No, the CDC did not drop its COVID-19 death count to 37,000. In fact, it didn't 'revise' it at all.

I'm losing count of how many times in the past few days I've seen someone post something along the lines of this tweet:

"The CDC has actually ADMITTED that they overcounted COVID-19 deaths!"

"Look at the numbers—they're right there on the CDC website plain as day!"

"See, it's all overblown! We did this whole shutdown thing and tanked the economy for nothing!"

First of all, no, the CDC did not revise anything. Let's dive into these numbers because they actually are a bit confusing when you don't read the whole page (and frankly, some parts are a little confusing even if you do—get it together, CDC).

There are different methods of counting COVID-19 deaths, and the CDC's website includes numbers for two very different methods. We have:

1) The official CDC death count, which you can find on the CDC's home page. This count comes directly from public health departments in each state and territory daily. As of the writing of this article, that count stands at 68,279.

2) The Provisional Death Count, which is where that ~37,000 number comes from. This count comes from the National Vital Statistics System—the system that processes and logs death certificates. The notable thing about the Provisional Death Count is that it's not up-to-date. The CDC site itself states that the numbers on the Provisional Death chart lag weeks behind other counts:

"It is important to note that it can take several weeks for death records to be submitted to National Center for Health Statistics (NCHS), processed, coded, and tabulated. Therefore, the data shown on this page may be incomplete, and will likely not include all deaths that occurred during a given time period, especially for the more recent time periods. Death counts for earlier weeks are continually revised and may increase or decrease as new and updated death certificate data are received from the states by NCHS. COVID-19 death counts shown here may differ from other published sources, as data currently are lagged by an average of 1–2 weeks."

Here's a real-world example of what this looks like:

This is a screenshot of the Provisional Death Count as of April 16, 2020 (which you can access at this CDC link). As you can see, the COVID death count for the week of 4/11/20 was 3,542.

And here is the Provisional Death Count as of the writing of this article, which you can view in real time at this CDC link. As you can see, the week of 4/11/20 has been updated from 3,542 deaths to 12,628—a nearly four-fold increase since the April 16 publication.

When the numbers were published on 4/16/20, there were still 9,086 death certificates that hadn't been processed yet from the week prior—that's what they mean by a lag. Three weeks later, the numbers are very different.

So that 37,000 total (well, 39,000 right now) will change as the death certificates get processed. The Provisional Death Count simply isn't accurate yet. And the lag means it will never be an up-to-date count, so it's not a reliable source for current death numbers.

The problem is that people have been sharing the not-up-to-date Provisional Death Count link as a way to make it sound like the COVID-19 death numbers are actually smaller. They are not.

It's worth noting that all COVID-19 death counts include both lab-confirmed and "presumed" COVID-19 deaths. This has also been a source of confusion, not to mention conspiracy. But "presumed" doesn't mean just a wild guess.

Test results for coronavirus have a high false negative rate—from 5% to 30%—according to Dr. Alan Wells, professor of pathology at University of Pittsburgh. So relying solely on positive lab test results for COVID deaths would miss thousands. At this point, doctors and medical examiners can generally recognize clear COVID symptoms in a critically ill or deceased patient, and if a patient meets the clinical, epidemiological, or vital records criteria for the COVID being the cause of death, that's considered "presumed."

Each state has different requirements for coding COVID-19 deaths, and it's generally a very small percentage that are counted as "presumed."

Adding to the confusion on this front, Dr. Birx, from the White House Coronavirus Task Force, said that the U.S. was taking a "liberal" approach to counting COVID-19 deaths, and ""The intent is, right now, that . . . if someone dies with COVID-19, we are counting that as a COVID-19 death."

People unfortunately did not take that statement in the context of underlying conditions, which is what Dr. Birx was talking about. Here's what she actually said:

"There are other countries that if you had a pre-existing condition and let's say the virus caused you to go to the ICU and then have a heart or kidney problem -- some countries are recording that as a heart issue or a kidney issue and not a COVID-19 death. Right now ... if someone dies with COVID-19 we are counting that as a COVID-19 death."

If a person has a heart condition and they get sick with COVID-19 and die, COVID is counted as a cause of their death, even if they died of a heart attack—the reasonable assumption being that the disease led the patient's weakened heart to give in. Dr. Birx did not mean that a gunshot victim or a fatal car accident victim would be certified as a COVID-19 death just because they tested positive for the disease. That would be silly, not to mention illegal.

Read more on how COVID-19 deaths are counted from a forensic pathologist here.

You can also see an email from the Louisiana Health Department specifying how doctors are to log coronavirus deaths here:

So, no, COVID-19 death counts have not been revised downward, nor are they artificially inflated. In fact, it's more likely that they've been undercounted than overcounted, since only deaths that had been confirmed by tests were being counted for at least the first month of the outbreak in the U.S.

More importantly, read the fine print on a website before you make any assumptions about what you're seeing. Health data tracking can be a confusing to dive into under normal circumstances, much less during a novel virus pandemic where we're all learning as we go.