I have been trying to understand why non-cuoronaviruses would have any effectiveness against COVID-19, and failing so far. Partly, scientists don’t completely understand it, but also it requires a level of medical understanding that I don’t have yet. I’m working on it, and will update this paragraph when I do.
I thought it had to do with live attenuated viruses boosting the innate immune system. It is old news that live attenuated virus vaccines (LAV vaccines) offer some general protection. (See this technical review article, this mass-media article, or all my posts tagged with “live attenuated virus”). However, some of the viruses that provide some protection against COVID-19 are LAV vaccines and some are not.
| Name | Against | Type |
| MMR | measles / mumps / rubella | LAV |
| Tdap | tetanus / diphtheria / pertussis | sometimes LAV? |
| BCG | tuberculosis | LAV, with many different manufacturers using many different strains |
| Shingrix | shingles | recombinant subunit, came out in 2017 |
| PCV13 | pneumonia | conjugate capsular polysaccharide |
| PPSV23 | pneumonia | capsular polysaccharide |
| OPV | polio | LAV |
| various | influenza | varies, but usually inactivated virus |
Shingrix does have an adjuvant which stimulates the innate immune system.
Interestingly, PCV13 gives some protection against COVID-19, but PPSV23 does not, despite them both being pneumonia vaccines. I don’t know if it is because
- the conjugate is important,
- PCV13 has some adjuvant which PPSV23 does not,
- the two strains of Streptococcus pneumoniae (6A and 9V) which PCV13 targets but PPSV23 does not are important.
Note that “getting vaccinated against anything ever” also correlates with higher income and closer attention to health, so one might expect that having had any vaccination would look like it gave some protection. The fact that there are some vaccines which do not show protection actually gives me some hope that these numbers are not spurious.
Note also that there are ton of different BCG vaccines from different manufacturers using different strains.
Okay, here’s the data. Note: some studies do double-blind where they give a shot of the vaccine or of the placebo and then follow the patient. I call those “Recent”. Others are electronic record searches where they say, “did this person ever get a shot of X?” Those don’t usually look at how long it has been since the shot. I put “ever” on those.
| Vaccine | against | where | 1 dose | full treatment |
| Recent MMR [1] | asymptomatic infection | Brazil, health care workers | ~0% | ~0% |
| Recent MMR [1] | symptomatic infection | Brazil, health care workers | 48% | 51% |
| Recent MMR [1] | treatment required | Brazil, health care workers | 76% | 78% |
| MMR ever | hospitalization | Massachussets, Ohio, Florida | x | 38% |
| MMR ever | death | Massachussets, Ohio, Florida | x | 32% |
| Tdap ever | hospitalization | Massachussets, Ohio, Florida | x | 23% |
| Tdap ever | death | Massachussets, Ohio, Florida | x | 20% |
| Recent BCG [2] | infection in over-50 with comorbidities | Greece | x | 68% |
| Recent BCG [2] | hospitalization in over-50 with comorbidities | Greece | x | ~80% |
| Recent BCG [3] | infection in HCW | UAE | x | 100% |
| Recent BCG [4] | infection in vulnerable over-60s | Netherlands | x | ~0% |
| Recent BCG (Tokyo strain) | infection | Massachusetts | x | 92% |
| Recent BCG | infection | South Africa | x | ~0% |
| Recent BCG (BCG Moscow vax) | infection | Brazil | x | ~30% not statistically significant |
| BCG ever | symptomatic infection | Italy | x | ~0%[5] |
| BCG ever | mortality | Germany | x | ~5% |
| Recent BCG | infection or hospitalization | Denmark | x | ~0% |
| PCV13 ever | infection in over-65s | California | x | 35% |
| PCV13 ever | hospitalization in over-65s | California | x | 32% |
| PCV13 ever | death in over-65s | California | x | 32% |
| PPSV23 ever | infection in over-65s | California | x | ~0% |
| Shingrix ever (since 2017) | infection | California | 14% | 19% |
| Shingrix ever (since 2017) | hospitalization | California | 24% | 36% |
| OPV ever | infection | World | x | ~30% |
| recent flu | infection | Qatar | 29.7% | N/A |
| recent flu | hospitalization | Qatar | 88.9% | N/A |
[1] Manufactured by Fiocruz under license from GSK. Trivalent live-attenuated virus.
[2] The preprint didn’t say which brand of BCG vaccine / which strain they used.
[3] Manufactured by the Serum Institute of India with the BCG-I strain from Moscow.
[4] I cannot find the paper — just a non-technical article — which did not specify which BCG brand / strain they used.
[5] There was a small protective effect, but it was not statistically significant. They did not say which brand / strain they used.