Arkansas has a low percentage of hospitalizations related to Covid-19 at around 7.5% of active cases. The CDC estimates that 20% of people who contract Covid-19 infections will be hospitalized (at the low end). The figure is made more mysterious by the fact that Arkansas deaths are in line with CDC estimates. In the last post, I noted a variety of different theories for the discrepancy and found one that I couldn’t debunk: potential unreliability in the data.

I wanted to know how hospitalizations were being reported, so I reached out to the Arkansas Department of Health. Here’s what I learned.

Question: At what point does a hospital consider a patient a Covid-19 patient for state reporting purposes?

Answer: Patients that appear on the Arkansas Department of Health COVID-19 Dashboard must have a laboratory-confirmed COVID-19 specimen.

Question: Does a patient require a positive Covid-19 test before they are reported?

Answer: Yes.

Question: Are presumed positive Covid-19 patients reported?

Answer: Many hospitals have a list of presumed positive cases that they monitor due to symptoms consistent with COVID-19, but these numbers are not reported on a regular basis and have not been requested at this point.

Question: How frequently are hospitals reporting Covid-19 hospitalizations?

Answer: Hospitals report data at least twice each day. In addition, ADH nurses conduct follow up contacting with positive COVID-19 patients.

Question: How are Covid-19 hospitalizations reported – online or through some other means?

Answer: They are reported via phone, email, fax, electronic lab reporting, and electronically through ADH’s internal software programs.

The key takeaway from these answers was that only patients that have a positive Covid-19 test, either before they are hospitalized or while they are hospitalized, are being reported.

Here are two things we know:

It has not been unusual for it to take a week or more for patients to receive test results – particularly if the patient isn’t a healthcare worker or first responder. We also know that the average hospital stay for Arkansans with Covid-19 is six days.

Based on these two points, it would seem likely that many patients have been hospitalized with Covid-19, tested, recovered, and released – all before their test results came back. And all without being considered a Covid-19 hospitalization for state reporting purposes or becoming an added number in the daily press briefings.

To learn more, I spoke in general terms to a few frontline hospital workers. They confirmed that many, if not a majority, of the patients they have treated for what they presumed was Covid-19 were in fact released before they ever knew if they had tested positive or negative. It is possible there are Covid-19 wards in hospitals housing several presumed positive patients with only a small percentage or none of them actually reportable according to the state standard.

What this tells us is that the length of time it takes for tests to return can provide an artificially low hospitalization rate. Dr. Nate Smith said in the governor’s press conference today that tests are beginning to come back more quickly. If that bears out and more patients are able to receive test results while they are still in the hospital, we can expect the percentage of hospitalizations to rise. That won’t be anything to panic over. Instead, it will be evidence of a testing system that is become more robust.

Disclaimer: This is a theory. There could be an equally or more valid explanation of which I am unaware. I do not think there is a government cover-up to keep us from learning how many hospitalizations there are. I believe the data is being reported honestly in a system that may have worked for routine illnesses, but is less useful in a pandemic with test backlogs.