What’s the right painkiller prescription to send home with a patient after gallbladder surgery or a cesarean section?

That question is front and center as conventional approaches to pain control in the United States contribute, in the opinion of some experts, to a culture of overprescribing that aggravates the nation’s epidemic of opioid overuse and abuse.

Last year, Marty Makary, a surgeon, researcher and professor of surgery at Johns Hopkins School of Medicine in Baltimore, wondered why the answer wasn’t clearer. Even he admits that for most of his career he gave painkillers out “like candy.”

So he took an innovative approach toward developing guidelines: matching a right number of opioid painkillers to be prescribed for each of many procedures — a substitute for the one-size-fits-all recommendations that doctors have usually followed.

In December, he gathered a group that included surgeons, nurses, patients and others, and asked them, “What should we be prescribing for operation X?”

No one had a precise response. Dr. Makary didn’t know. Nor did the resident in the group. The nurse practitioner, the person who most often follows up closely with patients, said the answer would vary.

“Wow,” Dr. Makary remembers thinking that day. “We’re the experts, the heads of this and that, and we don’t know.”

After a couple of weeks of intense discussion, however, Dr. Makary’s group reached consensus and gave its blessing to guidelines setting maximum numbers of opioid-containing pills for 20 common medical procedures.

In some cases, the right number of opioids is zero, the group concluded. Indeed, it recommends no opioids for patients heading home after uncomplicated labor and delivery, or after cardiac catheterization, a procedure in which a thin, hollow tube is inserted into the heart through a blood vessel to check for blockages.

For certain types of knee surgery, such as arthroscopic meniscectomy, the guidelines recommend no more than 12 pills upon discharge, while a patient going home after an open hysterectomy could require as many as 20.

Optimally, Dr. Makary said, “no one should be given more than five or 10 opioid tablets after a cesarean section.” And for cardiac bypass surgery? No more than 30 pills.

How to Address the Pain?

Tens of thousands of Americans are dependent on opioid medications. An increasing number are dying from overdoses, both from prescription medication and street drugs. And many experts view post-surgery opioid prescription painkiller use as a gateway to long-term use or dependence.

A study published last year in JAMA Surgery concluded that persistent use of opioids was “one of the most common complications after elective surgery.” In that study, University of Michigan researchers found that 6 percent of people who received opioids for the first time after surgery were still taking them three to six months later.

With about 50 million surgeries in the United States each year, “there are millions who may become newly dependent,” said Chad Brummett, an associate professor of anesthesiology at the University of Michigan Medical School, who was the study’s lead author.

Smokers and those diagnosed with conditions such as depression, anxiety or chronic pain before their operations were found to be most at risk of long-term use. And other studies have shown that each refill or additional week of use makes for a greater risk of misuse.

Further research points to another reason for concern: If patients don’t take all their prescribed pills, the leftovers can be stolen or diverted to other people, who then run the risk of becoming dependent.

Still, there is debate in medical circles about just how effective recommendations and guidelines will be in stemming the epidemic. For one thing, some experts worry that if focusing on safe prescriptions comes at the expense of seeking alternatives to opioids, it will miss safer opportunities.

“Are there better methods than opioids in the first place?” asked Lewis Nelson, chairman of emergency medicine at Rutgers New Jersey Medical School. “Could you put a lidocaine patch over the wound, or is there a better way to immobilize a joint?”

Studies have shown that sometimes a combination of ibuprofen and acetaminophen can be just as good as — or better than — opioids. Dr. Makary agreed that alternatives should always be considered first.

Another concern is that guidelines for prescribing relief — even those aimed at acute, short-duration pain like what often follows surgery — have carry-over effects on patients with long-term pain.

The worry is prescribing limits will have the unintended consequence of also making it difficult for patients with chronic, long-term pain to get the medications they need.

A Different Focus: Duration

Lawmakers — desperate to address overdose problems — have begun doing something they usually avoid: setting specific rules for doctors.

Legislatures in more than a dozen states, including New Jersey, Massachusetts and New York, have set restrictions on the number of days for which supplies of pills can be prescribed for acute pain.

“States said that since physicians haven’t self-regulated, we’re going to do it for them,” Dr. Nelson at Rutgers said.

Congress, too, has held hearings and is considering similar legislation. The recently passed federal spending bill includes $3 billion in new funding to help states and local governments with opioid prevention, treatment and law enforcement efforts.

To be sure, the medical profession has also responded to the crisis, with medical societies and other expert groups offering a growing number of standards for prescribing opioids. Some recommend the lowest dose for the shortest period of time for acute pain. Others are more prescriptive. None is meant to address the needs of chronic pain patients or those with cancer.

And state rules vary. New Jersey’s, for example, says patients with acute pain should initially get no more than a five-day supply, while Massachusetts sets the cap at seven days for a patient being prescribed opiates for the first time. The Centers for Disease Control and Prevention, on the other hand, recommends three days.

Dr. Makary and some other experts say that while well intentioned, such durational rules are too blunt. A day’s worth of pills can vary, depending on how often the doctor instructs patients to take them. Under many state rules, patients could still head home with more than 50 pills.

“No one should have 50 tabs sitting in their medicine cabinet” for acute pain, Dr. Makary said.

Andrew Kolodny, a co-director of opioid policy research at the Brandeis University Heller School for Social Policy and Management in Massachusetts, supports guidelines but wants states to take their rules a step further, requiring that physicians warn patients that a drug is addictive if it is prescribed for more than three days. That would create a disincentive to prescribing more than three days’ worth of opioid painkillers, he added, and leave patients better informed about the dangers.

Dr. Nelson at Rutgers, who sat on the panel of the Centers for Disease Control and Prevention that developed recommendations, said durational rules — like those adopted by the states — can be effective, but he also called the Johns Hopkins approach an “excellent idea” that he has tried to implement. “It’s a lot harder than it sounds because of the large number of procedures and the diversity of patient needs,” he said.

To get around overprescribing, or setting one-size-fits-all guidelines, physicians at Dartmouth-Hitchcock Medical Center in New Hampshire have a developed their own data-based approach.

Richard Barth, the chief of general surgery at Dartmouth, and colleagues studied 333 patients discharged from the hospital after six common surgeries that included bariatric procedures; operations on the stomach, liver, colon and pancreas; and hernia repair.

They asked the patients how many opioid pills they went home with, how many they took, how many were unused and how much pain they experienced. The data helped them develop a way to recommend a specific number of pills.

“If they took none the day before discharge, then over 85 percent of patients did not take any when they went home,” Dr. Barth said. Dartmouth-Hitchcock now uses that finding as a recommended starting point for physicians. Under the guidelines, patients taking no opioid pain pills the day before discharge go home with none. Those who take one to three pills get 15, an amount Dr. Barth’s study found satisfied 85 percent of patients, and those who took four or more get 30 pills. Dr. Barth described that guideline as “very easy to implement and remember.”

Dr. Brummett, at Michigan, says the Opioid Prescribing Engagement Network, a collaboration of hospitals, insurers, physicians and others in his state, has used similar data methods to develop procedure-specific guidelines. “We believe patient-reported outcomes are a better way to guide than expert consensus,” he said.

For his part, Dr. Makary admitted it is harder to develop guidelines like those at Hopkins and Dartmouth, but he said the effort is vital.

“It’s mind-boggling to me” that so many opioid-prescribing guidelines do not specify the procedure, Dr. Makary said. “An ingrown toenail is not the same as cardiac bypass surgery.”