Top news, reports and insights for today:
- Daily COVID-19 headline summaries for Friday:
- Seven weeks ago, South Korea and The U.S. had the same number of COVID-19 deaths. Today, South Korea has fewer than 300 and the U.S. has over 70,000. What did they do right? (The Atlantic)
- Chinese study found that some men had detectable SARS-CoV-2 in semen, raising possibility of sexual transmission (JAMA)
- Three White House staffers have tested positive for COVID-19 including the Vice President’s Press secretary and the President’s personal valet (Newsweek)
- Instacart is now delivering prescriptions from COSTCO (delish)
- Some COVID-19 patients may have long-term complications from damage to heart, lungs and other systems (Vox)
- More new U.S. cases were reported last week than in the entire month of March
On Thursday, 27,472 new U.S. cases were reported, a cumulative increase of 2%. The U.S. continues to report the highest number of cases across the globe with more than 1.2 million (That’s 1.3 million if including “probable” cases). The U.S. now has 29% of the planet’s lab-confirmed cases. That is more than Spain, Italy, the U.K., France, Germany, Belgium, Netherlands, Switzerland, Portugal and Sweden, combined. Last week, U.S. cases grew by 17 percent. More new cases were reported last week than were reported in the entire month of March (U.S. cumulative total was 185,253 on the last day of March).
What this means: Four consecutive days of rising cases dampen enthusiasm for declining COVID-19 transmission. This rise is not yet related to reopening, but may reflect waning adherence to social distancing. The longer-term trend is toward continued heavy penetration of the epidemic in the interior of the country as widespread community transmission continues in 35 states.
- The more you test, the more you find: Arizona testing fewer of it’s citizens than the African nation of Djibouti
The number of confirmed cases in a given state is a function of 2 factors: a) the number of true cases and b) the number of tests that are done. The U.S. has tested only about 2% of the population, but the variation across states has been dramatic (as I showed yesterday), differing by as much as 6-fold between the highest (Rhode Island at 7.5%) and the lowest (Arizona at 1.3%). It may seem painfully obvious but holding the true number of cases constant, the more testing a state does, the more cases there will be. So what does this actually look like? The answer is in the figure below, which is a followup on yesterday’s report on state variation in testing. The figure plots all 50 states, DC and Guam according to testing rate (tests per 10,000) and confirmed cases per 100,000 (on the log scale). The conclusion here is that there is a pretty strong correlation between testing and COVID-19 prevalence, here explaining a bit more than a third of the variance. I further break the graph down into four “zones”. Why? Because I am an epidemiologist and we can’t help but turn everything into a 2×2 table (inside joke). A few states like New York, Massachusetts, Rhode Island are HIGH CASES + HIGH TESTING (pink). Some states like North Dakota, New Mexico and Vermont have LOW CASES + HIGH TESTING (light blue). That’s the zone we wish all states were in. The largest group of states have LOW CASES + LOW TESTING (yellow) including Arizona, Texas, Kentucky and Maine. The big fear here is that some of those states probably have much more cases hidden by low testing. The final group of states have both HIGH CASES + LOW TESTING (purple). This includes Pennsylvania, Indiana, Colorado and Georgia, many of which are rushing to reopen, despite below average rates of testing. Most if not all of these states have dramatically undercounted their outbreaks.
In the nation with the largest outbreak on the planet, Arizona, with more than 10,000 cases, has tested a lower fraction of its population than Saudi Arabia, Chile, Peru, Serbia, Bhutan, Kazakhstan and the African nation of Djibouti (where 1,135 cases are reported).
Why this matters: We have an election coming up. State governors are highly motivated to keep their reported cases down. One way they can do this is by undermining testing. You do the math.