Which doctor a person happens to see at a local emergency room can have long-term consequences when it comes to opioid use.

Within the same hospital, some doctors are three times more likely to prescribe an opioid than other doctors, and patients treated by high-prescribing doctors are more likely to become long-term opioid users, according to a study published Wednesday in the New England Journal of Medicine.

“Physicians are just doing things all over the map,” said Dr. Michael Barnett, an assistant professor at the Harvard T.H. Chan School of Public Health and one of the study’s authors. “This is a call to arms for people to start paying a lot more attention to having a unified approach.”

The study looked at how many opioid prescriptions emergency physicians gave to about 377,000 Medicare beneficiaries from 2008 through 2011. The lowest-prescribing quartile of doctors prescribed opioids to just 7 percent of patients, while the highest prescribed opioids to 24 percent — more than three times as often.

Patients who saw a high-intensity prescriber were about 30 percent more likely to end up with a long-term opioid prescription of at least six months within the year following their hospital visit. They were also more likely to return to the hospital in the next 12 months with an opioid-related fall or fracture, a risk factor for seniors who take the powerful painkillers.

Overall, about one in every 48 Medicare patients prescribed an opioid in the study were likely to become a long-term opioid user.

There is a growing consensus among doctors that opioids have long been overprescribed. In 2010, there were enough prescriptions written to supply every American adult with hydrocodone for a month, according to the Centers for Disease Control and Prevention.

Part of the problem, Barnett said, is that there isn’t enough guidance for doctors on when it’s appropriate to prescribe an opioid. Much of the evidence for when they are appropriate comes from small studies sponsored by drug companies.

“It’s kind of a gray area and there’s not very clear evidence around what you should do, so we use our own judgment. And there’s a huge gulf between what one doctor thinks and another,” Barnett explained.

Take, for example, a patient who comes to the emergency room complaining of back pain. There’s evidence that opioids are not necessary in that situation, but many doctors prescribe them anyway, said Barnett. “The world of pain treatment outside of opioids is limited and can take time to figure out. Opioids are an easy fix.”

The problem, he said, is that “even one prescription for opioids carries risks with it, that from my own experience as a provider, we tend to underestimate and under-explain to patients.”

“It is very, very plausible that well-intentioned but perhaps overly aggressive prescribing of opioids makes it likely that a patient will continue a medication long-term even if they don’t truly need it,” said Dr. David Juurlink, a professor of medicine at the University of Toronto. He was not involved in the study. “The doctors in the lowest quartile are the ones whose prescribing we should seek to be emulating.”

The study did not look at whether the opioids were correctly prescribed in each incidence. Dr. Carla Perissinotto, a geriatrician at the University of California-San Francisco, worries that some of the doctors in the lowest quartile might be under-prescribing. “We have to be careful to not make assumptions too quickly and assume they’re bad prescribers, because it could be the opposite,” said Perissinotto, who also was not involved in the study.

Usually, a patient is prescribed just a handful of pills by a doctor at the emergency department to tide them over until the patient can visit his or her primary care physician.

But many primary care doctors simply refill the opioid prescription for another 30 days or longer, a phenomenon Barnett calls clinical inertia. “There’s this cognitive bias to keep going with the flow especially if the patient still feels they’re in pain.”

That can have long-term implications: One-third of people who have taken prescription opioids for at least two months say they became addicted to or physically dependent on them, a recent Washington Post-Kaiser Family Foundation survey found.

“We know there is a population of people who will potentially get addicted if they’re exposed to an opioid,” said Dr. Lewis Nelson, chair of the department of emergency medicine at Rutgers New Jersey Medical School, who was not involved with the study. “It’s a numbers game. The more people you expose, the more people you are likely to hit in that population likely to get addicted.”

Some hospitals are starting to help doctors prescribe opioids more judiciously. Many of those efforts have taken place after 2011, and therefore any changes would not be seen in the New England Journal of Medicine study.

For the past year, St. Joseph’s Regional Medical Center in New Jersey has been instructing doctors to prescribe opioids only as a last resort, said Dr. Mark Rosenberg, who runs the hospital’s emergency department and is on the board of the American College of Emergency Physicians. Rosenberg helped institute protocols that have doctors try other methods of pain relief first, such as a Novocain injection. In just over a year, the hospital has managed to reduce the number of opioid prescriptions written in the emergency room by 50 percent.

New York City and Washington state have also tried to introduce opioid-prescribing guidelines into hospitals, and the ACEP expects to release new nationwide guidelines for emergency physicians in 2018.