Thread by @drabbyphilips

1/ Ivermectin+doxycycline for COVID-19 symptoms: a randomized trial
journals.sagepub.com/doi/10.1177/03…
New study included to ivermectin/doxycycline bandwagon

Upfront, luks legit - reported outcomes, as expected favours IVM+DOX in reducing progression to serious disease

But lets dig in

https://twitter.com/Fionaroha/status/1393208745581170694

2/ Patients included are mild-moderate

The title says - 'for treating Covid19 symptoms'

But primary outcome measure is clinical recovery

Secondary outcome is progression to mild/moderate/severe or death - mild or moderate is included, so ideally it should be severe/critical

3/ Primary measure should have been secondary and vice versa - but none of these fit the title.

Authors mention: We sought to enroll as many patients as possible in a short time. As a result, a priori power calculations were not performed - which means pre protocol sample size..

4/ ..was not calculated and study is possibly underpowered - this is a major flaw

To correct for this, authors did post-hoc (after study) sample/power estimation - for detecting 2 day difference in median 'recovery time' again - not clinical progression - this also weakens study

5/ Next look at patients included

Group assignment was not stratified according to disease severity, which means lesser severity patients could have been allocated to treatment category as inclusion is unclear

Clinical recovery is undefined

Moderate disease is included...

6/..
but authors excluded patients with hypoxia [oxygen saturation (SpO2) <90%] requiring supplemental oxygen

But moderate disease also includs oxygen requirement in hospital

Unsure if those requiring oxygen vs thos not requiring with moderate disease were allocated to placebo

7/
...becoz allocation according to severity was not clarified as per authors own statement

Now look at treatment offered

Paracetamol, antihistamines, cough suppressants, vitamins, oxygen therapy, low molecular weight heparin, antibiotics, remdesivir, other antiviral drugs!

8/
Which means, treatments received by the groups differed extremely between groups.

Hence a sub group analysis with respect to treatment heterogeinity was warranted which was not done in this study

Another major flaw

9/
So this study LIKE ALL other IVM studies - have inherent high risk of bias and if included in a pooled analysis would contribute to very low quality evidence.

Hence, no proper evidence for use of IVM+DOX in Covid19 can be conferred from this very inattentively done RCT.

10/
Main lesson to learn here is - RCT doesnt mean 'gold standard' - one shud have basic analytical prowess 2 understnd quality of trial:

Methodology
Inclusion/exclusion commitment
Power/sample size calculation
Ideal/solid (not vague) endpoints

#ivermectin isnt useful in Covid

@threadreaderapp compile