Back

More prevalent, less deadly? Bayesian inference of the COVID19 Infection Fatality Ratio from mortality data

Delius, G. W.; Powell, B. J.; Bees, M. A.; Constable, G. W. A.; MacKay, N. J.; Pitchford, J. W.

2020-04-22 epidemiology
10.1101/2020.04.19.20071811 medRxiv
Show abstract

We use an established semi-mechanistic Bayesian hierarchical model of the COVID-19 pandemic [1], driven by European mortality data, to estimate the prevalence of immunity. We allow the infection-fatality ratio (IFR) to vary, adapt the models priors to better reflect emerging information, and re-evaluate the model fitting in the light of current mortality data. The results indicate that the IFR of COVID-19 may be an order of magnitude smaller than the current consensus, with the corollary that the virus is more prevalent than currently believed. These results emerge from a simple model and ought to be treated with caution. They emphasise the value of rapid community-scale antibody testing when this becomes available.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.