Handicapping the COVID Vaccine Horse Race

F. Perry Wilson, MD, MSCE


February 01, 2021

Editor's note: Find the latest COVID-19 news and guidance in Medscape's Coronavirus Resource Center.

If ever there was a testament to the power of the scientific method, it is this: barely a year after SARS-CoV-2 hit our shores, we have two US Food and Drug Administration–authorized vaccines to prevent COVID-19. And from the looks of it, another is on the way.

Johnson & Johnson (J&J) on Friday published topline results of the ENSEMBLE trial, which examined the efficacy of its COVID-19 vaccine, touting an overall 66% efficacy rate.

Although not the 90% to 95% efficacy reported for the Pfizer and Moderna vaccines, 66% is certainly nothing to sneeze at. In fact, a single-dose vaccine that can stay in a refrigerator for 2 months is truly a game changer. Although the United States has sufficient infrastructure to manage the ultracold conditions needed for the Pfizer vaccine and the very cold conditions needed for Moderna's, many less developed nations do not have this capability. Additionally, a "one and done" shot dramatically increases the feasibility of mobile clinics, vaccine fairs, and direct outreach to rural individuals.

To understand the impact of this third coronavirus vaccine option, I want to compare it with the existing two across a number of domains: mechanism, efficacy, side effects, and the likelihood of vaccine escape. But a caveat here: I am working only from a press release from J&J. These data have not been peer reviewed, and there are many dark corners that we simply can't explore until we have more information.

Mechanism of Action

The Pfizer and Moderna vaccines are based on mRNA technology ― well established in research, but never before used on such a wide scale. The mRNA instructs human cells to produce the spike protein of the coronavirus, which binds the virus to cells, eliciting the production of antibodies.

The advantage to mRNA vaccines is the speed with which they can be developed. The downside? mRNA is fragile, prone to breaking down at room temperature into useless bits of sugar and acid.

The J&J vaccine is not your typical vaccine, either, relying on adenovirus vector technology. Again, this technology is not new, but this will be its first wide-scale test. The vaccine is a modified version of adenovirus 26; a specific gene has been removed to ensure that it cannot replicate in human cells. In its place, the gene for the SARS-CoV-2 spike protein is inserted. This complex is grown up in a bath of cells that have been engineered to express the missing gene, allowing the modified adenovirus to replicate. The virus is then purified and injected ― spike protein and all ― and provokes what appears to be a strong immune response.

The upside of this approach is its stability; no ultracold storage conditions here. One theoretical downside? Adenovirus 26 is a virus some of us may have had ― it is a cause of conjunctivitis ― and it is not yet clear whether prior infection with adenovirus 26 may attenuate the effect of the vaccine.



The startling efficacy of the mRNA vaccines was the high point of coronavirus news in 2020. Vaccines with greater than 90% efficacy are extraordinarily rare; to have two in such a short time bordered on miraculous. What to make, then, of the 66% reported efficacy for the J&J vaccine?

Most importantly, the vaccine efficacy numbers you hear are based on the prevention of symptomatic COVID-19, not severe COVID-19. The benefits of vaccination are not simply the prevention of symptoms. The true benefits come from reduction of hospitalizations, deaths, and further virus transmission. We don't know about the available vaccines' ability to reduce viral transmission, but we do have data regarding the clinically critical outcomes of hospitalization and death.

The Pfizer vaccine is 89% effective at preventing death and hospitalization. Moderna – 100%. J&J reports "complete" protection against hospitalization and death 28 days after vaccination, implying (without giving us the raw numbers) that no such events occurred in the vaccine group at that time interval. J&J notably did not provide the count of any such events in the placebo group, which would really help drive home the significance of those results.


Still, it is clear that even a vaccine that is imperfect at preventing all symptoms may be excellent at preventing hospitalization and death ― and that is a broad victory.

A brief aside: If hospitalizations and deaths are so important, why weren't they used as the primary outcomes in the vaccine trials? Expediency. Hospitalizations and deaths are (fortunately) uncommon among those with COVID-19. We thus would be waiting longer for enough of these outcomes to occur before we could estimate how effective the vaccine is. Using symptomatic infection is a decent surrogate marker ― the idea is that if a vaccine decreases symptoms, it almost certainly would also reduce hospitalizations and deaths. So far, the data have borne out that assumption.

Even if we focus on symptomatic infection, it's not exactly fair to compare the efficacy of the J&J vaccine with that of the mRNA vaccines. There are a lot of apples and oranges here. First, the J&J vaccine is a single-dose vaccine; were patients to receive a booster of the J&J vaccine, efficacy likely would increase. A single dose of the Pfizer or Moderna vaccines probably provides about 80% efficacy, given the data we have.

Another major issue is timing. The mRNA vaccines were studied earlier in the pandemic, before rampant infections gave rise to multiple potentially vaccine-resistant variants. Indeed, we can attribute virtually the entire underperformance of the J&J vaccine in South Africa to the B.1.351 variant endemic there.

Side Effects

J&J didn't release much data about adverse events except to report that the vaccine was "well tolerated," there were no cases of anaphylaxis, and severe adverse events were more common in the placebo group compared with the vaccine group. All good news, I suppose ― but few hard numbers. J&J reports a fever rate of 9%. You can compare that to a rate of around 15% after the second dose of the mRNA vaccines.


Vaccine Escape

Despite breathless news stories of each new coronavirus variant, it is worth mentioning that SARS-CoV-2 is not a particularly rapidly mutating virus. By the standards of influenza, this coronavirus is a staunch conservative. But hundreds of millions of infections worldwide have provided hundreds of millions of opportunities for new mutations, and some of these may be able to evade the immune surveillance that the vaccines engender.

So far, there are two sources of data that allow us to evaluate whether a variant may escape a vaccine. The first uses serum from vaccinated individuals. These neutralization assays ask whether the antibodies in that serum will keep the virus from growing in culture. So far, both Pfizer and Moderna show good results against the British B.1.17 variant. The Pfizer vaccine performs a bit better against the South African variant, and Moderna a bit worse, though it's still likely effective. These estimates probably underestimate true efficacy, because only the protection afforded by antibodies was evaluated, whereas vaccines lead to both humoral (antibody-based) and cellular immunity.

The other way to figure out whether a vaccine protects against a variant is to observe the variant infection rate among vaccinated people. Because the Pfizer and Moderna trials data came out before the variants were identified, we can't speak to this ― though they no doubt are still collecting these data on their participants. The J&J data, on the other hand, suggest diminished efficacy ― the protective effect in South Africa (where 95% of the infections were due to B.1.351) was just 57%.


In the end, we need to accept that variants will arise, and it will be hard to predict which vaccine might be less effective against any particular variant. My hope is that all vaccines will provide at least partial protection against variants; converting deadly COVID-19 to a bad cold or flu is a fine proposition.

Additionally, we should anticipate that we will need vaccine updates to account for new strains, much like we have for influenza currently. The mRNA vaccines have the advantage here in that an update simply requires adding the new sequence to the production line.

What to Do

We will soon have three vaccines with subtly different effect and side-effect profiles. Although supplies are scarce, the decision as to which vaccine to get is simple. Anyone eligible for a vaccine should take whatever they can get.

Soon, however, multiple options may be available, and I'll offer the following guidance. Where choice exists, the highly effective mRNA vaccines should be given to those at high risk for death from COVID-19: the elderly and those with certain preexisting conditions. The J&J vaccine may be ideal for those at high risk for exposure to COVID-19 but who are nevertheless at lower risk for bad outcomes: think frontline and essential workers.

This calculus might change when we know how these vaccines may (or may not) prevent asymptomatic transmission, but for now, this seems a reasonable first approach. Of course, a choice of vaccines is an extremely good problem to have.


Perry Wilson, MD, MSCE, is a nephrologist at Yale University and has a master's degree in clinical epidemiology. He frequently opines about clinical study design and interpretation, including his video series on Medscape, Impact Factor.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: