We really understand almost nothing about how viruses other than SARS-2 ordinarily behave

Many points of evidence suggest that we're worrying about reinfection and mortality rates that aren't that remarkable in the context of other viral pathogens

Reports of SARS-2 reinfection are everywhere. The vaccinated appear to be peculiarly subject to reinfection; many of them have had their third or even fourth experience with the virus already. This is one of the most common objections to my post on why you shouldn’t worry too much.

Surely something bad is happening here, right?

Corona, it is true, circulates preferentially among the vaccinated. This is probably because the vaccines fix immune response on the wild-type spike protein, giving new variants with a slightly different spike the keys to the kingdom. It is probably also because the vaccines downregulate generalised, innate immune reactions in the vaccinated, causing their immune systems to “function more like those of the bats who carry these viruses with them,” in the words of friend-of-the-blog Rintrah Radagast. Sub-optimal doesn’t mean catastrophic, though. There’s no evidence I can find that people are suffering substantially higher rates of respiratory illness now than in the past. It looks a lot like we got lucky: Omicron emerged just in time to spare us the worst consequences of our imprudence.

Reports of constant reinfections might sound bad, but the truth is that we’ve never tested this widely for any other virus in history. Our knowledge of SARS-2, its reaction to vaccines, and its associations with all-cause mortality, is totally unique and we have no basis for comparison.

For all that’s been written about the prodigious transmissibility of Corona, it was a fairly rare virus before this year. With the emergence of Omicron, it’s become more pervasive, but in terms of infectiousness, it’s not doing anything that other viruses aren’t doing:

In this chart of German National Reference Centre swabs, keyed to week numbers, we see that SARS-2 looked a lot like a fifth human-infecting coronavirus (hCoV) through 2021. It stole the show after Omicron in January, but never succeeded in becoming more than a plurality of infections across Germany, and by mid-April (week 16) it was even ceding ground to long-missing influenza (the red at the bottom). There’s nothing here that looks like a population-wide collapse in the immunity of vaccinated Germans.

What’s more, this chart is a survey of people who were sick enough to visit their doctors, but the wide availability of rapid antigen tests and mass testing programs means that we’re hearing about cases of SARS-2 reinfection that we’d never notice were it any other virus. Those rare studies that bother to test people regardless of symptoms find that garden-variety rhinoviruses and coronaviruses circulate with quite high frequency even in the dead of summer. This 2005 study from the Netherlands finds viral pathogens in almost 25% (adjusted) of asymptomatic controls. Children in particular have very high rates of rhinovirus and coronavirus infection whenever they’re tested; 20% of asymptomatic infants in this 2003 study tested positive for rhinovirus. These numbers suggest that reinfection from ordinary viruses must be quite common, especially among children – and we’re all children when it comes to SARS-2. There’s some limited cross-immunity from other coronaviruses, so we’re not exactly two year-olds, but we’re close. We’ve just never really quantified the rates of virus infection and reinfection before now, and we’re fixating on numbers that we don’t really understand.

As for mortality: We should be especially cautious about death statistics in the Omicron era. If you test the sick and the dying for any virus, you’ll find very high fatality rates. Nearly ten percent of the elderly people hospitalised for rhinovirus in this retrospective study died within thirty days. What if we developed nationwide rhinovirus testing programs that screened everyone admitted to hospital for rhinovirus and then counted their deaths, while allowing a great part of the mild cases in the young and the healthy to resolve out of sight? What kind of view would we have of rhinovirus and the threat it poses? This is why we need to wait for clear excess mortality signals that a lot of people are dying, before we conclude that Omicron is doing a lot of killing.

Many critics have thrown excess mortality statistics at me these past few days, and I’m loath to reopen the whole question. Just to clarify what I mean here, though, I invite you to consider this graph of excess mortality in Portugal since 2019:

I’ve annotated this chart in six places that I think provide useful perspective:

A) is the 2019 influenza season, and C) is the first real Portuguese mortality spike from Corona, in Winter 2020/21. Both of A) and C) involve more excess deaths than B) the March/April 2020 Corona wave, D) the pre-Omicron Fall 2021 wave, E) the first Omicron wave, and so far also F) the BA-5 wave that we’re currently worried about. Yes, the data are incomplete, they could get worse, but this is how things look right now.

Our strength has always lain with our willingness to provide context and perspective to disease statistics that sound scary in isolation. We shouldn’t lose sight of this, just because the vaccinators and their mRNA elixirs are now implicated in the dynamics of SARS-2 infections. The pandemic is over with. If we want any of the old normal back, we have to let exposure to Corona become a routine part of everyday life.