Skip to content
Lies, Damned Lies, and Statistics

Anti-Coagulants Misattributed as Reasons for Decreased Mortality Rate in COVID-19 Patients

The headline claim is that among patients who were on ventilators, mortality was 29.1% for those given anti-coagulants.

Radoslav Zilinsky/GettyImages

“Blood thinners show promise for boosting the survival chances of the sickest covid patients,” reads a Washington Post headline from May 2020. The basis for this claim was a May 2020 article in which the researchers had analyzed data on 2,773 COVID patients at the Mount Sinai Health System in New York City. The headline claim is that among patients who were on ventilators, mortality was 62.7% for patients with no anti-coagulation drugs but a mere 29.1% for those given anti-coagulants. Sounds like a winner, right?

But there’s a hitch: The overall mortality rate for the two groups was basically equal at 22.8% and 22.5% respectively. How then was it possible to find a mortality benefit for the folks put on ventilators?

Here’s how: Patients who got anti-coagulants were far more likely to be put on a ventilator in the first place (29.8% versus 8.1%)!

Now, there is no data in the study on why doctors chose anti-coagulants for some patients but not others — we don’t know whether the anti-coagulants caused more people to be put on ventilators in the first place, or whether doctors chose to give additional medicines to the healthiest people on ventilators whereas others died too quickly. Either way, there is no basis for saying that anti-coagulants cause better survival.


Arnold Ventures funds projects to understand problems and identify policy solutions.

Map of the U.S. made of Arnold Ventures icons