NedNotes (not blog): COVIData Round-up 14aug20

Introduction: Declines in global and U.S. static pool rates are slowing down and levelling at 5.3% and 7.2%, respectively. The rates remain significantly higher than the 1918 flu mortality rate of 2.3% and much higher than the expected mortality rates for this coronavirus epidemic of below 1%. U.S. mortality run rates changed little before levelling at 3.2%. This week, Russia released news of an approved vaccine met with a well-founded scepticism across the world; there may be a political or geo-political dimension to the likely mis-information.

¿Why call the Russia vaccine the true 'Russia hoax'? The time-line just does not add up. Assuming a three month development cycle for the prototype vaccine and keeping in mind that Russia under-reported its data for two or three months of her epidemic, it is unlikely that the Gamaleya Institute started serious work on the vaccines earlier than mid-April. That means a best-case development date of mid-July. Three weeks to test the vaccine? Do not think so.

Transmissibility holds the key. Unless transmissibility proves lower than expected, the U.S. death rate could be higher than is acceptable in other parts of the world. Leading indicators for the death portend a slowly improving trend. The sample states for the South and Southwest could produce 500,000 deaths by year's end from the data accumulated below, though epidemiologists temper that forecast to a range between 200,000 and 300,000 for the country. The latter figure is more reliable for two reasons.

1. The epidemiologists know what they are doing and I do not; their findings control the forecasts for fundamentals changing behaviours, and wider use of mitigating treatments.
2. The projections from these data are simple extrapolations applying the 'rule of 72' (versus seventy last week) to guesstimate the length of time that a given growth doubles the base number. The current time for doubling in 5⅓ weeks.
   

One possibility of higher death tolls reflects the findings published this week that, adjusting for factors like changing demographics, the United States reported in excess of 200,000 more deaths year-over-over for the past 4½ months than previous years. This report implies that COVID-19 may have claimed more lives than previously thought; up to 40% mores deaths.

The reporting going forward may be under-stated as the Trump Administration has switched the reporting protocol to a new web-site opened in late July by the Department of Health and Human Services. The rapid transition from the National Healthcare Safety Network, previously sponsored by the Centres for Disease Control, has challenged many health administrators, particularly those of smaller hospitals. Again political interference may be driving this switch; ¿control the data, control the narrative?

Analysis of normalised death tolls. The tables below present the compilation of general U.S. data on the coronavirus contagion from thirty-three states and six territories; these jurisdictions cover in excess of 80% of the U.S. population and greater than 90% of the deaths recorded to date. First, this illustration familiarises the reader with the methodology underlying the normalisation of death tolls, to enable direct comparisons between states, per the calculations in the larger table; the data themselves are stale.

The U.S. death toll increased by 4% during the past week. Of the states reviewed below, the ones causing the most concern with increasing death rates include:

• Florida with a 19.5% growth in the number of deaths;
• Texas with a 17.4% growth in deaths;
• South Carolina with 15.9% growth; and,
• Georgia with 13.0%

The states hardest hit early on -- where my extended family tends to live -- all did better with Connecticut, New York, New Jersey and Michigan reporting mortality growth rates uniformly less than one per cent!

One point to note: the State of Washington was one of ten states hit hardest in the beginning, with King County initially being the national epicentre. Kudos to Governor Jay Inslee for turning in the best performance of any of those ten states, including California, with a death toll currently running at less than half the national death rate.

Review of the five-state sample. The five states in the South and Southwest -- California, Texas, Florida, Arizona, and Oklahoma -- comprise 55-60% of the population, confirmed cases, and number of deaths of the two regions. The concern with these states, despite the levelling growth rates of confirmed cases, remains the acceleration of death rates by trend-weighted growth ahead of the compounded growth rates in Florida and Texas.

Arizona and California are moderating their weekly growth rates in death tolls though they remain high in the 10-16% range. The challenge for these states -- particulary Arizona, Texas, and Oklahoma -- and the regions beyond (i.e. states like Nevada, the Carolinas, Colorado, Louisiana et al.) will be the spread of the American epidemic into the rural areas of large states that enjoy fewer human, hospital and financial resources to manage the contagion.

.