HOUSTON – From the number of infections to total deaths, there’s no shortage of COVID-19 models that are trying to predict what’s going to happen with the coronavirus.
But what's the right way to look at the numbers and which models can you depend on?
KPRC 2 Investigative Reporter Joel Eisenbaum put these models through the Trust Index.
But what dangers come with all of this? Who really has that answer?
“We did so well for so long,” said Dr. Peter Hotez of Baylor College of Medicine. Hotez added, “I’m quite worried we’re going to start seeing an uptick in cases.”
Certainly, that’s the general consensus. The closer we get together, the riskier it is.
Experts by the dozens are citing various models that give varying predictions.
An internal White House study supposedly predicts 3,000 deaths and 200,000 new cases a day by June 1 in the U.S.
But is that information accurate and should we base our social distancing policies on it?
"I don't think that was ever meant for public consumption so I would basically discount it," said Hotez.
Hotez says his favorite model is the one from the Institute for Health Metrics and Evaluation at the University of Washington’s School of Medicine. Hotez likes that the data sets are transparent and can be verified, but it’s also ever-shifting.
On April 1, total deaths predicted by August 4 in Texas were 6,392. By April 7, the predicted deaths were dramatically down to 2,025. Now, a month later, with stay-home orders evaporating, they’re back up to 3,632. Presumably, the longer we go, the more trustworthy the prediction.
The COVID-19 modeling gig isn’t just for the professionals.
There are also all manner of amateur modelers, like Judith Oppenheim. She's a retired engineer who went to Rice University and lives in Friendswood.
For Oppenheim, it’s a hobby. She appears to be diligent and she reliably posts her updated modeling and predictions weekly in a local Facebook group. It’s chock-full of semi-pro soothsayers.
Another hobby in her retirement, she's crafty and she makes her own face coverings.
Oppenheim's conclusion is it's clearly too early to get out and about because we simply do not know enough. The Texas data set isn't big enough yet, she says.
"We need more of a validation, more of a predictive capability so that we can say we're under control and we can go to the next stage," said Oppenheim.
Trust Index: Be Careful
So can you trust the COVID-19 prediction models? Certainly, it depends on the model. But even the one Hotez likes for Texas has swung significantly since the pandemic started. So we rate this “be careful” on the Trust Index.