Austin (KXAN) — As the coronavirus pandemic and response continues, researchers at the University of Texas at Austin have continued to churn out papers and update models in an effort to give policymakers an informed look at the risk the virus poses.
The latest projections for many places (including the state of Texas) based on current data suggest that there is a 64% probability Texas has already seen its peak in COVID-19 deaths and a 99% probability that the U.S. has already seen its peak in COVID-19 deaths. These numbers reflect significant changes from what these models have shown over the past few days and come with a big caveat: they cannot account for the possibility of a second wave of the virus.
This particular UT model tracks the impact of social distancing in different areas by using GPS records from tens of millions of cell phones. UT says the data is anonymized and helps to illustrate where people are spending their time. The researchers behind this include Professor Lauren Ancel Meyers who has led many papers and modeling efforts at UT related to COVID-19 and pandemics.
“The researchers found clear evidence that the timing and extent of people’s distancing behavior differ across states—and that these differences matter in terms of ‘flattening the curve,'” said a release from UT about this modeling on Friday. “In Texas, for example, many large cities issued stay-at-home orders that began to curtail mobility many days before a statewide policy was enacted.”
Friday, the UT COVID-19 Modeling Consortium and the Texas Advanced Computing Center released a model that offers live, daily estimates on fatalities from COVID-19 across the US in the coming weeks. UT has a website that shows these projections of COVID-19 deaths. A spokesperson for UT’s College of Natural Sciences explained that this website is updated live based on the data they are getting.
As of Friday, April 17, that model suggested the number of deaths for COVID-19 across the US had not peaked and is not likely to peak in many states until after May 1. As of Wednesday (five days later) the model is now showing that the probability that the peak of COVID-19 deaths in the United States has already happened is 99%.
As of Wednesday morning, the model for Texas showed a 64% probability that the peak of COVID-19 deaths has already passed. The model Wednesday also showed an 89% chance that the peak will have passed in seven days and a 96% probability that the peak will have passed in fourteen days.
This projection reflects a shift in the model from earlier on in the week. On Monday (two days ago) the model showed the odds were below 50% that Texas had seen a peak in COVID-19 deaths, a spokesperson for UT’s College of Natural Sciences said.
A screenshot of the UT COVID-19 modeling web page for COVID-19 deaths in Texas. Data snapshot taken 10:30 a.m. April 22, 2020.
This model is projecting that by May 12 in Texas, the state will have seen cumulative number of 863 COVID-19 deaths, with a minimum projection of 679 deaths and a maximum projection of to 1370 deaths by that date.
Why offer minimum and maximum projections of deaths in this way?
The UT researchers recognize there is uncertainty in these numbers and want their models to show the range of uncertainty possible.
The numbers can only reflect confirmed COVID-19 deaths and the models assume that people will continue social distancing to the same extent they have been. This UT model was only designed to forecast a single wave of COVID-19 and can’t predict the possibility or characteristics of a possible second wave of the virus.
A visualization by UT Austin to show the uncertainty in the outcomes for projected COVID-19 deaths in the state of Texas. Data snapshot taken 10:30 a.m. April 22, 2020.
The UT researchers explained that they define a peak as the day when their model’s prediction for the average daily death stops increasing and starts decreasing instead.
UT Austin noted these findings are in contrast to one of the most frequently cited models, the University of Washington’s Institute for Health Metrics and Evaluation (IHME) COVID-19 projections. The UT model uses local movement data from individual U.S. states as opposed to looking at patterns from other countries. UT also says that this model “accounts for greater uncertainty further in the future” than the IHME model, noting that updates to the IHME model led to significant changes to the projected numbers of COVID-19 deaths.
“While more uncertain forecasts may be disconcerting, we believe that they reflect the true range of possibilities that could unfold in the weeks ahead,” said UT Austin Professor James Scott who helped develop this model. “Our model stands on the shoulders of the IHME model, but it corrects critical statistical flaws that led the IHME model to make many projections that, in retrospect, have turned out to be far too optimistic.”
UT says this new model is being provided to the U.S. Centers for Disease Control and Prevention as well as to the White House Coronavirus Task Force. This UT model (as with all the other papers and updates during this time from the UT COVID-19 Modeling Project) is being presented before scientific peer review “due to the time-sensitive nature of the subject.”
Has Austin seen a peak of COVID-19 deaths yet?
As for a peak in cases in Austin-Travis County, Austin’s Interim Health Authority Dr. Mark Escott said Tuesday that he believes it’s still too early to tell how close Austin-Travis County may be to a COVID-19 peak, but he hoped to have more clarity by the end of the week. Escott and other regional leaders are working with UT Austin and the modeling from Meyers and her colleagues to make policy decisions.
“We as a community will determine when that peak happens and when another peak happens in the future,” Escott said, citing the need for continued social distancing and use of healthy habits such as wearing face coverings.
“If we don’t continue to do that, we are going to back at a ‘stay-at-home, work safe’ order again in a short amount of time” Escott said.