Thread by @leticiakawano

1/ Some more information on our study Coalizao 1, on HCQ/Azithro for #Covid19 that explains some issues:

@f_g_zampieri @damiani_petri @tlisboa79 @azevedo_lcp @FlaviaSepsis @a_serpaneto
Before I start, the first thread on the study is here:

https://twitter.com/leticiakawano/status/1286379221896630272

2/ Back in March, we knew much less than we know today on Covid19 (and as a consequence v. little information to inform RCT design)
Regulatory agencies & WHO have defined the ordinal scale as an important outcome.

As @f_g_zampieri commented (in Portuguese) elsewhere

3/ We realized, after the trial started, that a clinical status ordinal scale would be potentially problematic given that the level of support offered could differ between countries and local practices not reflecting necessarily dz severity

4/ Before the first DMC analysis (completely blinded to the data) our intention was to change our primary outcome to days alive and free of oxygen support at 15 days (DAFOR15), but the Brazilian regulatory agency did not allows us - this is all in the Suppl.

5/ Note that this change in outcome is a consequence of "building the airplane while we fly it", and not bc we have spotted anything midway on our data. Remdesivir trial also changed its 1ary outcome.
DAFOR15 would have been more clinically meaningful. And we were powered...

6/ with DAFOR15 as 1ary outcome, we would be powered to detect small differences between groups (for ex. 1.5 day)...but that was kept as a secondary outcome:
[well, that is part of the game in clinical trials]

7/ Bottom line: Coalizao I brings important information to the table:
1. Results are consistent w different RCTs in diff. populations and different medical practices
2. The 2ary outcomes support the suggested direction of effect of our 1ary outcome
3. Side effects w HCQ/Azithro