In a previous post I analyzed the data from the Moderna and Pfizer vaccines trials. Recent data from the romanian authorities allows us to estimate values for the vaccine efficacy which are significantly lower than what was measured in the trials.

Latest data show that about 80% of the hospitalized patients and 90% of the dead are unvaccinated. The vaccine efficacy can be expressed as \(V_e=1-\frac{p_v}{p}\) where \(p\) (\(p_v\)) are the incidences of hospitalizations or deaths in the vaccinated (unvaccinated) populations.

Given the fact that the rate of vaccination is approximately 25% the proportion of vaccinated patients in the hospital satisfies

\[\frac{p*.75}{p*.75+p_v*.25}=.80.\]

This implies a much smaller than predicted hospitalization efficacy \(V_e=.25.\) Protection against death is higher (\(V_e=.66\)) but far from absolute.

Probably better data will increase these estimates but not enough to approach even the lowest estimates one can find (\(.6\) efficacy in preventing hospitalization). Many experts blame the lower efficacy on the prevalence of the new virus variant(s). My guess is that much of the difference is due to increasingly dangerous behaviour of the vaccinated. Also, the persons vaccinated first are not typical. The persons at risk (due to occupation or comorbidities) have been prioritized and are, probably, more open to vaccinate. Therefore the baseline risk (without the vaccine) of the vaccinated population is higher than average and it is normal to have more people than expected sick while vaccinated.

I expect in the future that more data and a better vaccine coverage will show higher effective efficacy. There will also be a public health communication problem as the proportion of vaccinated between the COVID hospitalizations and deaths can only increase.