“This is loony. This is crazy,” says Arthur Caplan, founding head of the Division of Medical Ethics at NYU School of Medicine in New York City.

He’s talking about Tuesday’s joint statement from Dr. Anne Schuchat of the Food and Drug Administration and Dr. Peter Marks of the Centers for Disease Control and Prevention regarding Johnson & Johnson’s COVID-19 vaccine. Their message: “We are recommending a pause in the use of this vaccine out of an abundance of caution.” Their concern is that after more than 6.8 million doses of J&J’s one-shot vaccine have been administered in the U.S., six cases of an extremely rare type of blood clot, including one death, have been reported. This type of clot requires treatment different from that used for most blood clots; the FDA and CDC issued their statement so health care providers can plan for “the unique treatment required” in those cases.

Is that response ethically right? And what about J&J’s response? The company attracts extra attention because in 1982 it set the standard for an ethical corporate response in cases when a product—Tylenol then—is harming or killing customers; the case has been studied in business schools ever since.

But despite seeming similarities, this case is nothing like Tylenol.

Caplan thinks the government’s response is all wrong. “I wouldn’t pause vaccinations,” he says. “In light of an explosion of cases in Michigan and worldwide disasters in Brazil, Iran, France, even Canada, you don’t want to undermine trust in vaccines. It’s exactly the wrong time.” And what about the risk of blood clots? “I don’t like messaging that doesn’t have comparative data,” he says. His preferred messaging would go like this, with data added: “J&J vaccine has a potential hyper-remote risk of rare blood clots that may be manageable. Here is today’s data on hospitalization, deaths, and burials due to COVID. Keep vaccinating while issue is resolved.”

J&J’s statements have been cautious. The company says “we strongly support” Tuesday’s statement by the FDA and CDC, and it simultaneously announced it will “proactively delay the rollout of our vaccine in Europe.” It also noted that “At present, no clear causal relationship has been established between these rare events and the Janssen COVID-19 vaccine.” Janssen is J&J’s pharmaceutical business.

But the company has not recalled any doses, a sharp contrast with its response in 1982, when it quickly recalled all Tylenol nationwide—its retail value was an estimated $100 million—and ran ads telling consumers not to use Tylenol. That was the ethical thing to do. But in today’s situation, based on today’s facts, it would not be.

In the Tylenol case, seven people died within a few days, and it became clear almost immediately that someone was putting cyanide-laced capsules in Tylenol containers. Today it’s far from clear that J&J’s vaccine has harmed anyone. Six cases in almost seven million patients vaccinated could reflect normal prevalence of the disorder, called cerebral venous sinus thrombosis (CVST).

But what if those few cases don’t reflect normal prevalence? What if investigation shows, as some are already hypothesizing, that the vaccine produces, in a few people, antibodies that trigger an immune response that causes the blood clots? That is, what if there’s a tiny risk that the vaccine might cause CVST blood clots that could, in the worst case, kill a patient? What’s the ethical thing to do?

For the answer, remember Caplan’s preference for comparative data. While Tylenol is a pain medication and comforts people, it doesn’t save their lives. But the vaccine is saving thousands of lives. Which leads to a truly wrenching question: Would ethicists condone risking the potential loss of a few lives per million as a tolerable price for saving thousands of lives per million?

“I would,” says Caplan, who co-edited a book on vaccine ethics in 2017. It would be necessary to inform patients by changing the consent statement they sign, and “you would prepare doctors to handle this type of blood clot,” which isn’t necessarily fatal and gives warning through symptoms including severe headache, abdominal pain, leg pain, or shortness of breath. But “you don’t pull a vaccine in the middle of a plague even if it did kill one or two people.”

Caplan adds one more ethical consideration. When the FDA and CDC pause the use of one vaccine, “the public just hears that vaccines are dangerous,” he says. “There are people out there who are smoking but aren’t getting vaccines because they think they’re dangerous. And if those are the headlines over the coming days, it bodes very, very ill for public health.”