The study is way underpowered (too small a sample). Based on the raw numbers, the HCQ group had less deaths and less ICU admissions, but the numbers we're talk

The study is way underpowered (too small a sample). Based on the raw numbers, the HCQ group had less deaths and less ICU admissions, but the numbers we're talking about are 3 vs. 4 deaths and 16 vs. 21 ICU admissions (group sizes were different). These numbers simply aren't enough to make a determination one way or the other. On mortality, the 95% confidence interval allows for a range of HCQ reducing death rates by 87% up to it increasing death rates by 2.9x.