The Washington PostDemocracy Dies in Darkness

Opinion The Checkup With Dr. Wen: Yes, N95s are still superior to surgical masks

A medical assistant working at a drive-up coronavirus testing clinic in Puyallup, Wash., on Jan. 4. (Ted S. Warren/AP)

You’re reading The Checkup With Dr. Wen, a newsletter on how to navigate covid-19 and other public health challenges. Click here to get it in your inbox.

A new study on masking is making waves by suggesting that N95 masks are no more effective in protecting against covid-19 than surgical masks. This is the wrong conclusion to draw — and can have dangerous consequences.

Let’s look at how researchers designed this Annals of Internal Medicine study. They enrolled about 1,000 health-care workers in four countries (Canada, Israel, Pakistan and Egypt). The participants were divided into two groups of workers caring for covid-19 patients: One that wore the gold-standard N95 respirators, and another that wore surgical masks. All wore N95s in the highest-risk settings (performing procedures that involved aerosolization).

This was a non-inferiority study, meaning that researchers were testing whether surgical masks were inferior to N95s in preventing covid infection. Importantly, the researchers decided at the outset that they would determine that surgical masks were inferior to N95s only if those wearing surgical masks had twice the risk of getting covid than those wearing N95s.

Based on that standard, surgical masks did not prove inferior. During the 10-week trial period, 9.3 percent of the N95 group contracted covid, compared with 10.5 percent of the surgical-mask group.

Submit your questions for Dr. Leana S. Wen’s holiday public health Q&A on Dec. 22

But is this really the right criterion for non-inferiority? The journal’s accompanying commentary pointed out that reducing risk by less than half can still be significant. “In fact, the finding of noninferiority in this trial was consistent with up to a relative 70% increased risk,” the commentary states. A lot of health-care workers — and the public at large — would be happy with this reduced risk.

Then there’s the fact that more than two-thirds of study participants were in Egypt and Pakistan during omicron surges, when community transmission was high. This study does not distinguish where someone contracted covid, and the large number of people infected outside clinical settings might be why the two groups’ infection rates were relatively similar.

A better subgroup is the Canadian enrollees, who entered the study at the beginning of the pandemic, when community transmission rates were low. Among Canadian clinicians, N95s provided a nearly threefold increase in protection over surgical masks.

Yet another limitation of the study is the adherence to mask-wearing. Participants in the surgical-mask group self-reported that they “always” masked 91 percent of the time, versus only 81 percent in the N95 group. A comparable level of mask-wearing could have resulted in more dramatic differences.

Given this study’s many limitations, I think it’s wrong to draw the conclusion that people are equally protected by N95s as they are surgical masks. This is not what the study shows, and it is contradicted by a wealth of research that shows N95s are highly effective at filtering out aerosolized particles.

I really hope this study will not be used to justify arguments that health-care workers do not need N95s if future shortages occur. And I worry that the public will misunderstand the results to mean that a surgical mask is enough to protect against very contagious omicron subvariants. It is not.

Of course, many people are choosing not to mask anymore because they aren’t prioritizing avoiding covid. There are others who cannot tolerate an N95. In those latter cases, a surgical mask would still provide some protection. And some people might need to wear a mask in some settings, not because they want to but because it’s required; if so, a surgical mask would be okay and superior to a simple cloth facial covering.

But for those who are trying to avoid infection, and are fine wearing N95s (or KN95s or KF94s), the science has not changed. They should keep wearing a well-fitting N95 or equivalent mask. These provide the best protection against the coronavirus.

Ask Dr. Wen

Newsletter subscribers are invited to submit questions to Dr. Wen. Not a subscriber yet? Click here to sign up.

“My wife and I are in our 70s. We have all our covid vaccinations, including the new booster, and have never had covid. Our massage therapist just informed us that he will no longer wear a mask during sessions. While we are fairly comfortable in activities where masks are optional, we are concerned about the proximity and length of time (45 minutes) involved in massage therapy. We would appreciate your advice.” — Richard from Virginia

If your massage therapist has covid and doesn’t know it, there is a chance he could transmit it to you during the 45-minute massage session, where I imagine you are in proximity in a small, enclosed indoor space. That chance is much reduced if you wear an N95 mask or something equivalent the entire time.

That said, this activity isn’t too different from, say, an indoor fitness class or indoor restaurant dining. The possible risk of getting covid from your therapist needs to be weighed against the value of continuing the therapy. And again, you can take matters in your own hands by making sure you wear the highest-quality mask during your sessions.

“How safe or unsafe is ride-sharing for a severely immunocompromised 85-year-old person in treatment for blood cancer (myeloma)? Testing early this year showed that my husband had no covid antibodies after three doses of Pfizer. He got Evusheld early this year and has continued with all boosters as available, including the most recent bivalent. He recently got his second dose of Evusheld. Our cancer center is 10 to 15 minutes away. If we had to use ride-sharing because of car trouble, are the risks reasonably low if he remains effectively masked in the back seat, even if the driver is unmasked?” — Kathy from California

There are two issues here. The first is how well your husband is protected from severe disease if he were to contract covid. He just received the preventive antibody Evusheld, which should give him some protection, though he remains very vulnerable due to his age and cancer treatment.

The second is how well he could be protected from infection while in a ride-share with a stranger who would probably be unmasked. A well-fitting N95 mask offers excellent protection even if the driver is not masked. Your husband should make sure to wear his mask during the entire drive. He should not drink or eat while in the car. To further reduce risk, he could ask the driver to open all the windows. If he takes both of these measures, the risk of contracting covid should be very low for a short, 10- to 15-minute drive.

“I have been reading your weekly newsletter since it started, and a while ago, you mentioned that you are not planning to get either of your children the booster at this time, as it’s not clear that protection against hospitalization wanes in children and that a booster would be required. I have a 2-year-old and a 4-year-old who are fully vaccinated (not yet eligible for the booster, but I suspect will likely be soon), and another one on the way in February. Would you change your mind and boost both your children when eligible if you were expecting a baby in the early part of 2023 like myself?” Lisa from New York

Yes, I would consider changing my recommendation in your situation. For a short time, a booster conveys some protection against symptomatic illness. Your baby will be extremely vulnerable to infections in the first few months after birth. I think it’s prudent to reduce potential exposure for your baby during the newborn period, and one way to do that is to boost your other kids about two weeks before your expected due date.

There are other steps you could consider, such as having your older kids mask while they are in day care or school until the baby is at least a month old. I’d also be very careful about all others who have contact with the baby, including visiting family members, and consider mini-quarantines and tests for them. And everyone should wash their hands vigilantly and try to avoid covid and other respiratory infections.

Congratulations in advance (and thank you for reading the newsletter!).

The Post has also compiled Q&As from my previous newsletters. You can read them here.

What I’m reading

Finally, there are real-life data showing that the new bivalent covid booster offers additional protection against symptomatic illness. A report from the Centers for Disease Control and Prevention compared individuals who received at least two doses of the original vaccine with those who subsequently had the bivalent vaccine. Among those 18 to 49 years old who had two or more doses, the bivalent vaccine increased vaccine effectiveness by 43 percent. Among those 50 to 64 years old, the increase was 28 percent. And among those 65 years old and above, it was 22 percent. These effects are significant. Keeping up to date with vaccines will provide enhanced protection against severe illness among those vulnerable to it.

Yet another study finds that hybrid immunity — the combination of vaccination and prior infection — conveys excellent protection against reinfection. This one, published in PLOS Medicine, involved more than 700,000 people in Denmark during periods of alpha, delta and omicron predominance. In all three periods, people previously infected and who were vaccinated had protection against reinfection, with the effect waning the most during omicron.

Is marijuana a “gateway drug” to opioids and other illicit substances? New research in last month’s JAMA Health Forum found an association with cannabis and increased alcohol usage. According to an extensive analysis of recreational cannabis laws in all 50 states from 2010 to 2019, legalizing nonmedical cannabis was associated with a 0.9 percentage point increase in alcohol use in the population overall. “These findings suggest that increased alcohol use may be an unintended consequence of recreational cannabis laws,” the authors write.