Wednesday evening’s announced numbers extend the trend of new Kentucky cases decreasing. This is evidenced by simple plots of daily cases and cemented by the falling of both 7-Day and 14-Day rolling averages. Based on available data, the Test Positivity Rate as calculated by our Public Health Department is falling in parallel. Even better, both rolling averages are falling away from the exponential growth trendline on semi-log plots suggesting a plateau of new cases if not better. Weekly new cases per week are on track to be much less than the last 4 weeks, but January will likely still be the worst month so far by perhaps as much as 5000 cases.
For deaths however, January continues to be the cruelest month. With 47 new deaths reported yesterday, the total of 880 so far for the month has already set a new record compared to the 754 in December. In my opinion it is too early to tell what is happening to the numbers of people entering hospitals because the day-to-day reported numbers vary dramatically. It is in hospitals and nursing homes that most of the recognized Covid-19 deaths occur. I have begun calculating the daily number of new additions to the reported “Ever Hospitalized” or “Ever ICU” counts using the same logic I use to count new cases and deaths. This takes much of the volatility away. Since mid-January fewer new people are being hospitalized. Nonetheless it is fair to say our hospitals are still working very hard. Despite that burden, some hospitals are taking on the workload of immunizing people against the virus!
As of the end of the week on January 23d, the jagged sawtooth profile of tall daily peaks of followed by deep dips of new cases over weekends and holidays persists, but the peaks are not as high as in early January and our 7-Day and 14-Day averages are back down to about where they were in early December. The number of new cases weekly has been dropping for the past three full weeks. To the extent that Test Positivity Rates are an indicator of community epidemic spread, they remain high but have been trending downwards over the past week. These are all good signs, but we would all like to see a decrease in new cases at this time of rapid epidemic spread nationally and the emergence of new strains of the virus whose impact in the USA remains to be determined. (It will also be reassuring news to learn that the state’s broken case-reporting pathway from its testing vendors had been repaired.)
Following this just past long holiday weekend, the outcome of settling back into a predictable weekend rhythm of data collection and reporting remains unsettled. There are hints that the rate of new cases is slowing, but last week (Sunday, Jan. 10th through Saturday, Jan. 16th) had the second highest weekly count, second only to the just prior week. Our 7-Day average of new cases has been falling from its record high for the last 5 days, but the 14-Day average remains at its highest levels. At the current rate of new cases, we will end January with over 100,000 cases for that single month– a new record high. The Aggregate Total Number of cases continues to climb exponentially. If nothing else changes, we could reach a total of 500,000 Covid-19 cases by the middle of February. Where we go the remainder of this week will likely be telling. I refer the reader to the full portfolio of KHPI’s data visualizations on Tableau Public.
Testing and Test Positivity Rate (TPR). I manually entered the Test Positivity Rates reported by the state since early October into the KHPI database. To make daily calculation of this metric less volitile, the state includes only PCR Viral-RNA tests from swabs, and only from those testing centers that report to the state electronically. Although the major stated reason for calculating this measurement is to determine if enough testing is being done, it is also a truism that if the same kinds of people are being tested and the same number of tests are being performed, that the positivity rate will also reflect the prevalence of active Covid-19 disease in the tested population. Indeed, inspection of the graphic plots shows that the rise and fall of the 14-Day average of new cases tracks the Test Positivity Rate. At least since early October, new cases increased about a week after the Positivity Rate did, and correspondingly decreased when the Positivity Rate did.
We have not yet seen the corner much less turned it.
We anticipate artifactually lower new reported cases and deaths from Covid-19 disease due to holiday work schedules and fewer individuals being tested. New case counts may also be lower recently due to failure of at least parts of the reporting systems of one of the state’s major viral testing laboratories that has not yet been either acknowledged or confirmed by state authorities. If delays and incomplete reporting is the reason, we should expect the record-breaking catch-up counts of the last week to begin to drop, followed much later by a decrease in daily reported Covid-19 deaths. As of January 10, that has not yet happened. At this point, until and unless we see major decreases in daily case and then death counts, we should assume that as is currently the case in many other places in the USA, we are in the middle of a rapidly expanding viral epidemic. The evidence supporting this concern is summarized below and displayed visually on KHPI’s Tableau Public website.
New Record Highs for Daily Cases. We expect in Kentucky and nationally that certain days of the week will have statistically higher or lower case counts than others. For example, Sunday and Monday have been low-count days. It is noteworthy that in each of the past seven weekdays save one, new record highs were set. The exception was last Tuesday, January 5th with 1693 new unduplicated cases. The last time a Tuesday count was lower than that was October 20th with 1297 new cases. You can step visually through the counts by weekday here.