Editor’s note: This commentary is by Dr. Turner Osler, a career academic trauma surgeon at the University of Vermont Medical Center turned research epidemiologist.
Over a 50-year career as a surgeon and researcher I’ve had a front row seat for the many advances in medicine that have saved lives and revolutionized how doctors practice their craft. So it’s shocking to admit that our most primitive technology, the face mask, is our best protection against Covid-19. And this anachronism will likely persist even after vaccines become available. The evidence from laboratory simulations and real-world reports is clear: for species from hamsters to humans, masks interrupt viral transmission and thus prevent disease spread. But for masks to save us from the current pandemic they must be worn, and so understanding who is wearing masks, and more importantly who isn’t, is crucial in the fight against Covid-19.
Epidemiologists are encouraged that several national surveys show 80% of Americans are wearing masks “always or frequently,” a rate slightly higher than that of Taiwan, France or Canada, and far higher than Scandinavian countries where wearing masks is in the single digits. But while 80% sounds good, it represents an overall average, and hides the fact that in many areas Americans are masking up at worrisomely low rates. Such areas are ripe for a resurgence of Covid-19, putting the health of all our citizens, and our economy, at risk.
So, where are masks not being worn? A recent survey of 250,000 people by the New York Times and Dyanata gives valuable insights into our mask-wearing habits. Broadly, masks are worn in the Northeast and West, but often eschewed in the Plains and South. To better understand the factors driving mask-wearing we merged the New York Times survey data with county level data for a number of factors. This is what we found:
Population density is a powerful driver of masking. In the most densely populated counties 85% of people wore masks, a rate that fell to 65% in the least populated counties. People in counties with older inhabitants were more likely to wear masks. Additionally, counties with the highest numbers of Covid-19 cases, and especially deaths, had higher rates of mask-wearing. All of these observations are expected, because people care greatly about their own health and that of their loved ones, and very accurately make day-to-day decisions to keep everyone safe.
One final finding surprised us, however. In counties that favored Trump in the last election people were less likely to wear masks, and the greater the margin for Trump the fewer masks are worn. This effect was pronounced: in counties that went avidly for Trump people had a 65% probability of wearing masks, compared to a 90% chance in counties with the least support for Trump. This “Trump effect” persisted when controlled for other factors; indeed, the “Trump effect” was by far the strongest predictor of failing to wear a mask. In this sole instance political affection seems to have overridden rational choice. While we don’t know why this happened (and is continuing to happen), we can speculate: decisions are based on citizens’ understanding of facts, and so a systematic campaign of disinformation can lead to irrational, even dangerous, choices. The efforts to discredit masks have had fatal consequences for thousands of Americans, one more unfortunate result of our current fractious political climate.
The haiku was the tweet of 17th century Japan, and may still be the best way to summarize our findings:
In these COVID times
It is so hard to mask up
Without losing face
David W. Hosmer, Ph.D., professor emeritus of biostatistics at the University of Massachusetts, and Stas Amato, MD, MSc, of the Department of Surgery at the University of Vermont, were involved in the data analysis.
