SAN JOSE, Calif. – In what is billed as the first study to measure the impact of Uber and other ride-booking services on the U.S. ambulance business, researchers have concluded that ambulance usage is dropping across the U.S.

A research paper released this week examined ambulance usage rates in 766 U.S. cities in 43 states as Uber entered their markets from 2013 to 2015.

Co-authors David Slusky, an assistant professor of economics at the University of Kansas, and Dr. Leon Moskatel, an internist at Scripps Mercy Hospital in San Diego, said they believe their study is the first to explain a trend that until now has been discussed only anecdotally.

Comparing volumes before and after Uber became available in each city, the two men found that the ambulance usage rate dipped significantly.

Slusky said after using different methodologies to obtain the "most conservative" decline in ambulance usage, the researchers calculated the drop to be at least 7 percent.

"My guess is it will go up a little bit and stabilize at 10 to 15 percent as Uber continues to expand as an alternative for people," Moskatel said.

Slusky said he and Moskatel are submitting the paper to journals for peer review.

San Francisco-based Uber quickly distanced itself from the notion that hailing an Uber driver is an acceptable substitute for calling an ambulance.

"We're grateful our service has helped people get to where they're going when they need it the most," said company spokesman Andrew Hasbun. "However, it's important to note that Uber is not a substitute for law enforcement or medical professionals. In the event of any medical emergency, we always encourage people to call 911."

Moskatel said many patients "tend to be pretty good at assessing their state and how quickly they need to come in and how sick they are."

But at least one prominent Bay Area emergency room physician disagreed.

Paul Kivela, president of the 37,000-member American College of Emergency Physicians, said he believes that for those low-risk patients who can't drive themselves to the emergency room, Uber is a good service.

But many people, he said, may not be able to differentiate between a life-threatening emergency and an innocuous medical issue. So, he said, calling 911 is always the safest bet.

"A paramedic has the training and the ability to deliver lifesaving care en route," Kivela said. He noted that in addition to his work as an ER doctor, he is also the medical director of an ambulance company in Solano County, Calif.

The researchers, however, insisted that ride-booking services such as Uber and San Francisco-based Lyft can sometimes be the best way to get to the hospital in a hurry.

Previous research, Moskatel said, "suggests that a fair number of people are using ambulances to get to the hospital because they simply don't have another way to get there" — especially those who live in areas with limited taxi service.

And, Slusky added, with health care taking a big chunk out of most people's budgets, many consumers these days have to weigh a few factors before calling an ambulance.

"They have to think about their health — and what it's going to cost," he said. "And for many of us with high-deductible plans, an ambulance ride would cost thousands of dollars."

Moskatel and Slusky said they focused on Uber because they needed a broad set of data and the company has operated for a longer period of time than competing services.

Ambulance rates were obtained from the National Emergency Medical Services Information System, a national repository for emergency medical services data.

Slusky said an agreement between the system and its members prevents release of information that could be used to identify rates in specific states, cities or ZIP codes.