Trust Index: Were models predicting COVID-19’s impact on Florida hospitals misleading?

JACKSONVILLE, Fla. – In a televised speech, Florida Gov. Ron DeSantis slammed a model that predicted how many people could have been hospitalized in the state by the end of April due to the novel coronavirus.

“They predicted 460,000. OK. Projections say the state could run out of ICU beds by April 14, very scary,” DeSantis said. “If you need to be in intensive care, you need certain types of treatment and there’s no beds available, what’s going to happen? Obviously that’s not going to be good. But what did happen, Florida’s ICU bed availability today, 36.5% of the ICU beds in the state of Florida sit empty.”

The projection DeSantis was referring to was published in mid-March by model Covid Act Now and reflected what the models said could happen if no actions were taken to slow the spread of the virus.

The same model predicted that with shelter in place order in place for three months, by July, there would be a little more than 18,129 hospitalizations.

Figures released by the state Thursday show that Florida is still well below that number, with about 5,500 people who have been hospitalized to date.

Later in his speech, DeSantis did acknowledge that actions to slow the spread in Florida were effective at changing the predicted outcomes of COVID-19. Actions like stay-at-home orders, bans on in-restaurant dining and visits to long-term-care facilities.

“The goal of reducing the curve was to keep the infections underneath that hospital capacity. That’s the reason why we went on this, what we’ve done, and you can see not only did we not reach hospital capacity, we’re far under what hospital capacity is in the state of Florida,” DeSantis said.

In a televised interview in March, Dr. Anthony Fauci said of the models predicting the impact of COVID-19: “Whenever the models come in, they give a worst-case scenario and a best-case scenario. Generally, the reality is somewhere in the middle. I’ve never seen a model of the diseases that I’ve dealt with where the worst case actually came out. They always overshoot.”

Covid Act Now has since changed its predictions for the outcomes of COVID-19, looking at what could happen based on current trends and what could happen if restrictions are lifted. The model often cited by the White House Coronavirus Task Force, the Institute for Health Metrics and Evaluation (IHME), has also changed over time.

Fauci said the following in late-March about why the IHME model has been revised over time.

“Every time we get more data, you feed it back in and relook at the model. Is the model really telling you what’s actually going on? And again, I know my modeling colleagues are going to not be happy with me, but models are as good as the assumptions you put into them, and as we get more data, then you put it in and that might change. So even though it says according to the model, which is a good model that we’re dealing with, this is full mitigation. As we get more data, as the weeks go by, that could be modified."

According to Fauci, models predict what could be the worst-case scenarios – not always what will actually happen.

Actions taken like social distancing or stay at home orders can result in very different outcomes, like much lower numbers, and the models changed as new data came in and gave closer estimates to what would happen.

Be Careful

After reviewing this topic, we've found some issues - Be Careful.

About the Author: