The Food and Drug Administration will require drug companies to study whether prescription opioids already on the market are effective in quelling chronic pain – another step in the government’s efforts to rein in use of the narcotics that spawned the drug epidemic.

Some studies already indicate that opioids are ineffective for pain beyond 12 weeks and many experts say long-term use can cause addiction, by prompting patients to build up tolerance to the drugs and seek higher doses. But conclusive, controlled research is scarce.

A finding of ineffectiveness in more rigorous studies supervised by the FDA could allow the agency to change the labeling on some opioids, impose special rules for prescribing, dispensing and taking them and even prohibit their use in some cases, according to FDA Commissioner Scott Gottlieb.

“We are going to impose a mandate on existing products . . . to answer the question that people have been posing for years: whether you have declining efficacy, and whether that declining efficacy can lead to addiction,” Gottlieb said in an interview.

But at least one longtime critic of the FDA’s response to the opioid crisis expressed frustration with the move. Andrew Kolodny, director of Physicians for Responsible Opioid Prescribing, said the FDA already has all the research it needs – and authority under existing law — to tighten restrictions on the use of opioids for chronic pain by changing instructions for how they should be prescribed.

Here we go again,” Kolodny said in an interview. “That’s exactly what the FDA said to us in 2013. . . . Five years later, we don’t have the studies and another FDA commissioner says ‘we’re going to do the studies.’ “

Gottlieb said the new research is aimed at all immediate, long-acting and extended-release opioid tablets taken by patients outside of health-care settings. It would cover products already on the market and new applications, Gottlieb said. It would not include painkillers used short-term inhospitals, for instance. Opioid painkillers were originally intended for the severe pain of cancer and end-of-life care, but beginning in the late 1990s, drug companies helped persuade doctors to offer them for a wide variety of less serious conditions.

Each manufacturer would be required to sponsor unbiased, controlled research of its opioid products, likely at universities, he said.

The FDA also is mandating a second study, Gottlieb said, to determine whether opioids can actually cause users to become more sensitive to pain, a condition known as “hyperalgesia.” Preparations for that research by a consortium of companies are farther along and it should start soon, he said.

In 2013, after Kolodny’s group complained that opioids should be labeled unsafe and ineffective for chronic pain, the FDA ordered similar research, including an attempt to determine whether painkillers cause hyperalgesia. Gottlieb said those studies were difficult to carry out because, at the time, the FDA had authority only to require post-market studies of safety, rather than effectiveness.

On Sunday, the CBS program “60 Minutes” explored the FDA’s decision in 2001 to allow long-term use of OxyContin despite the lack of research showing it was safe and effective. Gottlieb conceded that “it’s regrettable we didn’t do this many years ago.”

The vast majority of opioid prescriptions written in 2017 were for generic versions of the drugs. The research would be required only of companies that produce brand-name narcotics; generic producers would be required to adopt the same changes.

The FDA acquired the authority to demand the effectiveness research in a section of the Support for Patients and Communities Act. passed in October. Previously, Gottlieb said, the FDA could order only safety studies of drugs on the market, but the new power allows for effectiveness studies as well.

He said drug companies will soon be informed of the new requirement. A spokesman for Purdue Pharma, the company that pleaded guilty in 2007 to deceptive marketing of opioids, declined to comment. A spokeswoman for the industry lobbying group, the Pharmaceutical Research and Manufacturers of America, did not return an email seeking comment.

Gottlieb has previously said the FDA must take a more active role in battling the opioid crisis, which killed 47,600 people via overdoses in 2017. In November, when the FDA’s approval of a powerful new opioid, Dsuvia drew criticism, he said he would seek authority to consider an opioid’s benefits and harm to public health, its risk of being diverted to abusers and whether it has unique benefits before deciding on future applications.

After peaking in 2012 at more than 255 million, the number of opioid prescriptions written annually has declined sharply to 191.2 million in 2017, according to the Centers for Disease Control and Prevention, and overdose deaths from prescription narcotics leveled off as well. Still opioid overdose deaths are six times higher than they were in 1999, with more users succumbing to heroin and, in recent years, illicit fentanyl.