Why don't Covid-19 vaccine trials report statistics for the first 14 days?
Recently a pediatric colleague sent me this preprint link to a Danish cohort study of the Pfizer Covid-19 vaccine in long-term care facility residents and healthcare workers. (https://www.medrxiv.org/content/10.1101/2021.03.08.21252200v1 ) It showed real-world effectiveness of two doses of the mRNA vaccine: 64% and 90% VE in the two groups respectively beyond seven days after the second dose.
However, from 0 to 14 days after the first dose the risk of Covid-19 infection was actually increased in vaccine recipients: in the LTCF residents VE was -40%, CI -62% to -2%; among healthcare workers VE was -104%, CI -118% to -91%.(Table 2)…..By contrast, statistics for the 0 to 14 days after the first dose were not reported in the randomized trials of the Pfizer, Moderna, or AstraZeneca vaccines. (Polack et al, NEJM 2020;383:2603. Baden et al, NEJM 2021;384:403. Voysel et al, Lancet 2021;397:92) Why not? Are the manufacturers hiding negative data?
There are at least two previous examples of vaccines increasing the risk of illness from the target disease for the first week or two: The unconjugated PRP Hib vaccine increased the risk of Hib disease 2.6-fold in children within 7 days of vaccination, probably because vaccination produced a transient decrease in protective antibody. (https://www.ncbi.nih.gov/books/NBK236299/ pages 250-261)….A Danish study of pandemic influenza A/H1N1 vaccine among people with chronic diseases found a 3.58 risk increase of hospitalization for influenza 1 to 7 days after vaccination, CI 2.27 to 5.64, and a 2.12 risk increase of laboratory-confirmed infection, CI 1.58 to 2.87. (Table 2, Emborg et al, BMJ 2011;344:d7901)
A massive matched case-control study of the Pfizer vaccine in Israel found 62% VE against severe Covid-19 disease and 72% VE in preventing deaths for days 14 through 20 after the first dose. However, the results were biased against the unvaccinated controls: On day 1 there were already 359 documented infections in controls and 172 in the vaccinated group, and 227 versus 90 symptomatic cases, respectively. (Table S7, Dagan et al. NEJM online Feb 24, 2021. DOI:1056/NEJMoa2101765)
The vaccines may be effective, but we will need more transparency from vaccine authorities as we continue to study vaccine effectiveness and adverse effects.
ALLAN S. CUNNINGHAM 16 March 2021
Competing interests: No competing interests