This page gives some background to the graph data and website design. Go back to the graph itself.

The trendline calculation process is completely automatic and is repeated each time new data is available. The graphs initially display the calculated trendlines with the highest and lowest calculated growth rate. If you click on a red or green trendline segment, the graph both extends that segment and also draws a line at that same growth rate starting with the most recent data point. Trendlines are calculated by curve-fitting an exponential function to each segment of smoothed data, using the following steps:

1. Raw data is smoothed based on a seven day moving average.
2. An exponential curve fit is applied to each seven day period.
3. The time period with the best-fitting curve is considered. The algorithm considers extending that curve to include additional data preceding or following it. If the overall quality of the curve fit improves or remains constant by adding additional data points, the trend curve is extended to include those points.
4. The above step is repeated for other time periods.

Different analyses of the pandemic consider different types of data:

• Cases from testing in the general population. This should give the most rapid information (because tests can happen soon after infection) but suffers from data variability due to the availability and utilization of tests.
• Hospitalization data suffers from some time delay (time between the time of infection and time of hospitalization), but effectively removes some of the variability due to testing availbility.
• Death rates may be more easily (although not perfectly) quantified, but there is a significant time delay from time of infection.

This site focuses on hospitalization data as a best available blend between accuracy and time delay. Data used on this site comes from three sources:

• The Texas Medical Center provides daily updates of ICU and hospital bed census, as well as daily new hospitalizations. This website relies on manual data entry from the published data by the TMC.
• SETRAC provides daily updates of ICU and hospital bed census for many institutions in Houston and the surrounding region. This website relies on manual data entry from the published data.
• The Texas Department of State Health Services provides a daily update of hospitalizations for each Trauma Service Area in the state. This website attempts to automatically load the spreadsheet provided daily by the agency.

Other sources may be added over time. All available sources can be seen by clicking on the data source menu above the graph. Available data starts at about April 1 for these data sources.

Because COVID-19 information in Texas is not available in a database/API format, the information in this site has been parsed or manually entered. While it is intended to accurately reflect the available information, errors are possible.

Site developed by Matt Reisdorf, Houston, Texas This site is a volunteer project with no advertising revenue, personal information collection, or connection to any corporation. While this site uses available information from hospitals, regional organizations, and governments, this site is not affiliated with any of those institutions. Site graphs utilize the Chart.js library.