Corona 10-Day Trend Update

In brief, the trends observed ten days ago have persisted, with diagnostic test-positives going up while deaths and hospitalizations remain steady.

Test-positives again jumped dramatically, reaching twice the rate recorded during the plateau extending from early May through mid-June:

  • July 2 — July 12:   57.5K new test-positives per day
  • June 22 – July 2:   44.5K new test-positives per day
  • June 12 – June 22:   27.1K new test-positives per day
  • June 2 – June 12:   22.0K new test-positives per day
  • May 23 – June 2:   21.8K new test-positives per  day
  • May 13 – May 23:   24.2K new test-positives per day
  • May 3 – May 13:   24.5K new test-positives per day

The covid death rate, meanwhile, remain plateaued after having dropped by about two-thirds between early May and mid-June:

  • July 2 — July 12:   633 deaths per day
  • June 22 —  July 2:    630 deaths per day
  • June 12 —  June 22:   599 deaths per day
  • June 2 –  June 12:   853 deaths per day
  • May 23 –  June 2:   940 deaths per  day
  • May 13 –  May 23:   1,316 deaths per day
  • May 3 –  May 13:   1,700 deaths per day

Covid hospitalization rates — the CDC data lags by about a week — continued to parallel the death rate trend, remaining plateaued at around 1,000 per day through July 4 after having dropped from around 3,000 between early May and mid-June.

Infection follows an intrinsic timeline, so it’s possible that the lagging data trends for hospitalizations and deaths will soon catch up to the rising diagnosis rates. However, the current trends have persisted for a month now.

Another possible explanation for the misalignment of trends is that the increased test-positive rate is an artifact of increased testing. From the beginning of the epidemic, test-positives have undercounted actual infections by a wide margin. Current elevated diagnosis rates might merely be closing the counting gap between official diagnoses and actual infections, which might not be increasing at all. However, along with the overall increased testing, the last month has also witnessed a steady increase in test-positive rate, supporting the hypothesis that the infection rate really is on the rise.

It’s possible that death rates have remained steady because the virus has begun disproportionately infecting younger people, who know that they’re far less vulnerable to the virus than are older people and so are returning to work and to play. Older people know that they’re far more likely to die of covid, so they’re persisting more diligently in maintaining social distance from potential vectors of contagion. Is there evidence supporting a demographic shift? Since the beginning of May, the percentage of deaths accounted for by those aged 65+ decreased from 83% to 76%. That drop might account for about a 10% swing downward in the US age-adjusted mortality rate — significant, but not nearly enough to correspond with the 50% jump in dx-positives.

Or would it? Covid hospitalization age demographics have also shifted downward, with the percentage of inpatients aged 65+ dropping by one-fifth over the past month while those aged 18-49 have increased by one-fifth. According to my calcs, a 70-year-old who gets infected by the coronavirus is about 20 times as likely to die as a 40-year-old. The CDC used to provide periodic demographic compilations on test-positives, but the recent burgeoning of testing has made it impossible to compile accurate data. Public health departments in states experiencing surges report jumps in test-positives for younger people. If the average age of those newly infected with covid were to drop by 7 years in this country, the estimated covid mortality rate would go down from 0.6% to 0.3%. Under that scenario, a doubling of infections combined with a halving of mortality rate would result in a plateaued daily death rate — which is what’s happened over the past month.

There are too many gaps in the data, hence too many degrees of freedom, to zero in empirically on the most plausible explanation for recent trends. I’ll tentatively infer the interaction of multiple factors:

  • The recent increase in test-positives results from some combination of increased testing and increased infections.
  • The recent stability of the death rate despite increases in infection results from some combination of time lag from infection to death and a decrease in the average age of those who are getting infected.

Trends to look for over the next ten days:

  • If the testing rate and the test-positive rate both plateau or continue to climb, that would support the idea that recent increases in test-positives were an artifact of more widespread testing. If trends diverge between testing rate and test-positive rate, that would increase the likelihood that diagnostic testing is closing in on the actual rate of infection.
  • If the death rate starts to climb, that would lend support to the lagging indicator theory; if deaths stay steady, then a downshifting age demographic comes more decisively into play.




One thought on “Corona 10-Day Trend Update

  1. HHS just took over covid hospitalization data from the CDC. They’re going to prohibit public access, which is bad. However, the CDC’s hospital data has been a weak source, with long time lags. Here’s a better source. It shows hospitalizations increasing since June 24, contradicting the CDC’s plateaued data. The trends are lining up: first more test-positives and percent-positives, then more hospitalizations, then more deaths…


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s