A randomized trial set up to demonstrate the effectiveness of masks was conducted across 600 villages in rural Bangladesh. Ahmed Mushfiq Mobarak, a Bangladeshi economist at Yale University, discusses the findings with The World’s host Carol Hills
Masks work in helping to prevent the spread of Covid-19. That’s something that global health leaders have been stressing for more than a year. But preliminary results of a massive real-world study involving 340,000 people in Bangladesh are now adding some of the strongest evidence to date on how masks, especially surgical ones, can help communities slow the spread of the virus.
“We use the exact same technique that, say, Pfizer or Moderna used to test their vaccines -- some people get a vaccine, others get a placebo. Some people are getting the encouragement to wear masks and some other villages are not,” explained Ahmed Mushfiq Mobarak, a Bangladeshi economist at Yale University who helped lead the study with the country’s Ministry of Health.
“And just like in the Pfizer and Moderna case, which villagers received it versus which ones didn’t, had to be randomly assigned. That’s the only way for us to generate that gold standard evidence.”
Mobarak spoke with The World’s Carol Hills about the outcome of the study.
So, we’ve got the study set up. Half the villages receive free masks and education, the other half not. So, how different were the behaviours between the two sets of communities?
It turns out that even [in] the first part, getting people to wear masks was not an easy task. What we did was, given the massive scale of the trial with 342,000 people, we decided to just basically throw a lot of different strategies at it.
And then we came up with four strategies, when implemented in unison, [were] persistently increasing mask-wearing. We call it NORM. So, the N stands for no-cost mask distribution. The O stands for offering information. So, we had a nice video that explained the importance of mask-wearing, in which the prime minister of the country, the captain of the cricket team, and the director of the Imam Training Academy -- a religious leader -- all kindly recorded snippets for us.
And then, the M stands for role modeling of leaders. We went to the mosque and did a secondary mass distribution, but worked with the imam of the mosque, so that he would endorse the concept and tell his congregants. Finally, you can’t just distribute masks on day one and leave them. You have to go back and remind people. Basically, we had mass promoters who we tried to hire locally.
They would politely intercept people and say: “Look, we distributed masks to everyone in the village, why aren’t you wearing one? Here’s a replacement mask you can have, but please don’t forget it next time, because we don’t have too many of these to give out.” When we do all four of these strategies together, we saw a tripling of usage that, in fact, stayed persistent beyond the period of intervention.
And how did that affect the rates of Covid-19?
So, what we observe overall is about a 10% decrease in symptomatic seroprevalence. So, that’s our primary outcome.
The 10% overall success you mentioned, is that a lot?
So, the way to think about it is, think about exactly what happened in the first stage. In control areas, some people were wearing masks and in the treatment area, more people were wearing masks. And we didn’t introduce a mandate or anything like that, we didn’t force masks on people.
And just that little bit of a change, the 30% point change in mask-wearing, produced a 10% effect. If we were to move from nobody wearing a mask to everybody wearing a mask, that would have led to a much bigger change.
However, there was heterogeneity across two different types of villages. In some places, we randomly worked with fabric cloth masks, in another set of villages, we randomly chose them to work with surgical masks. We also have engineers on the team, and we did careful testing of masks at the outset. And we found that surgical masks start out at 95% to 97% filtration in efficiency, whereas cloth masks, even well-designed ones with three layers and interwoven fabric, they only start out at maybe 35% to 40% efficiency.
And we saw that even after 20 washes, the surgical masks retain up to about 81% filtration efficiency. The engineering data suggested that the surgical masks were great. So, then what does the health data show us? That overall 10% effect I’m talking about is actually much clearer and bigger for the surgical mask villages. In fact, in the surgical mask villages, 34% of all infections on people over the age of 60 get eliminated. It works particularly well for the elderly.
Now that you’ve learned that masks are important, really important, [are] there any ethical issues with this kind of study where you provide such resources to some communities, but not to others?
Masks were not being withheld from anybody. All we did was added encouragement, and promotion of masks in certain communities. And the reason that we had to do it in certain communities is that, honestly, I just didn’t have the budget to cover a country of 160 million people.
This is why we first received ethical approval. But when the crisis was hitting India, we immediately went into action where we had the first-stage results. We knew that we could get people to wear masks, and through a set of partnerships with the governments in Bangladesh, India, Pakistan, and Nepal, as well as large NGOs, we came up with targeted plans to distribute masks to over 100 million people that are being implemented as we speak. And that’s what I’ve already advocated for masks strongly in this kind of policy, very, very strongly.
This interview first appeared in The World, and can be accessed at https://www.pri.org/stories/2021-09-06/largest-randomized-study-mask-wearing-illustrates-their-importance-preventing.