Reinfections, Breakthrough Infections

If you don’t look at the website www.dailysceptic.org, it is worth checking out daily.  This article on the site is another big picture attempt to explain the pattern of the epidemic.  (DS Article)    It does seem that something beyond mere seasonality has to be a factor in the timing and shape of waves, and this seems like a reasonable possibility.

Okay, a crowdsourcing opportunity on data here, as I am trying to figure this out.   We have very fragmentary data regarding breakthrough infections, and even less data on reinfections.  Readers have been tremendously helpful by pointing me to some data from their state or region.  I simply can’t look at every state, but anyone who is aware of data and can collect and send it to me or send me a link, then maybe collectively we can gather enough data to get a better sense of what is happening.  Thanks to anyone who can help with that.

We got a little more from a CDC study released at the end of last week.  It has taken me a while to post on it because  frankly I am baffled by the study design and had a devil of a time rationalizing the reported results.  The intent of the study, following the CDC’s grotesque miscommunication around vaccines and masks, was blatantly to encourage people who have previously been infected to get vaccinated.  The study comes from Kentucky and claims to show that vaxed people who had previously been infected had lower reinfection rates than those who were previously infected but not vaxed.  (CDC Study).   This is supposedly a case control study, but I don’t understand the logic.  What would make sense is compare a group of fully vaccinated persons who had prior infection with a group of unvaccinated people who had a prior infection.  The researchers got there, but in a roundabout way.

The headline finding of the study was that people who had a prior infection and were not vaccinated were 2.3 times more likely to get reinfected than those who had been fully vaccinated after a prior infection.   A case was a person who had been infected in 2020 and then reinfected during May and June of 2021.  The controls were people infected in 2020 but not reinfected during May and June of 2021.  The cases and the controls were matched across a variety of factors.  There were 246 cases and 492 controls.  Of the case patients, 20.3% were fully vaccinated versus 34.3% of controls.  So that is how you get to your relative risk calculation.  50 of the reinfections in the cases occurred in the vaxed, so these are breakthrough infections.  So almost 20% of the reinfections were breakthrough infections.  That is fairly significant.  Again the way the analysis was set up, however, makes me uneasy.  Why not take all reinfections in the state, and compare how many of those occurred in vaccinated versus unvaccinated persons?  That would give you cleaner data.

Here is some of the data I have collected, with the help of others.  Minnesota is saying there have been 5599 cases, 514 hospitalizations and 56 deaths among the fully vaccinated.  I will point out again that this is a high hospitalization rate, almost 10%, compared to the overall one of a little over 5%, while the death rate is lower that the overall death rate for reported cases, at 1%.  These are both likely highly exaggerated.  I believe we are back where we were at the start of the epidemic, where a very low percent of all cases were detected.  Most vaccinated people probably don’t even know they are infected.  And we know from earlier reporting that half the hospitalizations aren’t actually because of CV-19.

Thanks to Twitter, some information from Dallas.  2676 breakthrough infections, or .8% of their total.  The age skewed old, as you would expect given the distribution of vaccination.  325 hospitalizations, so more than a 10% rate, but get this, and this is good transparency in reporting, only 139 were people admitted for CV-19 treatment, the rest just tested positive on admission.  That is a true 5% or so rate, which is consistent with the overall rate in Minnesota.  34 died, so a little over 1%, but again, Dallas is at least honest in saying that so far 3 of those people have already been determined to have died with, not because of CV-19 and a number of the deaths are under investigation.  One thing the breakthrough infection data is clearly revealing is that while cases are almost certainly way under-reported, hospitalizations are over-attributed to CV-19, as are deaths.

Here is another state with much better data than Minnesota:  Oklahoma.  Look at this presentation, very clean and accessible.  (Okla. Data)    Scroll down and you will see the breakthrough infection data.  Now this is really usable and useful data.  There are 1594255 fully vaccinated people in Oklahoma.  There have been 2514 reported breakthrough infections, 197 hospitalizations and 25 deaths.  The hospitalization rate is around 8% and the death rate per case is 1%.  There is also a very useful table showing the trends.  As you would expect, as vaccinations increase, so does the percent of cases in the vaccinated, which rose from .03% in February to a peak of 7.3% in June and has declined to 5.74% of all cases in July, as cases spiked in the state.  This indicates that in the current wave, more of the cases are occurring among the unvaxed.  But wait, there’s more, as the late nite TV ads used to say, look at the great table on hospitalizations among the vaxed and unvaxed by age group.  In the last 30 days, 24% of all CV-19 hospitalizations were among the vaxed, but per capita, not more than among the unvaxed.  The age structure is different, with a much older skew in the vaccination group.

Thanks to a reader who sent me some Illinois links.  Their data presentation is worse than Minnesota’s.  They give us breakthrough hospitalizations, 714, and deaths, 180, but no case numbers.  How bizarre is that.  (Ill. Data)

So it looks to me like we have a lot of undetected breakthrough “cases”, in those cases the hospitalization rate seems similar to the overall rate, maybe more skewed toward the elderly and that deaths occur at a lower case rate.  But be really nice if someone, like the CDC, actually did their job and collected and reported this data in a detailed and timely fashion.  And how about some reinfection data while you are at it.

And here is what I found on re-infections.  From the UK, a survey that apparently occurs regularly.  (UK Data)   Big picture is that reinfections are rare, the viral loads are lower, and the infections are likely to be mild or asymptomatic.  The risk of reinfection peaked in the first month after becoming at risk, which likely reflects the time it takes for adaptive immunity to fully develop.  In fact the peak is rather dramatic and falls rapidly.  Now this could also reflect lower community spread and the effect of vaccinations was also not reported, and I have to suspect at least some of this group was vaccinated.  Out of over 17,000 people surveyed, there were only 133 reinfections, and only 24 of those had a cycle number lower than 30.