COVID vaccines have proved to be magnificent successes, dramatically decreasing the number of cases, hospitalizations and deaths. However, there has been uncertainty about whether vaccinated people who still get infected—perhaps with very mild symptoms, or none at all—might pass on the virus to others. Such silent spread could complicate efforts to control the pandemic.

In recent months, there has been a deluge of data on the risk of transmission after vaccination. These findings have important implications for how quickly we can get the pandemic under control, and for what we say to those who are hesitant about getting vaccinated.

Vaccine trials are typically designed to determine whether an immunization prevents people from getting sick. These are the efficacy numbers in the headlines—as high as a 95 percent reduction in symptomatic COVID cases for the two FDA-authorized mRNA-based vaccines. But the trials provide little or no data on whether the vaccines can entirely block infection, which is the surest way of minimizing spread of the virus.

Considering that at least one third of COVID infections are entirely symptom-free, yet still potentially contagious, it is important to know whether vaccinated people are likely to become carriers of the virus. Not every carrier spreads the virus, though. If a carrier’s viral load is relatively low—meaning that fewer viral particles are shed while breathing and speaking—the risk of transmission is substantially reduced. One possible indirect benefit of a COVID vaccine, then, may be to reduce the viral load in so-called breakthrough cases, or vaccinated people who get infected.

Because many health care workers are now routinely tested for COVID, regardless of whether they have symptoms, much of the early real-world data on vaccine effectiveness in blocking infection has come from this population. In several studies of fully vaccinated health care workers—those more than two weeks past their second dose of either mRNA-based vaccine—the likelihood of having a symptomatic or asymptomatic infection was reduced by 80 to 90 percent, compared with those who were unvaccinated.

There has been good news, too, on the subject of viral load in breakthrough cases. Researchers in Israel studied vaccinated people who became infected. The viral load in these breakthrough cases was about three to four times lower than the viral load among infected people who were unvaccinated. Researchers in the U.K. reported a similar result. They also found that vaccinated people who became infected tested positive for about one week less than unvaccinated people.

We also now have evidence that infected people with lower viral load spread the virus to fewer people, based on contact-tracing studies in the U.S., India and Spain. This is supported by laboratory research demonstrating that nasal samples from infected people with lower viral load are less likely to contain infectious virus.

The Delta variant of the coronavirus, now dominant in the U.S. and many other countries, may induce a viral load that is 1,000 times higher than the level that was typically associated with the ancestral lineage of the virus in early 2020. This higher viral load makes Delta infections more contagious, which has led to a greater number of breakthrough cases, albeit many mild or asymptomatic. But the mRNA-based vaccines still provide strong protection, with efficacy against symptomatic infection in the range of 80 to 90 percent. And the ability of these vaccines to substantially reduce viral load in breakthrough cases could be a valuable tool in containing spread of the Delta variant.

The totality of these impressive data should bolster confidence that the COVID vaccines are extraordinarily effective in reducing transmission of the virus. This suggests that vaccinating a large majority of Americans throughout the country is our surest bet for returning to normal. Entirely eliminating spread of the virus may be an unreachable goal, but mass vaccination—in the U.S. and around the world—will relegate COVID to the background of our lives.

Until now, based on the available data from the vaccine trials, much of the public health messaging around vaccination has focused on the individual benefit. And the efficacy of the COVID vaccines in protecting against illness is indeed remarkable. But with mounting evidence that confirms effectiveness in reducing transmission, it is time to begin emphasizing the societal benefit—and the personal responsibility to avoid harming others.

Particularly because so many COVID cases are asymptomatic, an unvaccinated person is at risk of unknowingly becoming a carrier of the virus. Indeed, those who have the lowest risk of severe illness or death from COVID—the young and healthy—are most likely to serve as these unwitting carriers, because they live the most active lives, routinely coming into contact with many others.

The most important question for those who are hesitant about vaccination, then, is not, “What can the vaccine do for you?” It is, “How many people will you harm if you don’t get vaccinated?”

Early in the pandemic, in February 2020, an international business conference in Boston with about 175 attendees became a superspreading event. It has been estimated that, as of November 1, 2020, more than 300,000 infections in the U.S. had been sparked by that conference alone.

We are all linked in a vast, invisible network. “Six degrees of separation” is not an exaggeration for the surprising density of our multitudinous interconnections. If you decline vaccination and become infected with the coronavirus, the number of people whom you may inadvertently harm could be astronomical.

Mitigation efforts like testing, distancing and masking are imperfect tools for preventing spread of the virus. Vaccination is the closest thing we have to a sure thing in this pandemic. The data are in, and the vaccines are wondrously effective in reducing transmission.

The message to those who remain unvaccinated should be: Don’t let the virus use you to harm others.

This is an opinion and analysis article; the views expressed by the author or authors are not necessarily those of Scientific American.