For CDC Weeks 31 & 32 (ending 8/1 & 8/8) the COVID stats for Texas were all over the place, as the data become more confusing every day.

Fatalities

As I mentioned in this post, on July 25 Texas began reporting COVID fatalities based on death certificates filed with the state.  Previously the state had relied on fatality counts from local public health agencies.  The change revealed that many of the local agencies were woefully behind in counting fatalities.   In Harris County, for example, about twice as many death certificates listing COVID as a cause of death had been filed with the state than Harris County was reporting. As of August 8, the new count had Texas COVID fatalities at 8,343, up 73% from what Texas was reporting two weeks ago.  The new methodology reporting the actual date of death (as opposed to the date the death was reported) gives us a clearer picture of the course of the epidemic.  Here is the chart of fatalities in Texas as of the actual date of death.

At first glance, it would appear that fatalities peaked about the middle of July and have been decreasing since.  However, there is a lag time between when a person dies, and when the death certificate is filed with the state.  Under state law, the death certificate is supposed to be filed within 10 days.  However, on one day that I looked at recently, 40% of the deaths reported were from more than 10 days prior.  The red line in the chart above shows the number reported by July 25 versus the end of last week.  It shows the degree to which the fatalities were understated for the first two weeks in July.  So, whether mid-July really proves to be the peak is yet to be seen.

The highest daily death toll so far was on July 14 at 221.  For context, normally about 500-600 Texans die daily during this time of year.  The COVID fatalities so far represent about 4% of total annual fatalities for Texas in recent years.

Texas’ new methodology increased the per capita fatality rate to 30 per 100,000 population.  That is still well below the national average of 47 and ranked Texas 25 among all states and 8 among the 10 largest states.  So, far I have not been able to determine if other states are reporting on the same basis or not.

Testing Results

I am reluctant to even include a discussion of the testing results.  Every day it is becoming clearer that the testing is tardy and, at least as a measure of new infections, unreliable.  After I wrote recently about how my friend  received multiple positive tests before being cleared, several other people have written to me reporting the same thing.  In fact, I have begun to wonder if the number of people who have received just one positive test result might not be the minority.  If that is the case, then the numbers are greatly exaggerated.  Also, just about every day we hear more stories about false positives, including the Ohio governor just last week. With that rant in mind . . .  testing has begun to slow significantly.  After hitting 400,000 tests per week since mid-July, testing last week dropped to 264,000.  The number of positive results also declined significantly, but the positivity rate last week jumped up to its highest level ever at 19.3%.  So, clearly the drop in the number of positive test results has to be viewed with a jaundiced eye.

Hospitalizations

If you are not confused enough yet, the number of people hospitalized for COVID in Texas dropped starkly over the last two weeks.  There were 10,893 patients hospitalized for COVID two weeks ago, but that was down to 7,437 as of August 9, a 32% decline.  ICU beds with COVID patients dropped from about 3,300 to just over 2,600.  The drop has been fairly uniform across the state.

Overall, the data from the last two weeks seems to suggest that Texas saw a peak in its COVID outbreak in mid-July.  But the high positivity rate last week is not consistent with that interpretation.  Of course, as I mentioned earlier, I am having less and less confidence in the test data.  I think the most hopeful sign is the drop in hospitalizations.  The next couple of weeks should give us a clearer idea whether we are really on the downhill slope or not.