A vaccine created by Pfizer and partner BioNTech won the race to prove efficacy in a large trial. But is it the best vaccine of the dozens being tested? If it isn’t, will any slower but better vaccines have a chance?

A press release this week claims the vaccine is 90 percent effective at preventing symptomatic COVID-19. That would mean nearly all of those infections happened in people who got a placebo rather than the vaccine.

The trial isn’t over. It was slated to stop after 164 people of the 44,000 volunteers developed symptoms and tested positive for the virus, but an independent board was allowed to peek at the data early. They waited until they were up to 94 infections to share the preliminary news, according to an article in the medical website STATnews.

As promising as the announcement sounded, a critical question wasn’t answered: Does the vaccine prevent people from getting the infection and spreading it to others, or merely prevent symptoms? In experiments in which animals were deliberately exposed, some vaccine candidates did prevent infection, while others did not.

That makes an enormous difference, since vaccines are usually least likely to work in those who need them most — people whose immune systems are not working correctly because of their age or health problems. A vaccine that prevents younger, healthier people from transmitting the disease would save more lives by indirectly protecting those people through herd immunity. Safety will matter too — people might not agree to get a vaccine if there’s even a perception of serious side effects or incomplete safety testing.

This creates a big ethical puzzle. After it’s clear a drug or vaccine is working, it’s traditionally considered unethical to continue to give a clinical trial volunteers a placebo.

Another unanswered question is whether it will be possible to continue clinical trials of other vaccines. And if one later proves better than the front-runner, will it be safe to use it on people who’ve already had a vaccine with more modest benefits?

“The big discussion now is how we make the vaccine available without losing the data we need,” says immunologist Florian Krammer of the Icahn School of Medicine at Mount Sinai. How scientists solve that problem will determine when a vaccine is ready to be rolled out. Even before full scale approval, the FDA can fast track vaccines through so-called emergency use authorization (EUA), which could be issued for the Pfizer vaccine before the end of this month.

“What’s being discussed now is whether an emergency use authorization of the first vaccines will make it more difficult for other vaccines to complete clinical trials,” says Harvard researcher Dan Barouch, the co-creator of a different vaccine being manufactured by Johnson & Johnson and being tested in 60,000 people.

On the one hand, an EUA could prevent thousands of deaths. On the other, even more deaths could be prevented over the long haul by gathering enough data to get the safest, most effective vaccines out there.

Pfizer’s front-runner vaccine is one of two leading candidates based on a new technology that uses messenger RNA — a nucleic acid, like DNA. It sounds a lot scarier than it is, say the experts. The RNA can’t affect the genes in your cells. And its effect on the body is minimal compared to getting the virus. An infection hijacks all your cellular machinery and repurposes it toward making more virus, while the RNA would prompt cells to make one small part of the virus, stimulating the immune system to create antibodies that will attack the real thing if it’s encountered.

Rheumatologist Arthur Krieg, founder of Checkmate pharmaceuticals, isn’t worried that the unprecedented speed would hurt a vaccine’s safety. “What they changed was to dramatically accelerate the review and decision-making process, allowing trials to go much more rapidly than they normally would,” he says. “But we’re not really making a sacrifice in safety as far as I can see.”

In past vaccine trials, the vast majority of side effects occurred within the first two months of participants getting the shots, and most within the first two weeks, Krieg says. Pfizer should have that safety data by the third week in November, when they have two months of follow-up on at least half of the participants.

Still, the whole vaccine quest could be derailed by even the perception of danger. Once enough people are getting a vaccine, some will inevitably get various severe illnesses for reasons that have nothing to do with the vaccine. The scare that wrongly associated autism with childhood vaccines came about that way.

Mount Sinai’s Krammer says he has high hopes that a vaccine could end the pandemic, even though he thinks it’s unlikely to eradicate the virus. What he expects is that eventually a vaccine will make COVID-19 milder and less common. If it’s not severe and there’s very little of it, then it doesn’t matter, he says, and we’re basically back to normal.

This week’s news is promising for other vaccines. One being developed by Moderna also uses messenger RNA, and most of the others are, one way or another, aimed at producing antibodies to the so-called spike protein, which the virus uses to get into cells. Eleven vaccine candidates are in late-stage testing and a number of other promising vaccines aren’t far behind. It’s an abundance that’s never happened in the history of vaccines. Balancing urgency with research ethics and the longer-term need for data won’t be easy. But so much rides on getting it right.

Faye Flam is a Bloomberg columnist.