During the Pfizer/BioNtech and Moderna phase 3 vaccine trials, 7 instances of facial paralysis were noted out of 35,654 recipients who received the real vaccine. In contrast, just 1 such instance was reported among the 35,611 who received a placebo shot. While to the casual observer, that may seem like a big difference, the reality is that the numbers were too small to be interpretable. Now that millions of people have gotten the vaccines, researchers assessed rates of facial paralysis and facial weakness among recipients of the mRNA vaccines during the early rollout. The results were published today in JAMA Internal Medicine.
The vaccines were not found to impart a higher rate of these adverse events. In order to determine this, the researchers did something interesting. They compared the rates of facial paralysis and weakness to rates among influenza vaccine recipients and other viral vaccines. The reason for this is, in essence, two-fold. First, baseline rates of conditions like facial paralysis are extremely hard to know. Many of these conditions are temporary (most cases of "Bell's Palsy, a weakness of one side of the face) improve spontaneously or with medications. Moreover, it's not as though every single clinician who diagnoses one of these conditions reports it into some central database, the way public health officials must do with causes of death. However, virtually all side effects that people experience after a vaccination are reported to centralized databases. (If anything, swaths of random unrelated symptoms are often over-reported after vaccinations; people have random events and blame the vaccines, whether rightly or wrongly.) As a result, we have a much better idea of how common these conditions are amongst people who recently received an influenza or other viral vaccine than we do about the rates of these conditions occurring in the general population. Secondly, comparing the rates of facial paralysis and weakness to the rates that are associated with other vaccines to the rates that occurred after SARS-CoV-2 inoculation provides an important benchmark. If we are willing to accept some rate of temporary facial paralysis when we go to get our flu, shingles, or measles shot, it stands to reason that we'd accept a similar rate when protecting ourselves from SARS-CoV-2.
The researchers found that the coronavirus mRNA vaccines were associated with similar rates of facial nerve problems compared to other viral vaccines. Additionally, the coronavirus mRNA vaccines were associated with lower rates of facial nerve problems than has been seen among influenza vaccine recipients.
The takeaways are clear. First, the phase 3 coronavirus vaccine trials gave us important information on possible associations with some important side effects which needed to be tracked. This new study provided just such follow-up information. It reminds us that the hint of an adverse effect signal in a clinical trial does not necessarily correlate to real findings in larger datasets. Second, the authors have provided us with some insightful reasoning regarding our own implicit risk tolerance. The achievable goal, as they imply without saying directly, is not to create vaccines with absolutely no risk of adverse events. The goal is to create vaccines with tolerable effects that are in line with our usual risk tolerance when we set out to prevent deadly diseases.