Rotavirus Vaccination Deserves Nuanced Discussion
RFK Jr and Malcolm Gladwell overlook important data
Clearly, there is overlap between readers of Sensible Medicine and listeners of Malcolm Gladwell’s Revisionist History podcast. A recent episode of Gladwell’s podcast go me to write a post Abortion, Vaccines, and Autism. The same episode inspired frequent contributor Wil Ward to write this piece.
Adam Cifu
On a recent episode of Malcolm Gladwell’s Revisionist History podcast, Gladwell took issue with RFK Jr’s claim that RotaTeq, a vaccine against rotavirus, “almost certainly kills and injures more children in the United States than the rotavirus disease killed and injured prior to the vaccine's introduction.” Gladwell countered that RotaTeq is unequivocally life saving and does not cause any more adverse reactions than a placebo. He calls it “one of the most important public health triumphs of the past 100 years”. Both of their perspectives are informed by the RotaTeq package insert. How could their views be so divergent? I believe Malcolm Gladwell and RFK Jr. selectively report and misinterpret data while overlooking opportunities for nuanced debate.
The Package Insert
The approval of RotaTeq hinged on four RCTs (n=73,086), the largest being REST, the Rotavirus Efficacy and Safety Trial (n=68,038). In REST, Finnish and American infants were randomized to RotaTeq or a saline placebo.
Aligned with Gladwell’s commentary, the package insert reveals no difference in serious adverse events between placebo and vaccine groups (2.4% RotaTeq vs. 2.6% placebo). However, Gladwell incorrectly characterizes all adverse events as identical. RotaTeq leads to a statistically significant increase in non-severe diarrhea (24.1% vs. 21.3%), vomiting (15.2% vs. 13.6%), and otitis media (14.5% vs. 13.0%).
Yet, the benefits of vaccination are remarkable. REST found the number of severe rotavirus gastroenteritis cases was markedly reduced (50 vs 1, ARR 2.13%). Also, 191 unvaccinated infants and 13 vaccinated infants went to the emergency department (ARR 0.63%) and 138 unvaccinated infants and 6 vaccinated infants were hospitalized (ARR 0.46%) for G1–G4 rotavirus gastroenteritis. Vaccination reduced all-cause gastroenteritis hospitalizations, as well.
The Case For Vaccination
Hospitalizations are better avoided whenever possible. The patient who avoids hospitalization benefits, as do patients who are admitted to less crowded hospitals. A decrease in hospitalization probably also makes rotavirus vaccination a cost-effective intervention.
The benefits are likely most substantial in low-income countries. The 5th leading cause of death for children under 5 years old worldwide is diarrheal illness—rotavirus being the most common infectious agent. The only RCT conducted in low-income countries confirmed vaccination reduces severe rotavirus gastroenteritis in South Africa and Malawi. Logically, vaccination should be life-saving where hospitals are absent.
Rotavirus vaccine RCTs have cumulatively enrolled over 100,000 infants. If anything, the lack of increased mortality or severe adverse events in REST and large meta-analysis is reassuring.
The Case Against Vaccination
Rotavirus mortality was rare in high-income countries, including the United States, prior to RotaTeq. This was likely due to better public health measures (clean drinking water) and better access to medical care. When mortality rates are already low, there is little opportunity for vaccination to save lives. Perhaps for this reason, the rotavirus vaccine is not routinely given in several European countries.
RFK Jr may worry that a few severe and undetected vaccine reactions could outweigh the benefits, as was the case with the original rotavirus vaccine, RotaShield. This vaccine was on the market for one year before it was pulled due to a 29-times greater incidence of intussusception. REST did not reveal an intussusception risk, it is possible that other poorly understood adverse reactions are overlooked by clinical trials.
REST found no significant mortality difference between placebo and vaccine arms (20 vs 24). But this was reported as a safety, rather than an efficacy, endpoint. The largest rotavirus vaccine trials are not powered to detect a mortality benefit.
Common Ground & Learning Opportunities
RFK Jr’s claim that RotaTeq has likely caused more deaths than lives saved is almost certainly false. This is not supported by the package insert he cites or any publicly available data. There is little likelihood of severe, unknown reactions. Yet, the evidence base leaves us with more to learn. While a mortality benefit appears likely in low income countries, a healthy vaccine bias may inflate our expectations. A Kenyan observational study found rotavirus vaccination was associated with reduced mortality (p < 0.0001), but not diarrheal mortality (p = 0.40).
RCTs and natural experiments may help low-income countries better understand the true effect size of vaccination and which infants should receive prioritized access.
While RFK Jr’s analysis is reckless, Malcolm Gladwell’s framing is also problematic. He interviews physicians from Pakistan and India, where rotavirus mortality is high. The mortality difference between countries is not disclosed. This omission is relevant, considering Gladwell’s audience lives almost exclusively in low-mortality countries. Also, RFK Jr's original criticism was directed towards vaccination in the US.
While talking about mortality is more dramatic, Gladwell’s and RFK Jr’s audiences would have come away better informed had they analyzed the impact of vaccination on hospitalizations, the clearest benefit.
Neither do we need dramatic claims to recognize that new RCTs could help refine every country’s guidelines. Which childhood vaccine combinations and schedules produce the largest specific and non-specific benefits? Ongoing vaccination efforts should be guided by RCTs that are powered to detect patient-centered outcomes such as mortality and all-cause hospitalizations (I did not find this in the literature). They could also focus on long-term and uncommon events, possibly overlooked by observational studies and VAERS.1
Improving Scientific Communication
Neither Malcolm Gladwell nor RFK Jr left me feeling informed. A closer look at the evidence reveals relevant gaps in their reporting. And both engage in personal attacks that distract from substantive debate. To restore trust in medicine, we should first accurately represent vaccine science. Everyone—journalists, policymakers, and physicians—should be held to high standards of communication. As HHS Secretary, RFK Jr should be held to the highest standard. Whenever nuance or ambiguity exists, this should be at the forefront, not sidelined.
Critical appraisal can be both provocative and instructive. And debate can inform the public on important vaccine topics. But we must focus on answering the most important questions and leave vendettas behind.
We know that the reporting of adverse vaccine events to VAERS is poor. As a new family medicine-trained attending, I can testify that I was never instructed during my training how or when to report adverse vaccine reactions.
Great to see a nuanced conversation about the pros and cons of specific vaccines in countries with low child mortality. If nothing else, RFK has made it possible to have these discussions. Until now childhood vaccines has been a sacred cow of medicine whose risk/benefit value could never be debated in the same way we debate the value of other therapeutic interventions. I see that as a win for evidence-based medicine.
I can't find any evidence for your claim that the REST trial used a "saline placebo". Where are you getting this from?
The only description i could find in the linked study (or the package insert) is --- " Infants were randomly assigned, in a 1:1 ratio, to receive three 2-ml oral doses of vaccine or visibly indistinguishable placebo, 4 to 10 weeks apart."
"visibly indistinguishable placebo" does NOT mean saline. It might, but i'm willing to bet that like every other childhood vaccine, it's not. Probably the same solution as the vaccine, minus the antigen. This is a common trick to hide adverse events.