Projections vs Reality: How Models for Coronavirus Hospitalizations Presented to Lawmakers in March Compare to Reality

RALEIGH – In the first tranche of Pandemic Panic, everything was about the models. Models for deaths; models for cases; models for hospitalizations; models for spring breakers returning home; model projections for everything. And, they were all scary.

The infamous Imperial College model that projected more than 2 million deaths in the United States has been cited as the impetus for a wave of lockdown policies. In North Carolina, at the end of March, state lawmakers were also treated to models from the N.C. Medical Society that presented a few different scenarios for the impact of the state in terms of hospitalizations.

After all, ‘Flattening the Curve’ to avoid overwhelming our hospitals was the sole justification offered for stay-at-home orders and the long list of unprecedented restrictions, some of which are still ongoing. So how does our actual experience over the last 6 months compared to the models?

Jon Sanders, an economist and Director of Regulatory Studies at the John Locke Foundation, has done the work to insert actual hospitalization data through September 10 into those same charts presented to the General Assembly back in March projecting the fear scenarios that fueled Pandemic Panic. The juxtaposition is remarkable (you might need a magnifying glass):

From Sanders on the Lockerroom Blog:

“COVID-19 hospitalizations continue to rise as NC corrects technical issue with reporting system,” blared WTVD on Twitter earlier today. Worrisome, isn’t it?

This is an odd headline considering the fact that daily COVID-19 hospitalizations have been below 1,000 for 15 days in a row. In fact, you’ve probably never even heard that hospitalizations have been trending downward the past few weeks. Media aren’t going to tell you; doing so wouldn’t keep you properly terrified.

But as I wrote on June 19:

Unless you are in possession of a memory, you forgot that the best-case scenario was that we’d have a small peak of hospitalizations not in April, not even in May, but in July.

That best-case scenario was presented by representatives of the North Carolina Medical Society (NCMS) to the General Assembly’s House Select Committee on COVID-19 on March 26, the day before Gov. Roy Cooper dropped his stay-at-home executive order. The committee heard testimony from NCMA representatives, who presented the General Assembly with a report on “COVID-19: Analysis and Response.”

In her testimony, NCMS board member and Duke University professor Dr. Eileen Raynor highlighted a graph from that report with three projected hospitalization curves through July 6.

I have updated the NCMS graph to reflect the actual hospitalizations data (7-day rolling average) through today, September 10. Click here for the full size of this chart […]

Because it’s very difficult to see NC’s actual hospitalizations curve on that chart, owing to the enormous scale of its y-axis, I’ve isolated it with a y-axis based on the NC Medical Society’s estimated July 6 peak of around 12,000 hospitalizations. It’s the chart atop this post.

The peak 7-day rolling average of hospitalizations occurred on July 31 at 1,177.6 — about one-tenth of the NC Medical Society’s best-case scenario. The single-day peak was 1,236 on July 29.”

The actual hospitalization rate was ONE-TENTH of the BEST CASE SCENARIO.

And here we are, with businesses still forced closed, and the governor rejecting the requests of parents to attend their college athlete child’s football game in an outdoor, 50,000 seat stadium.