99.8% Bayesian Credible Interval for COVID IFR

Post #650

In this prior report, the range of plausibility on the infection fatality rate (IFR) for Alpha variant COVID was discovered. But after re-watching Dr.'s Dan Erikson and Artin Massihi make a permanent record out of early COVID numbers (a wise thing for them to do), I had to re-run the simulation to check for feasibility of what they said.

To recap, by late April in California, there had been 281,000 COVID tests with 33,865 confirmed positives — which is a test positivity rate of 12%. Applied to California’s population of 39.5 million, that makes for 4.74 million COVID infections. But at the time, only 1,227 COVID deaths had been recorded (~1 death per 3,863 infections).

But that is less lethal than seasonal flu, which can cause up to 1 death per 700 infections in a bad year. Using known death data from the UK Tech Briefing #5, I was able to discover the limits on plausibility for a COVID IFR:

All of the plausible IFR values — those able to re-create the known UK death data — were between 0.13% and 0.20%. At 0.13%, there is one death per 769 infections. At 0.20%, there is 1 death per 500 infections.

One explanation for why UK death was higher than early California death would be the difference in baseline vitamin D levels. The British are notorious for having deficient and unhealthy vitamin D levels, making them more susceptible to respiratory infection and death.

Either way, it is not plausible that COVID was ever twice as lethal as seasonal flu.

Reference

[As of 19 Jan 2021, from 117,000 COVID infections, there were 169 deaths by Day 28 (corrected to 190 total deaths, by using upper bound estimates from Linton et. al.)] — Page 3. Epidemiological findings. UK Technical Briefing #5. https://www.gov.uk/government/publications/investigation-of-novel-sars-cov-2-variant-variant-of-concern-20201201

[“Time-to-death” model for COVID; 89% of death occurs by Day 28 using the 95% UB of both the mean and SD of the lognormal model that fit the actual deaths best] — Linton NM, Kobayashi T, Yang Y, et al. Incubation Period and Other Epidemiological Characteristics of 2019 Novel Coronavirus Infections with Right Truncation: A Statistical Analysis of Publicly Available Case Data. Journal of Clinical Medicine. 2020 Feb;9(2). DOI: 10.3390/jcm9020538. PMID: 32079150; PMCID: PMC7074197. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7074197/

[What level of evidence it takes to make it appropriate to talk about something as being, for at least practical purposes, “proven”] — A Few Things You Should Know About Paternity Tests (But Were Afraid to Ask). https://digitalcommons.law.scu.edu/cgi/viewcontent.cgi?referer=&httpsredir=1&article=2024&context=lawreview

[Imperial College model using an evidence-contradicting IFR of 0.9%] — Imperial College COVID-19 Response Team. Report 9: Impact of non-pharmaceutical interventions (NPIs) to reduce COVID-19 mortality and healthcare demand. https://www.imperial.ac.uk/media/imperial-college/medicine/sph/ide/gida-fellowships/Imperial-College-COVID19-NPI-modelling-16-03-2020.pdf