Immunity after COVID does not appear to be long-lasting

Dr. Keith Roach M.D.

DEAR DR. ROACH: In your recent column about the COVID-19 vaccine, you review the Pfizer vaccine results along with your suggested approach. You say: "People with a history of COVID-19 infection benefitted from the vaccine just as much as those without, suggesting no natural immunity was present." I'm curious what specific results from the study led you to this conclusion. Thank you! -- Anon. 

       ANSWER: There is some degree of immunity after natural COVID-19 infection, for at least the first three months after infection. After that, immunity begins to wane.

       In the large study for the Pfizer vaccine, longer-term immunity was evaluated by comparing the risk of acquiring COVID-19 (in the group that did NOT receive vaccine) between those who had antibodies showing previous infection against those who did not. The "attack rate" (the proportion infected by COVID-19) was 1.3% in people with and without COVID-19 antibodies at baseline.

       The numbers were small: Only 3% of participants had evidence of prior infection at study enrollment, and additional analyses showed that very few COVID-19 cases occurred in these participants over the course of the entire study. The placebo group attack rate from enrollment to the end of the study was 1.3% both for participants without evidence of prior infection at enrollment (259 cases in 19,818 participants) and for participants with evidence of prior infection at enrollment (9 cases in 670 participants). While limited, these data do suggest that previously infected individuals can be at risk of COVID-19 (i.e., reinfection) and could benefit from vaccination.

       I have attached the briefing document for the Pfizer vaccine FDA EUA, where these data may be found, to my page at

       DEAR DR. ROACH: My son contracted encephalitis at age 7 months in 1963. He had no upper respiratory symptoms; he just fell asleep and the developed high fever, which left him deaf and with expressive aphasia. We never understood what happened.

       I recently read an article about a correlation between herpes simplex and encephalitis. My husband had many episodes of cold sores. We wonder if that could have been the source of infection. It doesn't help my son, but could be a warning to other parents not to let people kiss their babies. What do you think? -- M.

      ANSWER: It's been nearly 60 years, but I am still sorry for what happened. Seeing a child become disabled is an extraordinarily difficult thing for a parent, and raising a disabled child is a challenge for any family.

       There are many causes of encephalitis, and most can potentially lead to neurological disease, including deafness and aphasia. While herpes simplex viruses are a common cause, there are many viruses, bacteria and even parasites and fungi that could have been the cause. Since this was 1963, diseases that are (nearly) eliminated now, such as measles and rubella, would be common causes, though the vast majority of those have characteristic skin findings that no doctor in 1963 would have missed (many doctors now have never seen a case).

       The incidence of encephalitis in children is quite low, less than 1:10,000. Only a small percentage of those will be due to herpes simplex virus. I wouldn't advise parents in general to avoid kissing their children; however, it's a good idea not to do so when a parent is having an outbreak or oral herpes (cold sores).