Some important context about the COVID-19 hospitalization numbers we're seeing from NC DHHS.

https://www.ncdhhs.gov/divisions/public-health/covid19/covid-19-nc-case-count

The state's been publishing these numbers daily since March 27.

Since then, we've seen hospitalizations drop three times, the largest of which was from Friday, through the weekend, and today.

In that time, hospitalizations dropped more than one-quarter from 423 to 313.

But, caution.

This number is based on a daily survey sent by DHHS to hospitals across NC.

Not every hospital responds every day, and the state reports the response rate as a percentage here (That's also why you see the total bed number fluctuate up/down over time): https://www.ncdhhs.gov/divisions/public-health/covid19/covid-19-nc-case-count#by-reporting-hospitals

Response rates since Friday:
4/10 -> 87%
4/11 -> 78%
4/12 -> 69%
4/13 -> 66%

So even though we saw a big drop in hospitalizations, we don't know if that's because there are fewer COVID-19 patients in hospitals or if the hospitals just aren't reporting them.

We also don't know *which* hospitals aren't reporting, since NC DHHS isn't telling us that.

Are they the big ones? Small ones? Those serving a massive population? A tiny one?

That makes it tough to know how big the gap in our data could be.

All drops in hospitalizations so far have occurred during drops in response rates:

4/1-4/2
10% drop in patients
7 pct pt drop in response rate

4/4-4/5
4% drop in patients
23 pct pt drop in response rate

4-10-4/13
26% drop in patients
21 pct pt drop in response rate

But not every drop in response rate translates to drop in reported hospitalizations.

3/28-3/30
57% increase in patients
17 pctpt drop in response rate

4/1-4/4
33% increase in patients
9 pctpt drop in response rate

4/8-4/9
3% increase in patients
5 pctpt drop in response rate