Road Warriors

Columbia Mailman School researchers are mapping a path to healthier highways and safer transportation systems.

November 1, 2019

Cars play a fundamental role in our lives, whether we live in a bustling metropolis like New York City or a small farming town in Iowa. But they also play a key role in our deaths: An average of 101 people die in motor vehicle crashes in the United States each day. (Only opioids are more deadly.) The United States clocked some 37,000 motor vehicle deaths in 2017, and 3 million more people were injured, many of them permanently. The issue affects pedestrians and cyclists, too. Pedestrian fatalities in 2017 reached 6,000—a 27 percent jump over the previous decade.


“Despite our having made substantial inroads into safety over several decades, motor vehicles are still a huge burden on public health,” says Christopher Morrison, PhD, assistant professor of epidemiology. Public health interventions are a key reason that vehicle deaths have declined fairly steadily since the 1970s (seatbelts alone save thousands of lives in the United States each year). But more work by public health researchers like Morrison and others at the Columbia Mailman School will be needed to meet fresh challenges. With the rise of ride-sharing apps like Uber and Lyft and changing laws pertaining to marijuana use—not to mention a fleet of autonomous vehicles on the horizon—the field faces many new questions. Work underway at the School is providing answers, and in some cases already helping to change policy.


Uber turns 10 years old this year, and public health researchers see pros and cons to the explosive growth of ride-sharing. Being able to easily call a car after a night out drinking might drive down alcohol-related car accidents, but early studies have yielded mixed results. “A few cities already subsidize ride-sharing as a potential intervention, but we don’t yet know whether it works,” Morrison says. New York City is the perfect laboratory to study ride-sharing because of how frequently New Yorkers use it. From 2015 to 2017, ride-sharing use jumped almost seven-fold, from about 500,000 to 3 million rides per week. Morrison and his colleagues are now studying GPS data to see whether the number of trips in a particular region of the city correlates with the number of alcohol-related crashes.

But the public health effects of ride-sharing also raise other issues, says Joyce Pressley, PhD, MPH, associate professor of epidemiology and of health policy and management. Seatbelt and child restraint laws typically exempt vehicles for hire, and as ride-sharing ratchets up the overall number of passengers in such vehicles, the risks of riding unrestrained also increase. In a study undertaken with the New York State Department of Health and published recently in the journal Transportation Research Record, Pressley and her colleagues examined New York City motor vehicle crash injuries treated in a hospital or emergency department. Compared with children in other passenger vehicles, kids in taxis were one-tenth as likely to be restrained, twice as likely to be moderately to severely injured, and twice as likely to sustain a traumatic brain injury. Fewer than 6 percent of children in taxis were restrained in a child safety seat.

Pressley, who serves as chair of the Occupant Protection Committee of the Transportation Research Board of the National Academies of Sciences, Engineering, and Medicine, believes the injury and restraint differences may be even more extreme than what they observed, because they were not able to look at ridesharing vehicles. “Our surveillance systems don’t capture whether a private vehicle driven by an Uber or Lyft driver is being used to transport a customer or for personal use,” she says. Nevertheless, based on research by Pressley’s team and other experts, New York Governor Andrew Cuomo supports legislation that would close these gaps. “We believe our evidence is growing, and we are working to get that information into the hands of lawmakers,” she says.


An even bigger wild card for road safety than ride-sharing is the changing nature of drug use. The opioid epidemic spurred Guohua Li, DrPH, MD, professor of epidemiology, to study the role of opioids in motor vehicle crashes. In 2016, his team reported a 700 percent increase in the prevalence of opioids detected in drivers over the previous 15 years. In response to the finding, the Federal Motor Carrier Safety Administration added natural and synthetic opioids to the list of substances for which commercial truck drivers must be tested. More recently, in February 2019, Li published findings in the journal JAMA Open Network that drivers who are prescribed opioids are twice as likely as others to be involved in deadly car crashes.


Changing laws for marijuana use are also a major issue, Li says. In the past few years, some 35 states and Washington, DC, have legalized medical marijuana, and several states have legalized recreational marijuana use. His team is studying the effect of these policies on road safety. One emerging consensus is that consuming marijuana and alcohol at the same time has a synergistically debilitating effect. “But unfortunately, drivers and the general public are increasingly using them together,” Li says. Some bars and clubs are creating cannabis-infused alcoholic cocktails, and “there’s no regulation right now restricting these dangerous drinks,” he notes.

Opioids aren’t the only prescription drugs that can cause road accidents. Last year, Li’s team reported that one in five healthy drivers over age 65 take medicines that should be avoided by older people—and that many of these drugs impair driving. That work was part of one of Li’s major ongoing projects: a 10-year study that aims to address key issues affecting drivers over 65 years of age. The average crash rate for drivers this age is three times that of middle-aged drivers, he says. His study, which follows 3,000 older drivers from five regions across the United States, is exploring the risk factors for accidents, how people manage driving behaviors in the face of cognitive and functional decline, and how they use new vehicle technology like voice control and adaptive cruise control.

Although driving may be riskier in certain respects for older adults, stopping driving has big health consequences as well. Three years ago, Li and his colleagues reported in the Journal of the American Geriatrics Society that the risk of depression in older adults more than doubles when they stop driving, and their risk of death within five years jumps substantially as well. So the idea isn’t to cease driving but to make roads safer while maintaining people’s autonomy, he says. His study has generated specific and practical advice for using simple technological modifications such as voice control and special headlights to make driving safer. It is also informing policy. “We are working with state motor vehicle departments to come up with policies for older drivers that would restrict their license—say, to certain times of the day—rather than revoke it,” Li says.


Policy and investment in infrastructure have the potential to significantly affect road safety, notes Peter Muennig, MD, professor of health policy and management. Over the past few years, he and his colleagues have built computer models of transportation data to explore the cost-effectiveness of changes to urban infrastructure. For example, they found that building bike lanes and installing speed limit enforcement cameras more widely were each enormously cost-effective investments. This research may have contributed to recent legislation in New York that dramatically increased the number of cameras in use. Bike lanes the city built in 2015 reduced the risk of injury to bikers and the pollution they experienced, and boosted the number of people likely to ride and the benefits they gained through exercise, they reported in the journal Injury Prevention.

Currently, his team is taking a close look at New York City’s recently approved plan to charge cars a fee to enter certain parts of the city, starting in 2021—essentially, creating economic incentives to keep cars out. This will be the first such so-called congestion pricing plan to be adopted by a U.S. urban area, and policymakers are watching closely to see whether and how it might be adopted by other cities. In a paper currently under review, their research shows that setting up a simple series of tolls would have the same effect on human life—factors such as air quality and pedestrian safety as well as worker productivity—as developing a more complex high-tech system. “We are hoping our research will play into their calculations of the best approach,” Muennig says. “Policymakers really have to be careful,” he adds. “Often they get a totally different effect from the one they intended.” Case in point: When news of New York’s success with speed cameras broke, legislation to shut down traffic camera programs was proposed in some red states, which look to New York as a model of progressivism to avoid.

Looming in the near distance for public health researchers: autonomous vehicles. “They are coming more quickly than people initially thought,” says Pressley. Many experts think they will make the roads safer, but as recent problems with software in the Boeing 737 Max 8 plane demonstrate, technology is made by fallible humans, and it can misfire. “Autonomous vehicles are not going to be exempt from that,” Pressley says, and researchers at Columbia Mailman School are at the forefront of examining the new technology. Says Pressley, “We’re already trying to anticipate what data we will need to keep abreast of this issue.”

Alla Katsnelson is a contributing editor at The Scientist who has also written for Nature. She has a PhD in mammalian brain development.