Vaccine vs. Natural Immunity
After two years, it bears repeating -- epidemiological research is narrative and its power is in the application
(Starting off on a personal note. If you just want to read about immunity, skip to below the line)
Two years ago TODAY I was in my office (as in my real office, on campus; not my make-shift office in the basement of our home) when the phone rang. It was the Communications Manager at the college; I assumed he was calling to tell me my website was out of date. To my surprise, he was calling because one of the local news stations wanted me to do a live interview on TV that evening. He was asking do you want to go to the news station tonight and talk about the novel coronavirus that is spreading worldwide. After recovering from the surprise of the request, I agreed to do the interview. And as soon as we hung up, I simultaneously called my husband and googled “what do you wear for a TV interview.” I was a little excited & (a lot) scared.
During my first TV interview two years ago, I explained what a virus is (a packet of pathogenic genetic material) and how respiratory viruses spread (droplets and aerosols). I also predicted that we would begin to see cases of this new disease (the term COVID-19 was not coined until mid-February) in our region in the near future (it would be less than two months later).
I could have never imagined we would still be all-pandemic-all-the-time two years later, but here we are…
For me, that first interview turned into Ask the Expert panel discussions, my own TV spot Ask the Doctor, a weekly Facebook Live Q&A, and the title of Exclusive Medical Contributor. And though I am still a professor and researcher, the course of my career has taken off in an unexpected and wonderful direction —
I’m committed (as cliché as it may be) to putting the PUBLIC back in public health.
Specifically, this means I am committed to slowing the spread of COVID-19, protecting vulnerable communities, and fighting against the epidemic of misinformation and politicization of science.
And this blog is dedicated to sharing evidence-based information and actions you (and your communities) can take to slow the spread of disease and misinformation. If you haven’t subscribed, do so now. And then you can read more about the latest report detailing natural versus vaccine-induced immunity.
Last Thursday, COVID-19 Cases & Hospitalizations by COVID-19 Vaccination Status & Previous COVID-19 Diagnosis — California & New York, May-November 2021 was published in the MMWR. The findings summarized in the report have been used as anti-vaccination ammunition, misinterpreted by far too many people, and not used in conversation/context with the rest of the epidemiology cannon.
What was the main finding of the report?
The study found that during the Alpha and Delta variant surges individuals who are unvaccinated AND have not been previously infected with COVID were at the highest risk of contracting COVID and being hospitalized. Immunity was strongest among those who were both vaccinated and previously infected. The study also found that individuals who were vaccinated or previously infected with the SARS-CoV-2 virus had immunity that helped prevent breakthrough cases/reinfections and hospitalizations.
The study also showed (what we’ve known) that immunity, whether it is from a previous infection or vaccine, is NOT static. Our immune responses to COVID-19 (whether from previous infection or vaccination) wane over time, and when immunity wanes the likelihood of a breakthrough case/reinfection increases.
What evidence was provided to support this finding?
The study researchers created four cohorts/groups of people from data housed in databases and registries in New York and California. (Please note, this is not national data or even data that represents a specific sub-population. It is data from two states that could conveniently and with few limitations be combined.) The groups that were created so that comparisons could be made were —
Unvaccinated individuals with no previous lab-confirmed COVID infection (these are the individuals who were most likely to contract COVID and be hospitalized)
Unvaccinated individuals with a previous lab-confirmed COVID infection
Vaccinated individuals with no previous lab-confirmed COVID infection
Vaccinated individuals with a previous lab-confirmed COVID infection
The research team calculated the age-standardized hazard rate ratios. Age-standardized means that a mathematical process (known as direct age standardization) was used to ensure that the underlying age structure of each group was the same (all based on the 2000 U.S. Census). In essence, age standardization allows us to look at the data in a way such that age is not confusing the relationship between the outcome — new case of COVID — and the predictor variables — previous infection or vaccination. The process allows us to say that age is eliminated or standardized, and therefore cannot be an underlying cause of the associations we calculate in the study.
Hazard rate ratios are calculated by determining the incidence (proportion of new cases among the at-risk population in each group) and then dividing the incidence in one group by another; we are making comparisons between the groups listed above. For instance, the study states, “during the week beginning May 30, 2021, compared with COVID-19 case rates among unvaccinated persons without a previous COVID-19 diagnosis, COVID-19 case rates were 19.9-fold (California) and 18.4-fold (New York) lower among vaccinated persons without a previous diagnosis
The 19.9-fold and 18.4-fold decreases are the hazard rate ratios. What this is telling us is that there are far fewer cases of COVID among the group that was vaccinated and without a previous COVID infection (from both states) compared to the group that was unvaccinated and without previous COVID infection.
The study methodology requires a lot of mathematical and statistical knowledge/work, not to mention the ability to merge and combine big datasets across two different state systems.
Statistical significance versus real-life significance?
Like other epidemiological studies, this report focuses on statistical significance. With statistics, the research team is able to say with confidence that the differences seen between the groups did not occur by chance alone. They provide evidence that the decrease in COVID cases and hospitalizations is tied directly to vaccination status or previous diagnosis with COVID.
While statistical significance is important in determining associations/relationships between outcomes (COVID diagnosis) and predictors (vaccination status and/or previous infection), we need to remember that these results need to be used to create meaningful policies, interventions, or educational materials. For me, knowing that people in New York experienced a 19.9 fold decrease in COVID does NOT tell me anything about whether I am at increased or decreased risk over time.
What is needed is real-life significance.
I believe the real-life significance of this study is that vaccinations combined with previous COVID infections result in the strongest immune response. Individuals who have not been diagnosed with COVID, are most susceptible to disease. And both vaccines and previous infections provide some amount of immunity.
That is what I take away from this study to apply to public health practice.
How are the findings being misinterpreted? And used to spread misinformation?
Individuals have looked at the data provided in Table 2 of the study and said there were more cases of COVID among the vaccinated with no previous infection group compared to the unvaccinated group with a previous COVID diagnosis, and therefore vaccines do not work.
This is one study.
One study that was conducted before booster shots were available, as Delta was ravaging the globe, and where vaccine immunity was clearly waning.
This one study was also conducted as the J&J vaccine was being heralded as a one-and-done vaccine (subsequently, we have learned that at least two vaccine doses are needed for immunity to last longer than two months). Additionally, this study was conducted as mask mandates were being lifted for vaccinated individuals and not the unvaccinated — what part did that play in disease transmission and diagnosis? We don’t know; this study did not assess masking as a predictor.
To top it off the study is riddled with limitations (and listed here are just the limitations I see in the study; there are seven others listed by the authors themselves), this study was conducted in two states with robust public health surveillance systems and highly vaccinated populations (~71% of adults are vaccinated in both states). Making sweeping statements about vaccine immunity based on data from two states where the majority of people were vaccinated is not appropriate. You cannot simply apply the results of a single study from one population/state to another without thinking through the similarities and differences between the two.
What else do we know? How does this fit into the canon of epidemiology?
Since last Thursday (when this paper was published) additional data has been shared… during the Omicron surge (which we are in the middle of now), unvaccinated adults are 16 times more likely to be hospitalized with COVID compared to vaccinated adults.
And adults aged 50-64 who are unvaccinated are 44 times more likely to be hospitalized compared to a vaccinated + boosted adult 50-64. EACH DAY 3500 individuals are dying from COVID in the United States; the majority of these deaths (as in, almost every single one of them) is vaccine-preventable.
Taken as a whole, we can say that the initial two shots of Pfizer or Moderna or the single dose of the J&J vaccine were not enough to stimulate and sustain the immune system throughout the Delta surge. As vaccine-induced immunity waned, there were a lot of breakthrough infections. HOWEVER, the vaccines kept people out of the hospital and from dying (which is what we expect from the COVID vaccines). Now that booster shots are readily available for individuals 12 and older, we have an infusion of immunity at both the individual and community level to help us through the Omicron surge.
Natural immunity following infection with COVID does exist (and appeared to be sustained throughout the Delta surge). However, natural immunity comes with the cost of illness, risk of hospitalization, risk of death, and risk of long-COVID. Additionally, unvaccinated individuals on average take two additional days to clear the virus following infection and are, therefore, contagious for 48 hours longer than a vaccinated individual.
On the whole, vaccines provide greater protection than natural immunity. And they come with far fewer risks than a COVID infection. While the study from New York and California demonstrates what happens when immunity wanes, it does not account for the new Omicron variant, other forms of disease mitigation (masking), and the amount of time from vaccination to diagnosis. It cannot be used by itself to make claims about the utility and lack of benefit of vaccine-induced immunity.
A few weeks ago I quoted President Bartlet of The West Wing, who said —
“Every once in a while... every once in a while, there's a day with an absolute right and an absolute wrong, but those days almost always include body counts. Other than that, there aren't very many unnuanced moments in leading a country that's way too big for ten words.”
When it comes to using epidemiological data to develop policy or to direct our individual actions, there is rarely a single paper with an absolute right answer. We must look at each study independently AND then in concert with the rest of the epidemiological evidence/canon.
It is when data from various studies are put together in a meaningful way — acknowledging what we know and don’t know, what is applicable at the moment and what is not, what are the risks vs. the benefits — that we can turn the science into action.
And that is the goal —
To take what we have learned through research and take steps to slow the spread of disease, reduce the number of hospitalizations, and prevent deaths from a vaccine-preventable disease.
Get vaccinated. Get your booster shot (if you are eligible). And continue to wear a mask!