A trend is a trend until it ceases to be a trend. The U.S. remains at 33 new cases per 100K population, essentially unchanged from seven days ago. This is starting to look like a typical turning point in one of these COVID-19 waves. Case rates are rising in some areas and falling in others, and the net (average) just happens to work out to be flat. We have not yet reached the point where new cases are falling uniformly across the U.S., but it looks like we’re headed there.
Data source for this and other graphs of new case counts: Calculated from The New York Times. (2021). Coronavirus (Covid-19) Data in the United States. Retrieved 5/27/2022, from https://github.com/nytimes/covid-19-data.” The NY Times U.S. tracking page may be found at https://www.nytimes.com/interactive/2020/us/coronavirus-us-cases.html
At this point, hospital admissions for COVID-19 are tracking the new cases, albeit at a lower rate than has occurred in the past. And deaths still are not tracking new cases — they continued to decline throughout the Omicron-II wave, and still have not risen measurably during this latest wave. In short, the severity of the average case has fallen, compared to prior COVID-19 waves.
Source: CDC COVID data tracker, accessed 5/28/2022.
On the path to the new normal?
In past posts, I tried to speculate on what “endemic COVID-19” would look like. It’s hard to get a handle on it, because it’s not completely analogous to (say) flu. As with flu, there’s no permanent immunity. But unlike flu, COVID-19 is incredibly contagious and tends to spread in clusters, rather than a flu-like one-at-a-time transmission.
But maybe we’re already looking at the start of the new normal, with the Omicron-II wave.
Now that re-infections are common, COVID-19 new case rates should continue to flare up in response to any change that materially increases disease transmission. In this case, the change that caused the most recent wave is the spread of the somewhat-more-contagious BA.2.12.1 (Omicron-II) mutation. But it could just as well be a change in the weather (expect a winter COVID season parallel to the winter flu season). Or relaxation of COVID-19 hygiene, such as removal of mask mandates or restrictions on certain types of indoor gatherings. Or even the increased person-to-person contact from resumption of school each fall. (Although, so far, that last one has not proven to be a major vector of disease transmission.)
But of the factors that would cause new COVID waves, I would expect that mutation of COVID-19 will eventually fade into the background. The COVID-19 genome will eventually evolve as much as it can. It’s not as if it can get infinitely more transmissible, forever.
For all we know, we may have reached that point already with BA.2.12.1. Omicron was first observed in November 2021, and the first U.S. case was announced 12/1/2021. Here we are, six months and several quintillion viral replications later*, Omicron remains the dominant viral strain. The World Health Organization list of variants of concern still ends with the various members of the Omicron family. Ditto for the European Union’s list. There doesn’t even appear to be a rumor of anything coming down the pike that’s significantly more transmissible than BA.2.12.1.
* There have been roughly 250 million confirmed cases world-wide since 12/1/2021 (reference), with a viral load between 1 billion and 100 billion per infected individual (reference). Taking 50 billion as the midpoint estimate, you end up with (250 x 10^6 x 50 x 10^9 = 12.5 x 10^18 or roughly 12 quintillion viral replications under Omicrion. And that’s only the officially-diagnosed cases. My point being that as far as evolution goes, it’s not the total time that a strain is the dominant strain, it’s the total number of opportunities to replicate that matters. Because Omicron resulted in such a huge wave world-wide, there has been more opportunity for Omicron to evolve than for any prior strain to evolve. And yet, it remains the dominant strain. Suggesting that maybe it has fully or nearly-fully exploited its genome, in terms of becoming ever-more-infectious.
As with flu, hospitalizations and (particularly) deaths will occur mainly among the unvaccinated and the elderly. The vaccine seems to provide reasonably long-lasting protection against a severe case of COVID. (Separately, flu also attacks infants and young to some degree, which is why CDC recommends flu shots for infants. That said, pediatric deaths from flu remain exceedingly rare in the U.S., with the record being 188 pediatric flu deaths in any one flu season (reference)
But you can’t dismiss the protective role of prior infection, even though our public health establishment doesn’t harp on it, the way it does vaccination. It really should come as no surprise that prior infection seems to provide a strong protective effect against hospitalization during a subsequent infection (e.g., reference, reference). Near as I can tell, those estimated effects for prior infection are every bit as large as those from vaccination. Which, again, would be no surprise, as immunity should be immunity, no matter how acquired. (Although vaccine-based immunity reacts to a narrower subset of surface proteins on COVID, as only the spike protein was used in the common MRNA vaccines.)
In fact, if you step back from it, a) average case severity is lower now than it was before Omicron, as noted above, and b) what really changed during the Omicron wave was not our overall vaccination rate, it was the fraction of the population that had some prior infection. Here I’ve lined up the cumulative percent of the population, starting 12/1/2021.
Source: CDC COVID data tracker, accessed 5/28/2022.
But in addition, we know from the seroprevalence surveys that the chart of known infections (bottom graph, above) understates the number of true infections by at least a factor of two. So, in round numbers, between 12/1/2021 and today, we’ve had maybe another 20 million citizens get fully immunized, and another 70 million citizens survive a COVID-19 infection.
Those numbers overlap, but it shows that the main driving force behind the recent decrease in average case severity is probably recovery from prior infection. That’s far different from the situation last year, where the initial wave of immunizations (and relatively low new-case rates) meant that immunization was, at that point, far more important for increasing population immunity to COVID-19.
And so, as a result of both immunization and infection-acquired immunity, we’ve seen a substantial decrease in average case severity.
And that’s likely the model going forward. We’re never going to acquire enough immunity to prevent new COVID-19 infections. But as each new COVID-19 wave results in an ever-smaller fraction of the population with no immunity (neither vaccination nor prior infection), the average severity of COVID-19 cases will continue to decline. (Absent any random mutations that would increase case severity, as with the Delta variant).
We’d like to think that there must be some benign endpoint to that process, but there’s no evidence to support that. Yet. There’s pretty good evidence now that average case severity will be lower. And the mechanism behind that is pretty clear. But there’s no basis from which to say “lower” means “low”, as in, no material threat to the public health. For that, it’s yet to be seen.
Let me sum up what I see as our current COVID-19 glide path.
- The continuing processes of vaccination and infection will result in ongoing declines in average case severity. (That’s absent some random mutation that results in BA.2.12.1 being replaced with a more virulent strain).
- But those processes aren’t necessarily going to result in ongoing declines in new cases. Reinfections are now common. (As they have always been with, say, flu — it’s just that nobody bothered to remark on it.) Nothing appears to provide strong and permanent immunity against any new infection.
- Likely, COVID-19 waves will be “spikier” than flu season, owing both to its vastly higher infectiousness (R-nought) and to its propensity to spread via large clusters, rather than a flu-like one-infection-at-a-time pattern.
- As with flu, you can shift your odds somewhat by getting re-vaccinated every year. In fact, the short-term effect of the COVID-19 booster (against any infection) appears to be much better than the average impact of flu vaccine (against any infection). At best, flu vaccine is only 60% effective (when they correctly predict the dominant strains), and on average (across all years), it’s only 40% effective against any infection with flu.
- But no matter what, this is probably going to continue to kill a lot of frail old people, the same as flu and pneumonia do now. That’s just the way the world works, and widespread COVID in the population permanently adds another end-of-life risk to the existing set of risks. So that’s not a fundamental change, it’s just more of the same.
Long COVID?
For me, the next risk to try to quantify is long COVID.
As with everything else that we incorrectly thought was unique to COVID, there has always been “long flu”, it’s just that nobody paid much attention to it (reference).
Unfortunately, almost everything I’ve read about it has been … how to say this … lacking in rigor.
Let me demonstrate that by returning to that “long flu” article just cited. Using the same questionnaire that was used to diagnose “long COVID”, researchers found that 30 percent of people who had flu had “long flu”. Like so:
The researchers found evidence that Covid-19 patients are more likely to suffer from long-term symptoms than flu patients, with around 42% and 30% respectively reporting at least one symptom three-to-six months after infection.
Source: Forbes, reporting on the underlying research, emphasis mine.
Really? Roughly one-third of flu cases result in “long flu”, using the same methodology as was used to measure “long COVID”? Really? And nobody noticed this before?
Or is it the case that the symptoms — particular the vague ones such as anxiety and fatigue — are both common in the general population, and fairly benign. So that this ultra-broad definition of “long COVID” — measured with no control group — is mostly picking up a lot of people who’d report those symptoms in any case, and a lot of symptoms that are in no way debilitating.
This, as opposed to (say) scarring of the lungs that is visible on xrays for months afterward. Or lifestyle-changing cardiac irregularities, observable via EKG. Or measured reductions in cognitive function that last months after the presumed end of the COVID-19 infection.
What I’m getting at is that the big numbers you see being thrown around for the prevalence of “long COVID” almost certainly exaggerate the extent of the problem. Most appear to have been generated without reference to a control group. What you see reported is not the increase in those symptoms observed in the COVID-19 population relative to those never infected, it’s the absolute rate at which the COVID-19 population reports at least one such symptom. And most appear to include a broad laundry-list of symptoms including vague ones that cannot be verified or disputed based on any objective physical evidence (e.g., anxiety, fatigue, “brain fog”).
So I guess my next task will be to try to sort through that mess, but I’m not very hopeful of finding anything useful. That said, clearly this risk is next in the hierarchy after risk of death and hospitalization. And based on popular-press reporting, this risk is quite large. So I guess I’d better pay attention to it at least long enough to figure out what I think is believable, and what appears to be just artifacts of sloppy methods.
