For weeks, we have URGED CAUTION about using medications not proven to benefit patients with covid-19.

Hydroxychloroquine (HC) was touted by the US President.

Use skyrocketed.

New data out suggests the drug likely HARMS patients.

A thread by editor-in-chief @jeremyfaust.

1/

Anecdotes are dangerous. "I did X, and Y happened" is not science. It's the opposite of science. It's random observation. If that worked, we wouldn't need science.

Many said they'd "seen HC" work.

Had they? Did they notice when it did the opposite?

https://www.medrxiv.org/content/10.1101/2020.04.16.20065920v1.full.pdf

This new study looked at veterans being treated in our VA hospitals. The heroes of our nation.

Patients got either HC, HC plus azithromycin, or standard care.

Note: this was NOT a randomized study. The patients got what doctors felt was warranted.

The researchers then looked back to see how everyone did.

The patients who got HC died more.

Death in patients getting :
HC: 27.%
HC+azithro: 22%
No HC: 11.4%

p=0.003

So 10.6% to 16.4% MORE patients given HC died than those who did not. Is this real?

Very possibly.

On one hand, the patients who received HC were indeed a bit sicker than the ones who didn't.

Docs saw the sicker patients and said "uh oh, I better try something"

But does the somewhat worse illness in the patients who got HC explain the mortality difference?

No. It does not appear so.

The researchers used statistical methods to assess that.

Researchers found that even after adjusting for how sick the patients were, the patients who got HC...

DID WORSE.

DIED MORE.

Ugh.

Here's what the researchers say...

the "increased risk of overall mortality in the hydroxychloroquine-only group persisted after adjusting for the propensity [baseline level of illness] of being treated with the drug."

"There was no increased risk of ventilation in the hydroxychloroquine-only group"

Meaning...?

The researchers believe increased mortality in the HC group might be because of "drug effects on or dysfunction in non-respiratory vital organ systems."

They then note that "hydroxychloroquine use in Covid-19 patients has been associated with cardiac toxicity."

Sound familiar?

It should.

We've discussed in the brief ( http://brief19.com ) and elsewhere about the cardiac effects of HC, especially in higher doses.

This is a real thing. "Torsades de pointes" is a lift-threatening cardiac problem.

Let's be clear what the take home message here is...

Giving drrugs that are unproven to help a disease "just in case" is NOT A SAFE PLAN.

A placebo IS ACTUALLY safer than a dangerous drug.

Therefore "depriving patients of a study drug" is indeed ethical--because the study drug could be hurting patients.

We must detect that.

This is yet another reminder that we must be humble with WHAT WE THINK WE KNOW.

"Hydroxychloroquine has been reported to inhibit SARS-CoV-2 replication"... IN THE LAB.

We know this.

But does that make it effective and safe IN HUMANS with covid-19?

Not one bit.

A randomized controlled trial would give us a more clear answer whether HC works for ANYONE.

But it may not even be ethical to do this now, given these data.

Does HC help some? Unclear. Doesn't look good.
Does HC hurt some? Absolutely appears to.

We can't be sure.

Routine use of hydroxychloroquine (or any untested medication) is roulette.

It can hurt people.

Giving it out en masse degrades science, and our ability to get answers

So....

Join us in demanding science.

Better studies.

This one is a start.

The methods are at least careful; though observational and "retrospective."

Several hundred patients (a good start).

But tons of good data to parse through.

Not good news. But we need to know this.

For more, follow us here, visit http://Brief19.com  for daily emails, and find us in 10-15 languages on @Brief19_itl.

Stay safe, everyone. -Jeremy Faust. ( @jeremyfaust)
/fin