Texas coronavirus cases and hospitalization continue to decline from July’s record highs

3,172
Texans are in the hospital for the coronavirus as of Sept. 18293 less than a week ago. They occupy 5.5% of hospital beds.
3,978
new cases were reported on Sept. 18. The latest positivity rate — the percentage of positive tests to total tests conducted over seven days — is 7.1%.
14,713
Texans have died as of Sept. 18123 more deaths reported than the day before and 716 more than a week ago.

The Texas Tribune is using data from the Texas Department of State Health Services to track how many people have tested positive for the novel coronavirus in Texas each day. The state data comes from 57 city and county health departments, about 600 hospitals and 340 laboratories and the state vital records registration. It may not represent all cases of the disease given lirevealedmited testing.

To get The Texas Tribune’s coronavirus coverage in your inbox, sign up for our daily coronavirus newsletter.

What you should know:

Where are most of the cases in Texas?

On March 4, DSHS reported Texas’ first positive case of the coronavirus, in Fort Bend County. The patient had recently traveled abroad. A month later on April 4, there were 6,110 cases in 151 counties. As of Sept. 18, there are 682,241 cases in 251 counties. The Tribune is measuring both the number of cases in each county and the rate of cases per 1,000 residents.

Number of cases
Harris and Dallas counties, the two largest in the state, have reported the most cases and deaths.
Cases per 1,000 residents
The rate of cases per 1,000 residents is especially high in the Panhandle’s Moore County, where infections were tied to a meatpacking plant, and in counties with state prisons such as Walker and Jones. Newer hotspots emerged in South Texas and the Coastal Bend — Nueces County, home to Corpus Christi, had one of the fastest growing outbreaks in July.
CountyNumber of casesCases per 1,000 peopleDeaths
Harris121,62826.432,483
Dallas76,60729.621,073
Tarrant42,81621.20684
Bexar41,21921.401,242
Hidalgo30,37535.761,403
Travis28,21223.45398
Cameron22,33952.97865
El Paso22,17926.48491
Fort Bend15,67821.21275
Nueces15,53543.09346
Statewide682,24123.7714,713

How many people are in the hospital?

On April 6, the state started reporting the number of patients with positive tests who are hospitalized. It was 1,153 that day and 3,172 on Sept. 18. This data does not account for people who are hospitalized but have not gotten a positive test.

Experts say there’s a lag before changes in people’s behaviors, like more social interaction, are reflected in coronavirus case data. It takes about nine to 16 days to see increased infections and generally another five to seven days to see changes in the numbers of people hospitalized, said Rebecca Fischer, an infectious disease epidemiologist at the Texas A&M University School of Public Health. (Some individuals are only diagnosed once they make it to the hospital.)

Total current hospitalizations

The state says roughly 2% to 6% of Texas hospitals do not report hospitalization data each day. The average number of hospitalizations reported over the past seven days shows how the situation has changed over time by deemphasizing daily swings.
  • 9% or more of hospitals reported incomplete data, falling outside of the typical range of missing data

On Sept. 18, the state reported 11,590 available staffed hospital beds, including 1,269 available staffed ICU beds statewide. COVID-19 patients currently occupy 5.5% of total hospital beds. In late April, Abbott ordered hospitals to reserve 15% of beds for COVID-19 patients.

According to DSHS, these numbers do not include beds at psychiatric hospitals or other psychiatric facilities. They do include psychiatric and pediatric beds at general hospitals, and pediatric beds at children’s hospitals.

Hospital beds in use each day

The percentage of hospital beds in use shows the strain the coronavirus can put on Texas hospitals.
  • Beds in use
  • Used by COVID-19 patients
  • 9% or more of hospitals reported incomplete data, falling outside of the typical range of missing data

ICU beds available

On April 9, the state started reporting the number of intensive care unit, or ICU, beds available in Texas hospitals. These specialized beds cater to patients with the most life-threatening conditions and include equipment such as ventilators and heart rate monitors. ICU units also have staff who are trained to care for the critically ill.
  • 9% or more of hospitals reported incomplete data, falling outside of the typical range of missing data

Regional differences exist in the availability of beds. In the Rio Grande Valley, the increase in cases in July stretched hospital staff, while in Houston some hospitals had to turn away patients because of overcrowded emergency rooms.

“One of the most critical measures of the ferocity of the COVID virus is the use of intensive care unit beds,” said Britt Berrett, a healthcare management professor at the University of Texas at Dallas. “Healthcare professionals will need to be vigilant in monitoring the time it takes from diagnosis in the emergency room to treatment and admission into the ICU bed.”

Hospital beds in use by region

The percentage of hospital beds in use for each trauma service region shows how the virus is currently impacting hospitals in different parts of the state. These regions are administered by Regional Advisory Councils (RACs).

The Texas Tribune thanks its sponsors. Become one.

How many people have died?

The first death linked to the coronavirus in Texas occurred March 16 in Matagorda County. As of Sept. 18, 14,713 people who tested positive for the virus have died.

On July 27, DSHS began reporting deaths based on death certificates that state the cause of death as COVID-19, instead of relying on counts released by local and regional health departments. On that date, the state added more than 400 previously unreported deaths to the cumulative total. This does not include the deaths of people with COVID-19 who died of an unrelated cause. Death certificates are required by law to be filed within 10 days.

Because of this change, it’s impossible to compare the rate of deaths before and after July 27.

Experts say the official state death toll is likely an undercount.

New deaths from coronavirus reported each day

The average number of deaths reported over the past seven days shows how the situation has changed over time by deemphasizing daily swings.

How have the number of cases increased each day?

The state only tracks confirmed cases of the coronavirus, not probable cases, based on criteria published by the Centers for Disease Control and Prevention. However, DSHS may still accidentally include probable cases for certain counties. When found, they are removed.

Because the state does not include probable cases, these numbers don’t include the results from tens of thousands of rapid-result antigen tests, which suggests the state is vastly underreporting the number of Texans who have tested positive for the virus. Antigen tests are taken by nasal or throat swab like other viral tests, but results are much faster.

Increases in testing have led to more detected cases. In May, a large one-day spike was reported after testing was done at meatpacking plants in the Amarillo region. Delays and backlogs in reporting can also create one-day surges, when cases from multiple days are added on the same day. A backlog consisting of hundreds of thousands of coronavirus tests was disclosed by the state in August — experts said that the backlog made state data hard to interpret, especially when assessing current trends.

In June, the number of new cases each day trended dramatically upwards, before dropping in late July. Abbott issued his mask order on July 2.

New cases of coronavirus each day

The average number of cases reported over the past seven days shows how the situation has changed over time by deemphasizing daily swings. The number of new cases reported drops on weekends, when labs are less likely to report new data to the state.

How has the positivity rate changed?

Abbott said he is watching the state's positivity rate — the percentage of positive tests to total tests conducted. The seven-day average positivity rate is calculated by dividing the average of positive tests by the average of total tests conducted over the last seven days. This shows how the situation has changed over time by de-emphasizing daily swings. Public health experts want the positivity rate to remain below 6%.

In early May, Abbott said a rate over 10% would be a “warning flag.” The state exceeded that mark in June and remained above 10% until late August.

DSHS started breaking viral tests out into molecular and antigen tests on their site in early August. The state began using only molecular tests to calculate its positivity rate, excluding antigen tests.

According to the Food and Drug Administration, molecular tests, or PCR tests, "detect the virus's genetic material" and are highly accurate. Antigen tests detect proteins on the virus and are typically faster, but can be less accurate and produce false negatives.

7-day average for the positivity rate

The state released a new, more accurate version of the positivity rate in September, which is calculated by dividing positive molecular tests by total molecular tests. Because this new formula relies on the date on which the test was administered, the rates for previous days will be recalculated as more test results from those dates come in.

On Sept. 14, DSHS began publishing a new version of the state’s positivity rate, which relies on the date a coronavirus test was administered. Previously, the state’s rate relied on the date a test was reported to health officials and verified as a case.

The change revealed that the state’s positivity rate was higher in the spring than originally disclosed.

The Texas Tribune thanks its sponsors. Become one.

How many people have been tested?

As of Sept. 17, Texas has administered 5,885,282 tests for the coronavirus since March. We do not know the number of Texans who have gotten a test because some people are tested more than once. The state’s tally also does not include pending tests.

The DSHS data also might not include all of the tests that have been run in Texas. The state has said it is not getting test data from every private lab, and as of mid-May only 3% of tests were coming from public labs. The state has since stopped differentiating between tests reported by public and private labs.

Coronavirus test results reported to the state each day

The average number of tests reported over the past seven days shows how the situation has changed over time by deemphasizing daily swings. In April, Abbott set a goal of 30,000 daily tests in the state.
  • Viral tests (molecular and antigen)
  • Antibody tests
  • The state did not release the breakdown of tests

Even as demand for testing has increased, both public and private labs continue to prioritize Texans who meet certain criteria, but every private lab sets its own criteria.

On May 21, DSHS disclosed for the first time that as of a day earlier, it had counted 49,313 antibody tests as part of its "Total Tests" tally. That represents 6.4% of the 770,241 total tests that the state had reported on May 20. Health experts have warned against counting antibody and standard viral tests together because they are distinctly different tests. Antibody tests detect whether someone was previously infected, while standard viral tests determine whether someone currently has the virus.

Antibody tests are typically reported a day late.

The Texas Tribune thanks its sponsors. Become one.

How is this impacting Texans of color?

The limited data released by state health officials offers a murky glimpse of the virus' impact on Texas communities of color. Race and ethnicity are reported as unknown for a significant portion of the completed case reports. (Agency officials said some people prefer not to provide the information.)

On July 27, the state released complete demographic data, including race and ethnicity, for coronavirus deaths after changing its method for counting deaths to rely on death certificates. The data shows that Hispanic Texans make up roughly 48% of the total deaths, but only 40% of the state’s population.

Case data gathered various parts of the state shows the disproportionate impact of the virus on Black and Hispanic neighborhoods. The areas with the highest positivity rates in Harris County are predominantly Hispanic, according to a UTHealth School of Public Health analysis. In Dallas County, lower-income Black communities have also reported some of the highest positivity rates.

Although state leaders acknowledge the demographic data for cases and tests is lacking, they have indicated the state won't be taking steps to mandate reporting to fill in the gaps. In June, the state announced they are planning on ramping up testing in areas of the state that are predominantly Black and Hispanic, as well as launching a study on the coronavirus’ effect on vulnerable populations.

What else should I know about this data?

These numbers come from the Texas Department of State Health Services, which updates statewide case counts by 4 p.m. each day. The data is from the same morning, and it may lag behind other local news reports.

In order to publish data quickly, the state has to bypass what is normally a months-long process of reviewing the COVID-19 data and performing quality checks before publishing. That’s why all of these numbers and information are provisional and subject to change.

The state’s data includes cases from federal immigration detention centers, federal prisons and starting in mid-May, some state prisons. It does not include cases reported at military bases.

From March 13 through March 24, the Tribune added cases from Lackland Air Force Base in San Antonio, where hundreds of American evacuees from China and cruise ships were quarantined.

Those case counts came from the Centers for Disease Control and Prevention.

Notes about the data:

On March 24, the state changed how it reported numbers resulting in a sharp increase in cases.

Antibody tests were included in the new total tests counts for each day before May 14. Previously, the state had counted about 50,000 total antibody tests as virus tests, artificially deflating the positivity rate.

After a system upgrade on June 7 resulting in incomplete test data, the state revised the test numbers for June 6 to show a decrease in total viral tests. The testing numbers for June 6 are not shown in the test results by day chart.

On June 16, the state included 1,476 cases previously reported by the Texas Department of Criminal Justice from Anderson and Brazoria County in its cumulative case count. The new cases for June 16 do not include those cases.

On July 17, the state received about 5,268 additional cases from Bexar County. The state only included 608 of those cases as new confirmed cases for that day and added the rest to the cumulative count.

On July 19, the state removed 3,676 duplicate antibody tests from the previous day’s total.

From July 23 to July 28, between 9% and 18% of hospitals reported incomplete hospitalization numbers due to changes in reporting to meet federal requirements.

On July 25, the state removed 2,092 probable cases from the Corpus Christi-Nueces County Public Health District that had been previously included in the cumulative case count.

On July 27, the state began reporting deaths based on death certificates that state COVID-19 as the cause of death. On that day, more than 400 previously unreported deaths were added to the total death toll due to the reporting change.

On July 30, the state said an “automation error” caused approximately 225 deaths to be incorrectly added to the overall death count; a subsequent quality check revealed COVID-19 was not the direct cause of death in these cases. We updated the cumulative numbers for July 27-29 to account for this error. The automation error also caused us to incorrectly state the percentage of Hispanics who have died of COVID-19 and the number of previously unreported deaths on July 27. These have been corrected.

On Aug. 3, the state removed 536 duplicate confirmed cases from the overall cases count for Bexar County. Bexar County reported 471 new confirmed cases on this date.

On Aug. 7, DSHS started breaking viral tests out into molecular and antigen tests on their site. We’ve combined the two to come up with the total number of viral tests.

On Aug. 7, DSHS reported that some molecular tests had been miscoded, inflating the number of antibody tests over the previous couple of days. This was corrected, resulting in the number of antibody tests to decline from the 6th to the 7th. Because a breakdown of these tests is not available, the charts are showing the inflated numbers on those days.

On Aug. 10, the number of new cases reported did not include new cases from Nueces County due to a “large backlog of positive lab reports” that the county was working through.

In mid-August, several labs submitted large backlogs of tests to the state, which could not have been added until coding errors were fixed and a system update was complete. This caused several changes to the data:

- On Aug. 13, the state reported a record number of tests. Of those 124,000 tests, approximately 95,000 were from one lab serving several hospitals.

- Thousands of previously unreported cases were added to the statewide total over the course of several days. On Aug. 17 and on Aug. 19, the state added a backlog of 5,195 and 550 positive cases, respectively, from Dallas County to the total. On Aug. 20, the state added 44 cases from Dallas County and 336 from Fort Bend County. On Aug. 21, the state added 206 cases from Dallas County and 164 from Montgomery County. On Aug. 22, the state added 459 cases from Dallas County, 134 cases from Montgomery County and 23 cases from Nueces County. On Aug. 23, the state added 862 cases from Dallas County and 43 cases from Nueces County. On Aug. 24, the state added 93 cases from Dallas County. On Aug. 25, the state added 84 cases from Dallas County and 171 cases from Montgomery County. On Aug. 26, the state added 94 cases from Montgomery County. On Aug. 27, the state added 424 cases from Dallas County and 117 cases from Montgomery County. On Aug. 28, the state added 59 cases from Dallas County. On Aug. 29, the state added 18 cases from Dallas County and 12 cases from Montgomery County. On Aug. 30 and on Aug. 31, the state added 65 and 241 cases, respectively, from Dallas County. On Sept. 1, Sept. 2, Sept. 3 and Sept. 4, the state added 235, 167, 35 and 1 case(s), respectively, from Dallas County and 13, 10, 27 and 28 cases, respectively, from Montgomery County. On Sept. 5, Sept. 6 and Sept. 7, the state added 30, 195 and 3 cases, respectively, from Dallas County. On Sept. 8, the state added 5 cases from Nueces County. On Sept. 9, the state added 168 cases from Montgomery County. On Sept. 10, the state added 110 cases from Dallas County and 56 cases from Montgomery County. On Sept. 11, the state added 59 cases from Montgomery County. On Sept. 12, the state added 64 cases from Dallas County, 46 cases from Galveston County and 309 cases from Harris County. On Sept. 13, the state added 5 cases from Dallas County.On Sept. 14, the state added 17 cases from Dallas County and 1,399 from Bexar County. On Sept. 15, the state added 100 cases from Dallas County, 110 cases from Montgomery County and 316 from Harris County. On Sept. 16, the state added 26 cases from Collin County, 68 cases from Dallas County, 2,411 cases from Harris County, 109 cases from Montgomery County and 3 cases from Houston County. On Sept. 17, the state added 13 cases from Dallas County, 167 cases from Harris County, 93 cases from Montgomery County, 239 cases from Tarrant County, 15 cases from Collin County, 1 case from Camp County and 1 case from Upshur County. On Sept. 18, the state added 52 cases from Dallas County, 13 cases from Galveston County, 305 cases from Harris County, 47 cases from Montgomery County, 94 cases from Collin County, 1 case from Bowie County, and 1 case from Harrison and Houston counties. In all instances, these cases were not added to the daily new case count.

On Aug. 24, the state was unable to update its testing numbers because of a power outage affecting multiple state agencies. The numbers, however, were added retroactively.

On Sept. 4, officials in public health region 11 found 1,238 older cases in 12 counties in south Texas that were previously unreported. These were added to the statewide case total but not included in the daily case counts for those counties.

On Sept. 9, TDCJ reported 453 fewer cases among inmates in a state prison in Walker County. These were removed from the county’s total case count, as well as the statewide total. We asked DSHS why these were removed but haven’t received a response yet.

On Sept. 12, the state said a data entry error caused 91 cases to be incorrectly added to the Sept. 11 statewide and Colorado County total case counts. We updated the cumulative numbers for Sept. 11 to account for this error.

On Sept. 14, the state published a new, more accurate version of the state’s positivity rate, which relies on the date a coronavirus test was administered. We changed our positivity rate chart to use this version. The numbers of tests used to calculate this new positivity rate do not match the numbers of tests shown in the test results by day chart.

The same day, the state started deduplicating their test results, causing a drop in overall tests. This made the seven-day average of tests incalculable for one week.

Also on Sept. 14, Lamar County overstated their case count by 41 cases. These were removed on Sept. 15. The statewide cumulative case count was also adjusted to reflect this change.

On Sept. 16, case counts decreased in 12 counties. TDCJ reported reduced case counts in Bee and La Salle Counties. Cases were deduplicated in Calhoun, Gonzales, Guadalupe, Kerr, Lavaca, Orange, Roberts and Wharton counties. Shackelford and Swisher Counties saw decreases after updating case information.

On Sept. 17, case counts decreased in 13 counties. TDCJ reported reduced case counts in Bee, Childress, Houston, Karnes, Madison and Walker Counties. Cases were deduplicated in Bandera, Kendall, Titus and Zavala counties. Archer, Swisher and Yoakum counties saw decreases after updating case information.

On Sept. 18, case counts from TDCJ decreased in Anderson County by 1,070, which decreased the total number of cases for that county. TDCJ reported reduced case counts in Duval, Fannin, Grimes and Liberty counties. Bailey County saw a decrease after updating case information.

Carla Astudillo, Mandi Cai, Darla Cameron, Chris Essig, Anna Novak, Emily Albracht and Alexa Ura contributed to this report.

Previously, The Texas Tribune incorrectly stated our formula for calculating the average daily positivity rate. This has been corrected.

The Texas Tribune thanks its sponsors. Become one.