomization - stratification on dz severity
Intervention: loading dose of 1200mg/d for 3d HCQ > 800mg/d for 11d.
bmj.com/content/369/bm…
@bmj_latest

2/
The primary outcome was changed throughout the study conduction:
Final 1ary outcome: negative SARS-CoV PCR conversion by 28d AND, if severe #covid19, whether clinical improvement had occurred by 28d #BadChoice ?

3/ That seems to have been "the typical writing the protocol while doing the study"? #ThisIsAProblem

Sample size started off as 200 pts, without any calculation behind it, then on Feb 10th, researchers got the chance to do some math, and the sample size was updated to 360 🤔

4/ But the study is published w 150 pts... #CanSomeonePleaseExplain?

5/ no difference was observed in their primary outcome: time to negative seroconversion

6/ Conclusion: equipoise (genuine uncertainty) still persists:

The study was under-powered, the primary outcome is a non-clinical relevant outcome, many alterations in the protocol while doing the study...

Let´s wait for more robust studies results #CoalitionStudies and others