The stats for the COVID outbreak in Texas improved for the last two weeks (ending September 19). However, Texas continues to struggle with its testing/new case data.
I am going to begin with the hospitalizations data because as I have previously discussed, I have become convinced this is the most reliable metric to judge the spread of serious disease associated with COVID. All the hospital indicators improved again over the last two weeks. Total hospitalizations and ICU bed usage both fell by over 20% again (3,973→3,124 and 1,409→1,083, respectively). They now stand at their lowest levels since mid-June.
Harris County’s numbers have fallen even more dramatically. According to SETRAC, ICU beds with COVID patients is now at the lowest since April 3.
Reported fatalities also fell by almost half over the last two weeks, but the decline really leveled off in the last week.
As we have discussed previously, Texas began reporting fatalities by the date of death in early July. So, the numbers we see each day are the number of fatalities confirmed by the State for that day, not the number of people that died that day. But we can use the death data to create a picture of the course of COVID fatalities over time. The chart below is the number of fatalities by the actual date of death as shown by the State on various report dates, including the most recent one yesterday. The third week in July continues to be the peak of the fatalities (actually July 23 at 273). Nearly all death certificates are filed within about thirty days. At that time, the daily fatalities were running about 125 per day, so down a little over 50% from the peak.
(Reminder – the number of fatalities on the right end of this chart are understated because of the lag time in the filing of death certificates and will, therefore, increase over time as is demonstrated by looking at earlier reports.)
Over the last two weeks, the State came clean that it had a huge backlog of test results that it was struggling to push out. (See Texas Tribune story here.) As a result, the daily “new cases” were actually test results from weeks or even months earlier. As noted in this Tribune story, the State has now indicated that it will publish three different positivity calculations. Also, it is still not clear whether the State has the capability to eliminate multiple positive tests results for the same person. And finally, in a problem not unique to Texas a number of researchers are questioning whether the PCR tests are returning positive results for people actually over the infection. As explained in this BBC story, this could also lead to overstating new cases.
As a result, any reporting regarding new cases, positivity or the R(t) factor, may or may not bear any resemblance to reality. Several media outlets are still reporting the positive test results as new cases, although it is absolutely clear that is not the case. The Texas Medical Center’s COVID dashboard has specifically begun segregating the old test results.
Considering that warning about the data, the State reported an increase in the number of tests from 543,000 to 716,000 for the two-week period. However, on three days during the last two weeks, the State reported approximately 100,000 each day, which almost certainly means they included some backlogged results. The State reported that the positivity rate dropped from just over 11% to just under 7%, resulting in a slight drop in the number of positive tests results for the two-week period (28,875→27,063).
Locally on testing, I saw this story in the Houston Chronicle that Judge Hidalgo is continuing to maintain the County at a red threat level because the County’s positivity rate is supposedly 7.7%, which is above her 5% target. Based on the current data, no one has any idea what the County’s positivity rate is at this time. But the fact that the positivity rate for COVID tests at the Texas Medical Center was 3.9% on Friday and has been below 5% for two weeks, strongly suggests the County’s number is inflated. To tell people that “they should avoid leaving home except for the most essential needs like going to the grocery store for food and medicine” based on a belief the positivity rate is still above 5% is absurd and is absolutely not “following the science.” It is past time to lower the County’s threat level.
So far, there is no sign of a surge in COVID disease from Labor Day or schools reopening. However, the decline in the metrics showed some flattening the second week, perhaps indicating that those factors may be slowing the rate of decline, which up until then, had been quite remarkable. Let’s keep our fingers crossed that the downtrend picks back up over the next two weeks.