George Frey / Reuters

Updated at 12:15 p.m. on August 7, 2019

In the early days of the opioid crisis, public officials had reasons to blame it on all the pills. News stories featured people who, to the shock of their neighbors and loved ones, had died unexpectedly of a drug overdose. In an emergency, authorities do what they can with the tools at hand. In tightening controls on doctors who prescribed pain relievers, state and federal agencies were focusing on the aspect of the problem most subject to regulatory intervention.

To some degree, that strategy worked. According to the Centers for Disease Control and Prevention, overdose deaths declined by about 5 percent in 2018—a dip attributable almost exclusively to fewer deaths from oxycodone, hydrocodone, and other prescription opioids. (Fentanyl deaths are still climbing.) Now that the fever of the opioid crisis may be breaking, Americans can revisit some of the stories we have told ourselves about the role of prescription medication in the crisis.

Did policy makers and public-health experts correctly assess who was at risk of becoming addicted to opioid medications? Were their views on the addictive potential of such drugs realistic? Did they anticipate the consequences of policies devised to constrain doctors from overprescribing? In retrospect, policy makers seriously misjudged the answers to these questions, overestimating the risk that these drugs posed to the average patient while simultaneously doing too little to urge clinicians to identify those most vulnerable to addiction. The best time to correct course is now—while the opioid problem still commands public attention, and before the restrictions imposed at the height of the crisis harden into permanent practice.

By now, the outlines of the story are familiar: Opioid prescribing began to rise in the early 1990s, powered by two forces. One was a campaign by oncologists and pain specialists to correct the undertreatment of pain. The other was the introduction in 1996 of the potent time-release oxycodone medication Oxycontin, which the drug company Purdue Pharma vigorously marketed to doctors.

Packed with up to 80 milligrams of oxycodone a pill, Oxycontin became a magnet for opioid abusers. And with a street value of $1 a milligram, Oxycontin became valuable as local currency. By 1999, the diversion of pills into the black market was well under way in Maine and rural Appalachia. Medication was stolen from medicine chests, dealt on the street, and sold in cash-only pill mills.

Starting around 2010 or 2011, events converged in ways that made prescription pills less widely available. Law enforcement cracked down on pill mills, the maker of Oxycontin made the pill harder to crush, physicians tightened their prescribing practices, and more states created prescription registries to help identify people who were obtaining prescriptions by “doctor shopping”—that is, by seeking prescriptions from multiple physicians at the same time. Many people who abused pills turned to heroin, which was cheaper and easier to get. Several years later, illicitly sourced fentanyl, 50 times as potent as heroin, intensified the death toll.

From 1999 to 2011, the CDC documented a fourfold rise in prescription-related overdose fatalities. In 2011, the agency officially declared “prescription painkiller overdoses at epidemic levels.” On charts, the data were straightforward—casualties mounted in parallel with the rise in opioid prescribing—but the dynamics behind the trend were not. What became the popular but essentially dubious interpretation of the trends was that the typical opioid victim was a patient who received Oxycontin, Percocet, or Vicodin from her doctor for a tooth extraction or sprained ankle and then stumbled into addiction.

In a HuffPost essay two years ago, Brooke Feldman, a 38-year-old social worker in Philadelphia who at one time struggled mightily with an addiction to pills, summarized the staple narrative like this: “Johnny is from a white middle class community. He had everything going for him—was a great kid, star athlete, nobody ever thought of him as somebody who would use drugs. Johnny was prescribed narcotic pain medication by an irresponsible doctor peddling a nefarious pharmaceutical company’s wares. Unknowingly, Johnny got hooked and eventually moved on to using heroin. If it weren’t for that evil prescription he received, none of this would have happened.”

The real story, Feldman continued, was almost always more complicated: “For myself and most people I have met who are living with or in recovery, substance use or misuse was already taking place long before opioids entered the scene.”

In fact, only 22 to 35 percent of “misusers” of pain medication report receiving drugs from their doctor, according to the Substance Abuse and Mental Health Services Administration. (Misuse is a term that includes anything from taking an extra pill beyond the quantity prescribed by a doctor to full-blown addiction.) About half obtain pain relievers from a friend or relative, while others either steal or buy pills from someone they know, buy from a dealer, or go out looking for a doctor willing to write prescriptions.

People who abuse pills are rarely new to drugs. The federal government’s 2014 National Survey on Drug Use and Health, for example, revealed that more than three-fourths of misusers had used non-prescribed benzodiazepines, such as Valium or Xanax, or inhalants. A study of Oxycontin users in treatment found that they “were not naive individuals with accidental addictions who were introduced to painkillers by their physicians as reported by the media … [Instead they had] extensive drug use histories.”

Among people who are prescribed opioids, addiction is relatively uncommon. The percentage of patients who become addicted after taking opioids for chronic pain is measured in the single digits; studies show an incidence from less than 1 percent to 8 percent. Most of the estimates are skewed toward the low end of this range, when those at risk (due to a history of substance abuse or, to a lesser but meaningful extent, a concurrent mental illness) are removed from the sample. In Feldman’s case, the nature of the risk was constant anguish. When she was 4 years old, her heroin-addicted mother left the family and died of an overdose before she was 12. “For so much of my childhood, I felt abandoned, worthless, unlovable, and confused,” she told me. Her first Percocet came from a girlfriend. “Being numb helped,” she said. Before Percocet, though, she had achieved “escape” with marijuana, alcohol, PCP, benzodiazepines, and cocaine.

As for “Johnny,” the hypothetical shiny all-American kid who seemed to have it all—well, I met a real-life version of him this year. I am a psychiatrist who, for the past 11 months, has taken a break from urban life to do some clinical work at a behavioral-health clinic in a small community in southeastern Ohio. I have gotten into the habit, as I go around town, of chatting with anyone who seems friendly. While stopping at a fast-food place one day, I met a 23-year-old man who was on a vape break from his job there, and he was game to talk. He told me that he had been a star football player at his public high school. At 18, he had been thrilled to receive a full football scholarship to Ohio State University. He planned to major in engineering but truly aspired to the National Football League.

Unlike many of his friends, who felt stymied in the small Appalachian town where pills and heroin were short-term cures for boredom, this young man could imagine a future for himself. But a month before graduation, he was in a car accident. The injury to his right shoulder ruined his ability to throw overhead—and the damage went far beyond his anatomy. He developed post-traumatic stress disorder from the crash. Even worse, he told me, his universe was shattered. Football had given him both social status and a sense of purpose. What made this devastation bearable was the hydrocodone that his orthopedic surgeon gave him. Within six months, he was drinking and still using the pills that the sympathetic surgeon continued to prescribe. Eventually, he moved to heroin. By the time we met, he was enrolled in a buprenorphine clinic and was otherwise drug-free, but was still trying to regain his footing. Every day, he went to work at a fast-food restaurant and felt bitter about the direction of his life.

All of us have stressors in our lives, but when the turmoil seems insurmountable, some people find drugs unexpectedly seductive. One day last summer, after I gave a talk on addiction in New York, a 25-year-old graduate student approached me. He told me what had happened to him as a teen and the only child of working-class parents in Queens. (Both he and the young man in Ohio gave me permission to describe their stories but asked me not to identify them by name.) During his junior year of high school, the graduate student had been laboring under what he called “a boulder of stress.” He was ashamed of being overweight, anxious about excelling at the private school his strict and intimidating parents could barely afford, and quietly panicking about what they would say when one day he would have to them tell he was gay. At the end of his junior year, he broke his front teeth in a bike accident. The dental surgeon gave him two weeks of Percocet—a drug the then-teenager had never even heard of. “But I loved it,” he told me. The drug gave him freedom from dread. He managed to get two more refills from the dentist, and then bought Percocets or Vicodins from kids who loitered outside the high school near his house. One day, his father could not rouse him and called an ambulance. His parents were shocked by this crisis, and so was he. He weaned himself off the pills and they all entered family therapy. Within a year, he felt less overwhelmed and was ready to apply to college. He could talk to his parents more freely about himself and was working up to coming out to them.

Reducing both of these cases to one-dimensional, good-kids-get-hooked-on-opioids narratives may be tempting. But these two men’s stories are more nuanced than that, and reveal much about the complex nature of addiction. Opioids acquire their dark power when they keep souls—not just broken shoulders and teeth—from throbbing. If those two young men in Ohio and New York had never been given pain relievers by their doctors, would they have gotten past their turmoil without turning to drugs?

I’m less sanguine about the former Ohio football phenom, not least because he lived in a community where so many people were trying to medicate their troubles away, whether with opioids or alcohol. In the eyes of some people, certain problems are worse than a drug habit. Addiction often conceals depression, and, if not for drugs, the Ohio man might have slid even deeper into it—at least until the toll of addiction became too great. As for the young man from Queens, seeking out Percocet would never have occurred to him if he had not been prescribed it. So yes, in a sense, his dentist precipitated his addiction. His parents initially understood his addiction to be entirely the result of a dentist and a prescription. Yet when they learned the depth of their son’s unhappiness, they realized that the prescription was not the full problem.

The focus on prescriptions as the dominant problem, however, was embraced by policy makers of all kinds, from legislators to pharmacy-benefits managers. It drove their vigorous efforts to rein in prescribing, and it resulted in a one-third reduction in the number of opioid prescriptions nationwide from 2012 to 2017. Some form of pill control was warranted, to be sure. Too many doctors and dentists were routinely overprescribing, sometimes dispensing a month’s supply of pills when only several days’ worth, if any, was needed. But the bluntly designed pill-control policies enacted by insurers, pharmacies, and regulators did great damage to patients with chronic pain who had been functioning well on prescription opioids.

Doctors felt Drug Enforcement Administration agents, their state medical boards, attorneys general, and other health-care agencies breathing down their neck. They reduced patients’ dosages or cut them off altogether, leaving them in misery, unable to find another physician who would treat them, and sometimes contemplating suicide. Policy makers have frequently used a reduction in total opioid prescriptions as a metric for success, but that metric does not account for how demand for those medications is distributed among patients. The potency of opioids is often measured in “morphine milligram equivalents,” or MMEs; 60 milligrams of oxycodone equals 90 MMEs*. It turns out that only a small minority of chronically ill people, about 10 percent (many of whom take high doses), account for 70 percent of the total MMEs prescribed. When an insurer boasts of a 25 percent reduction in total MMEs prescribed, that could mean that some chronically ill patients have successfully been shifted to a nonaddictive form of analgesia—or that patients who badly need opiates aren’t getting them.

Now that the CDC’s preliminary data on prescription-related overdoses show an encouraging trend, I hope some of the pressure on doctors to curtail prescribing as a general practice will give way to making more careful distinctions based on what is appropriate for each patient and what is not. But breaking free of the simple narrative that opioids are indiscriminately dangerous to patients is difficult. Recent headlines about massive quantities of pills hauled into small towns reinforce that perception.

There is little question that the more drugs coursing through a neighborhood or town, the more they will be used. Whether drug companies were responsible purveyors and what the DEA and distributors knew and when they knew it are pressing supply-side questions that need to be resolved. A high-profile trial pitting the state of Oklahoma against an opioid maker whose product allegedly addicted swaths of its population just drew to a close in mid-July. A decision in that case is expected soon, and courts elsewhere will litigate these issues for some time.

But the deeper story of the demand side of the now-receding prescription-pill crisis must get attention, too. This is where Feldman and those two troubled young men come in. Most people take prescribed opioids uneventfully, but in these three cases the individuals did not. Clearly, the addictive potential of drugs is not random. And while exposure is necessary for addiction to develop, exposure is almost never sufficient. Addiction is a dynamic process, and depending on whether someone is suffering in a certain way under certain circumstances, a drug will either be profoundly seductive or it won’t.

*Earlier versions of this article incorrectly described the potency of oxycodone.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.