Over the last two weeks (ending 1/2), the country’s epidemic metrics were mixed and perhaps skewed by holiday reporting disruptions, but there were increasing signs that we may be approaching the much anticipated post-holiday peak.
Hospitalizations
Total hospitalizations, ICU bed usage and ventilated patients all increased over the last two weeks but the rate of increase again slowed. Total hospitalizations and ICU bed usage were up by 9% and 7%, respectively. More significantly, ventilator use came close to leveling out at just a 2% increase. All three indicators fell in the last three days.
Generally, hospitalization data is considered to be mostly free from reporting delays because of the holidays. But even if that is not the case and the decline in the last few days obscured some hospitalizations, the following chart of the daily percentage change in hospitalizations over the course of the epidemic shows the degree to which the increase has stabilized at just under 1% over the last month.
California has continued to drive a large portion of the increase in hospitalizations but thankfully its hospitalizations have leveled off over the last several days. Several other states, such as Tennessee, Arizona, Alabama and, to a lesser degree, Texas and New York, have contributed to the increases. Hospitalizations have declined in 24 states in last thirty days.
Several of you have asked about the per capita hospitalizations by state. The COVID Tracking Project has added this interesting map which shows the number of hospitalizations per million by state.
Fatalities
Average daily reported fatalities dropped slightly from 2,529 to 2,411. However, there was likely some delay in reporting over this two-week period because of the holidays, so I doubt there was really any improvement. But it does continue to appear that reported daily fatalities have stabilized in mid-2,000 range.
The 30-day lagging case fatality rate leveled out at 1.6%, down from the 2% we had seen for months. The 20-day lag, which increased in the previous two weeks, took a significant leg down to 1.2%. I think with the vaccination of seniors, we could really see this rate begin to fall precipitously.
In that same vein of thought, the IHME model has recently added a vaccination parameter. As a result, it is now projecting that daily fatalities will peak on January 11 at 3,800 and then rapidly fall to just over 700 per day by the end of March.
Testing
The public’s interest in getting tested appears to be waning. Total tests are now falling (down 10%) after plateauing for several weeks. Positive test results were also down by a similar percentage (8%). The ratio of tests coming back with a positive result was virtually unchanged at just over 11%. I suspect as we see hospitalizations and fatalities begin to decline, testing could really collapse.
Summary
It is risky to draw many conclusions from the data from the last two weeks because we do not know the degree to which the reporting systems were disrupted by the holidays. Many think that we are in for a difficult 2-3 weeks because the Christmas and New Year celebrations will set off another surge. But as I discussed in my last US report, the holiday gathering theory has never had much support in the data.
I am increasingly convinced that we are approaching the tipping point when natural immunity, supplemented by the vaccine rollout, is going to begin to seriously put the brakes on the spread of the virus. The IHME model projects that the US reached peak infections on December 24 and will drop by 80% by the end of March.
In a couple of days, I will be writing about the latest research on herd immunity and provide you with a spreadsheet from which you can make your own projections. There is still a lot of uncertainty over how this will play out but I am more encouraged than ever that the end is in sight.
