Incompatible Covid Findings from the UK

Here are some surprising findings based on diagnostic tests administered to nearly 20,000 people by the UK’s Office for National Statistics:

At any given point between May 17 and May 30, an estimated average of 0.10% of the community population had COVID-19, or around 53,000 people.

That seems awfully low. The median age of the UK population is 40 — two years older than the US, a year younger than France. Based on earlier estimates from well-designed international serology studies, the age-adjusted covid mortality rate in the UK would likely be about 0.8%. From May 17-30, 4,200 people died of covid in the UK — an average of 300 per day. If 0.8% of those who are covid-infected eventually die from the disease, then those 300 daily deaths represent a daily infection rate of 300/.008 = 37,500. Optimistically, assume that a covid infection lasts two weeks: 37.5K x 14 = 525,000 people in the UK had covid on any given day — that’s ten times the ONS estimate, or around 1% of the population currently infected.

During the two-week study interval between 17 and 30 May, 29,000 people tested diagnosis-positive for covid in the UK. The ONS estimate acknowledges that the official case count underestimates the true spread of the virus. But they bump the Dx+ number by only 80%; more likely it should be increased by many times that. Alternatively, the covid mortality rate in the UK would have to be far higher: 37.5K deaths divided by 525K infections = 7%, more than 10 times as high as the US covid mortality rate, and 7 times as high as France. Not bloody likely, as they say in the Old Country.

The ONS found that the 53K infected people represented a decrease from 136K earlier in May. If a typical covid death happens two weeks after initial infection, then those who died between 17 and 30 May would have gotten infected two weeks earlier, or between 3 and 16 May. Is there reason to believe that the new infection rate dropped substantially during that two-week interval, lending support to the ONS estimate? Running tallies for new diagnoses show 53.6K new diagnoses in the UK for 3-16 May, dropping to 29.1K for 17-30 May: a 46% decrease. So it’s plausible that the average number of currently infected people during the ONS study interval would have been something like 284K, or 0.4% of the population. Based on other analyses, diagnostic testing grossly underestimates the number of covid infections in a population. Still, calculating an estimate based on diagnostic data comes up with a current infection level that’s twice as high as the 136K reported by ONS.

The ONS website clarifies that their survey refers to

the number of coronavirus (COVID-19) infections within the community population; community in this instance refers to private households, and it excludes those in hospitals, care homes or other institutional settings.

Those exclusions would skew the ONS results substantially toward underestimation.

It’s extremely notable that, in a separate study, the ONS has found that close to 7% of the population tests positive for covid antibodies. Forty thousand people have died of covid in the UK. If the UK covid mortality rate is 0.8%, then 40K/.008 = 5 million people would have been infected since the beginning of the epidemic. 5 million/67 million = 7.5% of the population: very close to the ONS’s antibody study results.

About 90% of all UK infections have been diagnosed over the past nine weeks. If the average person remains infected for 2 weeks, then on any given day there would have been 7% x 2/9 = 1.5% of the population across the entire 9-week span. However, the test-positive diagnostic rate over the past 2 weeks has been less than half of what it had been over the preceding 7 weeks.

Taking all things into consideration, something on the order of 0.7 percent of the UK population being currently infected with covid seems likely — an estimate that’s 7 times the ONS estimate, but that meshes closely with the ONS’s own antibody survey results and with findings from other countries.








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