Sarah Holder is a staff writer at CityLab covering local policy, housing, labor, and technology.
In cities where the popular ride-sharing app is used, ambulance usage went down by 7 percent, according to a new study.
In cities where ride-sharing apps proliferate, patients are potentially saving thousands of dollars and alleviating the burden on emergency systems by traveling to the hospital in an Uber instead of an ambulance, according to a new study.
Economists at the University of Kansas analyzed ambulances deployed in the more than 700 cities where UberX cars are available, and found that in the time after Uber entered a given city, there was at least a 7 percent decrease in rate of ambulance use.
The study chose the Uber brand specifically because “it’s the first and the biggest,” but analyzing the impact of other ride-sharing services like Lyft and Via would generate similar findings, says David Slusky, assistant professor of economics and one of the co-authors of the report.
Using data on ambulance use from the National Emergency Medical Services Information System paired with Uber’s own data, the economists analyzed deployment rates starting from the date of Uber’s entry into each district until the end of 2015. The research, which has not yet been peer-reviewed, looked at how ambulance rates changed between the months before Uber's entry and the months after.
The authors infer from the drop in usage that now, for some trips to the hospital, people are taking Uber when they would have otherwise taken an ambulance. That doesn’t mean traditional ambulances will—or could—become obsolete. While more data on this has not yet been analyzed, Slusky suspects that as ambulance usage decreases, wait times for the available ones will fall. When more people with less urgent medical needs take private cars, ambulances are freed up to serve the most vulnerable.
Ambulances serve a unique purpose that can’t be disrupted entirely by an app: They’re outfitted with specific medical machinery, equipped with trained staff, and able to disobey the laws of traffic to get to emergency rooms fast. When dealing with cardiac arrests, head wounds, and seizures, it makes sense to stick with the pros.
But as the volume of 911 calls increases, often for non-urgent emergencies, ambulance demand has outpaced supply. The expansion of Medicaid under the Affordable Care Act slowed response times nationwide by an average of 19 percent: Today, ambulances in the U.S. take an average of 8 minutes to arrive in urban areas, and 14 in rural ones. In some cities and neighborhoods, the wait times can be far longer. And in critical cases, those minutes matter.
Ambulance rides are also expensive, especially for the many Americans whose health insurance includes high deductibles: One trip could cost more than $1,000, paid out of pocket.
Inequities in transportation access already lead to health disparities for low-income patients, which the high cost of ambulances only exacerbates. Those who don’t have a car might use slower forms of public transportation in lieu of incurring exorbitant ambulance costs. And those who live far from a high quality hospital might choose whatever’s closest, even if the treatment they receive there is worse.
“If we had infinite ambulances, we could take everybody in them, and therefore if somebody had a situation where an ambulance would save their life they’d always be in one,” said Slusky. But hospitals don’t.
What some cities do have, however, is a cheaper, faster way to travel. In 2015, a product designer reported on Medium that in New York City, the median wait time for an ambulance was 6.1 minutes, versus an Uber’s approximate three, depending on the borough. Riffing off those stats, Jimmy Kimmel joked that Ubers should start offering “Ubulance” services. While the company itself hasn’t adopted any protocols since, judging from the University of Kansas’s results, patients have taken matters into their own hands.
The U.S. has already adopted a patient-centric form of health care, in which consumers are in charge of many of their own medical decisions, including scope of treatment and cost. As the availability of ride-sharing apps has increased, patients are also able to choose the mode of transportation they take to and from the hospital.
“I’ve heard a lot of people say, individuals shouldn’t be making these decisions for themselves—they should be calling 911,” said Slusky. “And my answer to that is, well then why are they letting them have all this skin in the game and making all these cost decisions themselves? We’re already way down that route.”
The study’s association between more Ubers and fewer ambulances, while significant, could be explained at least in part by other factors. After Uber entered California, for example, researchers observed a decrease in fatal drunk driving accidents and DUIs—by reducing the number of emergencies, those phenomena could impact ambulance usage. That assumption has since been challenged, however, by another nationwide survey of metropolitan counties with Uber between 2005 and 2014, in which no such association was found.
Cities, hospitals and insurance companies can nudge people into cars through partnerships with ride-sharing companies. Last summer, Washington, D.C. emergency services considered implementing a new system to bring “low-priority patients” to doctors instead of emergency rooms, using Uber instead of an ambulance. Memphis is experimenting with diverting some 911 traffic by sending doctors to patients’ homes instead of transporting them all the way to the hospital. And some insurance companies already give patients positive incentives like $50 gift cards for using cheaper MRIs: Slusky envisions similar benefits for ride-sharers.
On the hospital level, 911 dispatchers themselves have the power to begin telling callers, on a case by case basis: You don’t need an ambulance—call yourself an Uber.
“At the emergency room, triage systems like this already exist. You show up at the emergency room, and they tell you to wait,” said Slusky. If you can sit for three hours without your condition getting worse, someone else is pushed to the front of the list. “It’s not illogical at all to move it one step back such that it’s dealt with [before],” he said.
The study does not include a control group of cities who didn’t get Uber in the seven-year block. By 2017, the app is more ubiquitous than ever—5 billion rides have been given, across 76 countries and 450 cities, leaving open some possibility that people called ambulances less frequently for other reasons nationwide. Slusky says that the granular quarterly analysis tempers that complication. The decreases in ambulance use were associated not with year or month, but with the times of Uber’s penetration, which vary by city.
And since ride-sharing apps are most prevalent in the suburbs and in urban areas, this trend would also leave out rural towns where public transportation and ride-sharing infrastructure are both lacking. There—and elsewhere—hospitals could offer modified ride-shares of their own, said Slusky. “You could imagine a future where emergency medical services has regular cars that don’t violate traffic cars, driven by someone who’s an EMT,” he said. “Just a car, driven by somebody who’s better trained in CPR.”
This article is part of our project, “The Diagnosis,” which is supported by a grant from the Robert Wood Johnson Foundation.