By Andrea Leinfelder, Staff writer
A popular model used to forecast deaths related to COVID-19 has readjusted its outlook for Texas, dramatically increasing the number of deaths projected for the state.
The Institute for Health Metrics and Evaluation at the University of Washington, which has been used by the White House, is now forecasting 3,632 deaths in Texas by Aug. 4. That’s up from 1,288 forecast on April 29.
Nationwide, IHME is forecasting 134,475 deaths, up from 72,433 forecast on April 29.
According to a news release, the new projections reflect rising mobility as social distancing measures are eased. Increased testing, more contact tracing and warmer weather (with more needing to be learned on the latter) are factors that could help slow the disease’s spread, but they don’t offset people leaving their homes and intermingling.
Previously from IHME: The model showed 6,000 would die in Texas. Now it’s 1K. What’s changed?
The projections released Monday came with a major overhaul to the IHME model. Its initial model released on March 26 to estimate hospital resource demand is now combined with a disease transmission model. The new hybrid model captures the impact of changes in social distancing mandates, mobility, testing and contact tracing. This model can predict a resurgence in the disease that might be caused as social distancing mandates are relaxed, according to the news release.
“This new model is the basis for the sobering new estimate of U.S. deaths,” IHME Director Dr. Christopher Murray said in the news release. “The model will allow for regular updating as new data are released on cases, hospitalizations, deaths, testing, and mobility. It can also be used to identify what may be the trajectory to progressively relax social distancing while still limiting the risk of large-scale resurgence.”
Models often change and adapt to new data and policies. The IHME model initially forecast around 6,000 deaths for Texas. Then it dropped to 4,000. It went down lower to around 1,000, but now it’s back up to 3,600.
And while models provide one specific forecast number, they also tend to have a range to indicate what else is possible. IHME, for instance, is forecasting 3,632 deaths in Texas by Aug. 4 but says the actual number of deaths could be as low as 1,470 or as high as 10,721.
Youyang Gu, a data scientist who graduated from Massachusetts Institute of Technology, created covid19-projections.com and is forecasting 3,985 deaths in Texas by Aug. 4, with a range from 1,527 to 8,409.
Houston coronavirus updates: What you need to know for May 5
Some other models listed on the Centers for Disease Control and Prevention’s website don’t look as far out. It is, after all, easier to forecast the weather tomorrow than next week.
Massachusetts Institute of Technology forecast 1,336 deaths in Texas by June 15. Los Alamos National Laboratory forecast about 1,900 deaths in Texas by June 14, with a range from 1,200 to 4,500 deaths. The University of Texas at Austin forecast 1,405 deaths in Texas by May 25, with a range from 1,144 to 2,041. For the Houston metropolitan area, it’s forecasting 420 deaths, with a minimum of 294 and maximum of 770.
“When our models project far into the future, the outputs are not forecasts but rather plausible scenarios,” Lauren Ancel Meyers, who heads the University of Texas COVID-19 Modeling Consortium, previously told the Chronicle. “The fate of ongoing COVID-19 epidemics will depend on social distancing and other efforts to mitigate the spread of the virus.”
As of Monday, a Houston Chronicle analysis reported 901 deaths related to COVID-19 in Texas and 219 deaths in the Houston region.