There was a large-scale study from Israel posted in April that concluded
This study suggests that both the BNT162b2 vaccine and prior SARS-CoV-2 infection are effective against both subsequent SARS-CoV-2 infection and other COVID-19–related outcomes. Moreover, the effectiveness seems similar for both cohorts. This puts into question the need to vaccinate recent (up to six month) previously-infected individuals.
The Cleveland Clinic study I posted followed over 50k employees for ~5 months (Dec '20 - May '21) & divided the subject employees into four categories:
- Previously infected*, vaccinated** (n ≈ 1,220)
- Previously infected, unvaccinated (n ≈ 1,359)
- Not previously infected, vaccinated (n ≈ 29,461)
- Not previously infected, unvaccinated (n ≈ 22,777)
*Previously infected = positive SARS-Cov2 test prior to 11/4/20. CCHS was not routinely testing asymptomatic employees, but about 12% of these previous infections had no symptom onset date in the data.
**“Fully Vaccinated” = 14 days after 2nd dose (81 employees who got J&J were censored on date of receipt), treated as a time-dependent variable, so as the study went on the unvaccinated cohort shrunk while the vaccinated grew.
2,154 SARS-Cov2 infections were observed in the 5 months of the study. 2,139 of those were in group 4 while the other 15 infections were in group 3, so 0 infections in the previously infected groups (1 & 2), regardless of vaccination.
Some interesting (to me) items in this paper:
Roughly 50% of previously infected employees chose to get vaccinated, and about 60% of non-previously-infected employees were vaccinated by the end of the observation period. These are employees of a major Ohio health system.
The study was not designed to determine duration of protection from natural immunity (nor from vaccines) but the median duration from previous infection to start of study was about 5 months, plus the 5 months of the study, suggests natural immunity protects against reinfection for at least 10 months (probably longer).
The discussion section is particularly interesting (to me, at least), when the authors get into what they perceive as strengths & limitations of their study:
Strengths:
- large sample size
- follow-up of up to 5 months, which is longer than the period in the published mRNA efficacy studies
- health system tracked infection among employees (relatively) accurately (but see limitation 1)
Limitations:
- no regular screening testing, so asymptomatic infections may have gone undetected - both in the classification as “previously infected” as well as during the observation period
- No children, few elderly subjects, few if any immunosuppressed subjects
- Lack of access to detailed clinical information on employees (e.g. severity of illness)
Conclusion
Individuals who have laboratory-confirmed symptomatic SARS-CoV-2 infection are unlikely to benefit from COVID-19 vaccination, and vaccines can be safely prioritized to those who have not been infected before.