Last week, Texas hit the grim milestone of 10,000 COVID-19 deaths.
That is, it did if you went by local news outlets and the Johns Hopkins Coronavirus Resource Center. The Texas Department of State Health Services didn’t report a similar figure until around 5 p.m. Eastern time on Monday.
That small, seemingly modest inconsistency might be easily brushed off—especially in the context of divergent COVID-19 stats across the country—if it weren’t for the fact that the state has endured weeks of data woes, sowing distrust across Texas. State officials have blamed coding errors, system upgrades, backlogs, changes in methodology, and closures of testing centers for a recent surge of red flags in its coronavirus stats.
ADVERTISEMENT
Regardless of the cause, the deluge of questions about figures ranging from infection rates to total deaths have made it difficult to get a clear picture of what’s happening in the virus-ravaged state, public health experts told The Daily Beast. That would be concerning in its own right, but it’s especially glaring given the state is using that data to make decisions about reopening schools and other facilities as summer turns to fall.
“There’s a fog around Texas right now as to what’s actually going on,” said David Rubin, director of PolicyLab at Children’s Hospital of Philadelphia, whose COVID-19 models have played a role in White House policy-making. “Things are kind of messy.”
“It’s not making sense,” Dr. David Lakey, vice chancellor for health affairs and chief medical officer at the University of Texas system and a member of the Texas Medical Association COVID-19 Task Force, told The Daily Beast of the state’s positivity rates last week.
Positivity rates reflect the proportion of tests that come back COVID-19 positive out of the total number performed. It’s an indicator, experts generally agree, for whether enough tests are being performed to paint a complete picture of a region’s caseload—as well as, more obviously, how widespread the infection is. As of Monday, Texas had 555,394 cumulative cases of the virus, according to Johns Hopkins data, and a positivity rate over the past week of 12.95 percent.
But as Lakey explained it, things got weird around July 27, when the state changed its method for reporting deaths, rather than infections. Specifically, that responsibility went from one that was placed on local health departments to a tally based on death certificates. More than 400 previously unreported deaths were added to the total death toll in that single day.
“All of a sudden it looked like deaths sky-rocketed,” Lakey told The Daily Beast.
Days later, according to the Texas Tribune, an “automation error” caused approximately 225 deaths to be incorrectly added to the overall state tally. They were later removed after a quality check on the information revealed that COVID-19 was not the direct cause of those deaths.
But as soon as those discrepancies were explained and began to even out, following the shift in reporting, another—equally disturbing—one started to arise.
An early August upgrade increased the Texas Department of State Health Services’ capacity from 48,000 to more than 100,000 electronically-submitted test results per day, a spokeswoman told the Tribune.
Last week, the seven-day positivity rate suddenly climbed up dramatically, to around 25 percent on average in the week ending Aug. 11. For comparison, positivity rates in New York state have hovered below 1 percent in recent days. Texas Gov. Greg Abbott said in May that if the state’s positivity rate went over 10 percent, it would be considered a red flag.
That red flag has been present every single day since late June. Even so, local and state experts on the ground told The Daily Beast last week that the actual information coming out of hospitals all over the state seemed to clash with such enormous positivity rates. Those findings were bolstered by national experts who said the rates seemed out of line with hospitalization numbers.
Meanwhile, the number of COVID-19 tests reported in Texas jumped by 124,693 in a single day on Thursday, Aug. 13, after declining over the previous days, according to the Texas Tribune. A spokesperson for the Texas Department of State Health Services told the news site that the jump was the “result of coding errors and a system upgrade,” as well as backlogs at labs that could not be added until the errors were corrected.
Before the pandemic, about 1,200 lab results per day were sent to the Department of State Health Services, but that has increased dramatically. Now, the data comes in much larger quantities from labs all over the state, which is sorted according to its format—including faxes and electronic records—and then uploaded into the Center for Disease Control and Prevention’s National Electronic Disease Surveillance System.
By Sunday, experts told The Daily Beast that Texas’ positive test rate seemed to be looking more credible, as it had dropped down to 11.25 percent after a fourth straight day of a decline, according to a Houston Chronicle analysis of state data.
As these issues arose, the data changed in one other (likely related) way: The number of reported tests went down, which Abbott has said he would investigate. The governor has since noted that a possible explanation for the drop was the closure of temporary testing sites, which were opened in July to help handle hot spots. Others have noted a drop in demand for diagnostic tests, a claim seemingly bolstered by the state health department’s claims, to the Tribune, that community testing sites have “plenty of capacity” and few, if any, lines.
In another twist, a coding error by Walgreens Pharmacy led to an undercount of about 59,000 coronavirus test results across the state, KVUE-TV reported.
Diana Fite, president of the Texas Medical Association and an emergency physician based in Houston, explained that what keeps a positivity rate high is when “most people going for tests are doing so because they’re sick.” A lower positivity rate will only be observed once enough people are going in for tests that more of them come back negative, but with current testing capacities in the state, “you just can’t practically do that right now,” said Fite.
And though to her mind, “that 11.25 percent makes more sense,” she still “can’t really say how accurate it is.”
Those fluctuations, which Fite acknowledged were “confusing” even to professionals, have raised concerns about Texas’s ability to track—let alone handle—the spread of the virus as schools make crucial reopening decisions in cities from Houston to San Antonio this week.
Big changes in the numbers, said Fite, “make everybody not trust anything.”
Neither Gov. Abbott’s office nor the Texas Department of State Health Services responded to requests for comment from The Daily Beast. But the latter told CNBC last week that it was “looking into the situation and has reached out to clinical labs and statewide health-care associations.”
“We are looking into why the number of lab tests reported to DSHS has been trending lower over the past few days,” Lyndsey Rosales, a spokesperson, told the news site. “We are closely analyzing our data to uncover any anomalies.”
So what do experts tell school boards or counties or the average Texan trying to make decisions about their children’s safety?
“Without really good data, it’s going to be hard to navigate and know where we are in the response,” Lakey said. “If you don’t know that you can trust that positivity rate and trust the case counts, then you’re flying blind.”
Still, according to Fite, the answer in the interim is the same as it always was: Pay attention to the broad trends instead of the specific numbers. Look at all the metrics together, instead of relying on just one. And be cautious.
Whatever the positivity rate, said Fite, “People who are at risk, who have chronic illnesses or elderly, need to stay quarantined.”
“No matter what, it’s still out there,” she added.
