Top news, reports and insights for today:
- Today’s headlines for Wednesday:
- Coronavirus has now killed more Americans than died over 2 decades in the Vietnam war (NPR).
- Antibody testing demonstrates that COVID-19 is significantly more lethal than regular seasonal flu (Washington Post).
- Vaccine update: 90 candidate vaccines are in development against SARS-CoV-2, 6 groups now in human testing (Nature).
- At least 1,300 federal prisoners test positive as testing ramps up, 70% are in California, Texas and North Carolina (USA Today).
- Stay-at-home orders exact a psychological toll (Los Angeles Times).
- U.S. deaths jump by 4%, more evidence of weekend reporting lag
On Tuesday, an additional 2,173 Americans lost their lives to COVID-19, a rise of 4%. The data confirm a consistent pattern of lagged reporting of deaths on Sunday and Monday. About 1,000 more deaths were reported Tuesday compared to the two previous days. This emphasizes the importance of looking at 7-day moving averages. Eight states matched or set new record high death tolls, including Illinois (142), Wisconsin (19), Arkansas (6), Florida (83), North Carolina (36), South Carolina (15), Delaware (12), and New Jersey (398). Despite rising deaths, Florida will soon allow stay-at-home restrictions to lapse and Arkansas has not implemented such restrictions. In the last week, 12 states saw death totals rise by 50% or more (see bottom graph). Minnesota, New Mexico, Iowa, Alabama, North Carolina, Washington DC, Delaware, and Massachussetts all saw greater than 60% increases last week.
What it means? Daily death totals are bouncing around. This requires us to look at longer time chunks. Sunday and Monday reports reflect systematic reporting delays. Death hot spots remain clear in all regions of the U.S.. These data show that flattening the curve has worked. It does not yet show that we are out of the woods as many wish to believe.
- Countries vary dramatically in testing, high and low testing countries have lower death rates
Here is a new way to look at international comparisons on testing and how it relates to mortality. The figure compares 51 countries that have done 100,000 tests or more, comparing tests per 1 million population (bar length) and the current COVID-19 deaths per 1 million (bar color). The data come from the WORLDOMETER site as of yesterday. Looking at both testing and deaths per 1 million population is a way of approaching an apples-to-apples comparison between countries of wildly different sizes and age structures. It’s in no way a perfect method, but I would bore you with the technically better adjustments and the pattern is largely the same. The top 12 countries have achieved at least 20,000 tests per million. All but Spain, Italy and Ireland have deaths per 1 million population of less than 200. The 11 countries that have tested fewer than 10,000 per million also have low death rates with the exception of France. Japan remains an enigma with extremely low testing and epidemic intensity. The U.S. is testing at half the rate of nations at the top like Portugal and Israel, both of which have low death rates.
What this means: I can’t say that more testing explains low deaths in some countries. The picture is more complex. What strikes me most is how tremendous the variation is in testing rates even among developed Western nations. Hard-hit nations like the UK and France are testing a third less vigorously than the top testing countries. This far into the epidemic, I expected more consistency at least within Europe as the relative success of Germany and Denmark has been so closely watched.
- Georgia: Where is the outbreak most severe? The answer might surprise you.
Georgia has been on my mind lately. The governor has been aggressive in re-opening his state. As collective “quarantine fatigue” sets in, it is hardly surprising that states would want to restart their economies. I decided to have a look at the epidemiologic patterns there. Before looking at the graphs below, take a guess where the epidemic is the worst.
Georgia is a state like many, with one major big metropolitan area, a few smaller cities, and a great expanse of rural communities and small towns. We are accustomed to thinking about states like Michigan (Detroit) and Illinois (Chicago) as examples of how COVID-19 is a big city problem. The top map shows total COVID-19 cases in Georgia by county as of Monday. Data come from the Georgia Department of Public Health. The dark area is Atlanta. But as epidemiologists, we prefer to look at rates of disease. That’s the lower map, which shows infections per 100,000 population. It shows something really different. The highest rates of infection in Georgia are not in urban areas at all. It’s a handful of rural counties in the southwest portion of the state where we find the highest infection rates, some are almost 10-times higher than the national average (now at 307/100,000). Turns out, the hardest hit county (Randolph) had a major cluster of 47 cases in a nursing home in Cuthbert. Nearby Dougherty county has been a national hot spot for some time with 1,498 cases and 120 deaths in a population of under 90,000. Reports suggest a range of factors that might account for this pattern. One county held a big grand opening celebration. Funerals have drawn big crowds in some communities. Many have high rates of poverty and health problems. Tests are reported to be unavailable still. The death toll has fallen heavily on African-Americans, who are a majority in several hard-hit counties.
Bottom Line: According to the Washington Post, of the 20 U.S. counties with the most COVID-19 deaths per capita, five are in southwest Georgia. And this is the state that rushes to reopen first? May and June are likely to be about COVID-19 in rural America.