This week the media has been screaming about Texas and Florida. You would think that a nuclear bomb had obliterated Houston, Dallas, and Miami with mass casualties. So what’s going on for real? Let’s dig in.
- Texas, Florida Governors Order Bars Closed, Impose New Restrictions as Cases Surge-Politico
- Coronavirus Update: Florida and Texas Take Steps Back as US Infections Soar-Yahoo Finance
- Texas Governor Rolls Back Reopening as US Coronavirus Cases Hit Record-WSJ
- Florida Reports Nearly 9,000 New Coronavirus Cases; Texas Reverses Course…-Washington Post
I’ve also seen stories of overflowing hospitals now, projections of overflowing Texas hospitals by mid-July, and general armageddon. Given that these are two different states and there is a bunch of data to review to vet these claims, I’ll just focus on Texas today.
How We’ll Vet “Texageddon”
I’ll use three data sources:
- The University of Washington IHME website
- The Texas Department of State Health Services (DSHS) COVID dashboard
- The Texas Hospital Association (THA) website
The IHME Projections
When I’m researching COVID headlines, I always begin with the University of Washington’s website and their IHME COVID model (1). Here are the available hospital beds in the state of Texas right now:
Huh? Right now, Texas seems to have ample hospital bed capacity. However, Austin Mayor Steve Alder (D) said during an appearance on CNN: “The trajectory that we’re on right now has our hospitals being overwhelmed, probably about mid-July.” So what does the model product for mid-July? See below:
Huh? Still not close to overwhelmed. However, the IHME model counts all beds available and it’s likely that the state of Texas has allocated certain beds for COVID-19. Hence, maybe the IHME model is different than the data reported by the Texas Department of Health? This is from the Texas DSHS website (2):
The Texas DSHS website has more in-hospital cases listed than IHME (1,645 vs 5,102), but note those are patients with a positive COVID test. We really don’t know if these are patients with the disease of COVID-19 or if they’re just in the hospital for something else and tested positive. Even if we use the DSHS number of beds allocated for COVID (12,398), the current availability statewide is only about 40% of capacity. However, the number of positive cases in the hospital is going up this past week (top bar chart above) in a concerning way. Why could that be? Let’s first look at the testing capacity as that drives reported positive cases (but generally not hospitalizations for the disease of COVID-19):
Here the number of tests being performed are going up each day. For example, in early May the state of Texas had performed about 500,000 tests. By the end of June, almost 2 million tests had been run. Meaning 75% of all the tests performed in the state have been performed in the past 6 weeks. Hence, that will definitely explain a spike in reported positive test results in the general public.
Not to discount the overall COVID hospital trend on the DSHS website, which is concerning, but let’s start digging into those numbers as well. For example, as far as hospital patients with a positive SARS-CoV2 test, some of these positive hospital cases could be due to increased testing of patients who are hospitalized for other causes. So can we posit why asymptomatic people who tested positive for the SARS-Cov2 virus might be in Texas hospitals without being actually sick with the disease called COVID-19? What could cause more testing of asymptomatic people in the hospital? If you’re a doctor who is on a hospital staff that answer is clear, surgery, and other procedures. Note this early May message (right before that spike in tests seen above) from Baylor Hospital in Dallas to its staff (obtained from a Texas colleague):
Here, Baylor, like many hospitals across the nation, is testing all routine surgical and procedure patients for COVID twice BEFORE they are cleared to proceed. Hence, some patients in the hospital who have no COVID symptoms and who are scheduled to have surgery will test positive for the COVID viral or antibody test and be counted as a COVID patient. I’ve discussed this issue of having a positive SARS-CoV2 test and no significant disease versus having the COVID-19 disease. In fact, in our state of Colorado, changing the death numbers from “died with COVID-19″ to “died of COVID-19″ reduced the number of COVID deaths by 25%.
This is the Texas positive test rate (percentage of positive tests over time) which is always a better metric of viral spread than the number of positive cases, as this number doesn’t vary with the number of tests you perform (while the number of positive tests goes up the more you test):
So the percent positive tests were running as high as 14% in mid-April, went down during the lockdown, and are now inching back up to about 12% over the past three weeks. This is the most convincing metric of community spread of the disease.
So is Texas screwed? I found a press release put out this week by the Texas Hospital Association (3):
Huh? The DHS website looks like there could be a problem, the IHME model shows no problem, and on Wednesday, the Texas Hospital Association (THA) put out a press release stating that they had “substantial capacity”. Basically, they relayed that while there were COVID hotspots in the state, many rural areas remain little impacted so patients could be treated in these locations if push comes to shove. Guess where the THA is located? Austin, Texas! The same city whose Mayor went on CNN predicting COVIgeddon! You can’t make this stuff up!
There is no Texas hospital meltdown due to a huge spike in COVID cases! There are some concerning trends in hospital use, but some of that could be due to more people in the hospital getting tested for COVID. In addition, two of the three data sources I checked, argue that Texas has MORE THAN ENOUGH hospital capacity. Hence, making some changes in opening strategy is a reasonable way to deal with a trend in rising hospital cases.
Regional ShutDown to Prevent Hospital Overload
As I argued previously, a balanced approach to managing COVID-19 longterm is not to stay shutdown as a society until we get a vaccine. Doing that would be societal suicide. However, to protect both people and livelihoods, a better solution is to look at hospital usage and only perform regional shutdowns to prevent health system overload while protecting the most vulnerable. So while the media has clearly exaggerated what’s happening in Texas, the state seems to be making some smart moves to try to keep a lid on trends that could possibly overload hospital capacity. Hence, the goal now is to review hospitalized cases each week in hotspot areas and decide if the regional projections the state is making are accurate. If the projections are off and there remains more than enough hospital capacity, then Texas can proceed to slowly ease restrictions.
The upshot? Texas is not in danger of running out of hospital beds anytime soon. Hence, this is yet another case of a Texas-sized exaggeration and fomenting panic by the media. Texas has made some choices to roll back components of its reopening plan to try to protect its hospital system, which I generally agree with at this point. Hopefully, they will reassess on a weekly basis and relax those closures if the hospital capacity doesn’t look challenged.
(1) The University of Washington. Institute for Health Metrics Evaluation. Covid-19 Projections-Texas. https://covid19.healthdata.org/united-states-of-america/texas Accessed 6/27/20.
(2) Texas Department of State Health Services.
(3) Texas Hospital Association. Hospitals Report Substantial Capacity For COVID-19. https://www.tha.org/Public-Policy/Newsroom/Hospitals-Report-Substantial-Capacity-for-COVID-19. AUSTIN, Texas – June 24, 2020. Accessed 6/26/20.