Chapter Content
Okay, here we go. Chapter...uh...nine, I think it is? Overstories, superspreaders, and group proportions. "OxyContin is our ticket to the moon." Jeez.
So, let's talk about the opium poppy, right? Beautiful flower, long stalk, and after it blooms, the petals fall off, and you've got this pod, like, the size of a small egg, filled with this thick, yellowish sap. And for, like, thousands of years, humans have been obsessed with this stuff. It's a chemical cornucopia, apparently, according to one historian. Sugars, proteins, all sorts of crazy things in there.
Dry that sap, smoke it, you get opium. Kingdoms have, you know, risen and fallen over opium. But if you extract the alkaloids...that's where the real magic happens. Back in the early 1800s, this German pharmacist, Friedrich SertĂĽrner, he isolated the first of those alkaloids. He called it morphium...or morphine, after Morpheus, the Greek god of dreams. Morphine, you know, dulled pain and gave you this really nice, euphoric feeling. But it was also, like, super addictive.
Then came codeine, a little later. And then, like, forty years after that, this English chemist, C. R. Alder Wright, he boiled morphine with some other stuff, trying to find an opiate that wasn't addictive. He ended up with heroin. Yeah, heroin. And for a while, people thought it was, like, the safe alternative to morphine. Crazy, right?
And then in 1916, two German chemists, they took thebaine, which is kind of like codeine, messed with it, and came up with oxycodone. Now, oxycodone didn't become as infamous as, you know, heroin and morphine...not until, like, eighty years later. When this company, Purdue Pharma, decided to, like, reimagine it. They packaged it into this high-dose, extended-release tablet and marketed it like crazy. They called it OxyContin. I bet you've heard of it. It's, uh, become the most infamous prescription drug, like, ever.
So, this whole thing started with the testimony of these two executives from a company, and if you haven't guessed it, they were Sacklers, the family that founded Purdue and, you know, gave us OxyContin. And Kathe Sackler, the daughter of one of the founders, she was asked about her family's role in the opioid crisis. She said, "I've tried to figure out if there's anything I could have done differently, knowing what I knew then, not what I know now. And I can't...there's nothing I can find that I would have done differently."
Then there was David Sackler, the grandson. After Kathe denied any responsibility, he said, "I take a deep moral responsibility for it, because I believe our product, despite our best intentions and best efforts, has been associated with abuse and addiction." "Has been associated." You notice he used the passive voice there?
And I've been arguing that this kind of dissociation and denial, it's really common. We tend to think that epidemics are just mysterious things. That we're powerless, and we don't have any responsibility. But in each case, that's not true, right?
So, let's go back to the opioid crisis. Let's use the lessons we've learned from other situations, superspreaders, group proportions, overstories...Let's try to make sense of the chaos caused by OxyContin. Can we, like, really understand the decisions and circumstances that led to this? I think we can.
Okay, so there's this article in an academic journal by a demographer, Jessica Y. Ho. It's called "The Contemporary American Drug Overdose Epidemic in International Perspective." And there's this chart in the middle that shows drug overdose deaths in high-income countries. From, like, 1994 to 2015. When the most severe overdose crisis started.
And you don't need to spend too long looking at this thing to figure out what it's saying. Industrialized countries had, like, completely different experiences with opioids. Some countries, things got better. Others, it got worse, but nothing too extreme. Some countries barely even registered a problem at all. But there's one country that had a truly catastrophic experience with opioid overdoses...the United States.
Jessica Ho's chart tells us that this isn't, you know, a European problem or a North American problem. It's fundamentally an American problem. It's, like, large-area variation.
But is it? Maybe it's actually small-area variation, right? There was this analysis done by a group led by Lyna Z. Schieber, and they looked at opioid prescribing practices state by state. And there are a ton of charts and tables. But the most interesting one shows the amount of opioid painkillers prescribed in each state, from 2006 to 2017.
There's a lot of variation. Like, Alabama had almost twice the number that California did. Delaware was super high. Colorado wasn't. So it looks like the small-area variation thing again, right?
Illinois and Indiana are neighbors. Similar poverty rates, similar unemployment, similar income. So, why did Indiana have twice the problem that Illinois did?
People often say that the opioid crisis is a result of social and economic problems, lost jobs, broken families, soaring rates of depression, mental illness. All that's important. But it doesn't explain Ho's chart. Italy is poorer than the US, more unemployment. Where's their crisis? The UK has its share of problems. Why is their line so much lower? And those theories don't explain why Indiana was devastated while Illinois wasn't.
So we need to look for the overstory. Miami had its own overstory. Our understanding of the Holocaust changed because of a television series. So, is there an overstory that can help us make sense of the weird variation in opioid use? Turns out there is. It involves a man who's mostly forgotten by history. His name was Paul E. Madden.
Paul E. Madden was a lawyer from San Francisco. He was named director of the California Bureau of Narcotic Enforcement. He was, like, full of righteous energy. He was, you know, imperious, ponderous, puritanical. Climbed the political ladder through ambition and moral conviction.
He was obsessed with the perils of illicit narcotics. He thought people under the influence of marijuana might think they were either so small they were afraid to step off the curb, or of enormous size and superhuman strength and passion and in that condition commit crimes altogether foreign to his nature.
Madden liked to use hyperbole. Things were never just bad, they were evil. Illegal drugs didn't just compromise the user, they destroyed him. He was, like, a zealot, one of the first of the American anti-drug crusaders.
But his true obsession wasn't illegal drugs from overseas. It was the painkillers being prescribed by doctors. He was concerned that legal drugs were being diverted. That unscrupulous doctors were handing out opioids too easily. So, Madden came up with a solution. He made a list of drugs, morphine, opium, codeine, and he persuaded the California legislature to pass an amendment to the state's Health and Safety Code. It was called Assembly Bill No. 2606. The key part was Section 11166.06. Every time a doctor wrote a prescription for one of those opioids, they had to use a special prescription pad.
The key word was triplicate. Every prescription had two carbon copies. One copy the doctor kept, one went to the pharmacy, and the final copy had to be mailed to Madden's Bureau of Narcotic Enforcement.
Shortly after, Madden had his first big case. This doctor in San Francisco, Nathan Housman. Housman was shady. His name had already come up in another case involving a wealthy widow. But the case that caught Madden's attention was centered on Alma Elizabeth Black, who the newspapers described as "a patient [Housman] treated for seventeen years for an ailment which an autopsy—performed at his own demand—failed to detect.” Housman's "treatment" was morphine. And upon her death, Black left her entire estate, worth, like, millions of dollars, to Nathan Housman.
Madden's agents went to Housman's pharmacy. They found him there, copying his prescriptions. They found 345 prescriptions from Housman for 200 patients. Only four had been reported to Madden's office. "That is an intolerable situation," Madden said. So, Housman was arrested. Not for murder or malpractice, for failure to file Mrs. Black's morphine prescriptions in triplicate.
Housman ended up in prison, and it sent a message to every doctor in California. Paul Madden was serious. He didn't think all doctors were bad. But he wanted to use Housman to send a message to those few dangerous ones. You could not escape Paul Madden's watchful eye. He had a carbon copy of every opioid prescription in California.
In this book, we've looked at lots of ways overstories emerge. Paul Madden's filing cabinets don't seem to be in the same category. But the more Madden talked about his idea, the more it became something bigger. Writing a prescription had been a private thing. Now it was public, with consequences. "The great amount of good to be derived from this system is that the State Narcotic Enforcement Division will, every thirty days, have a complete report of the narcotics dispensed" in the state. Those two carbon copies made doctors pause and think.
Hawaii passed a version of Madden's rule. Illinois followed suit. What started as one man's crusade turned into a national phenomenon. States were telling doctors that when it comes to these drugs, you cannot be left to your own devices. A policy turned into an overstory.
Fifty years passed. And then a second overstory emerged.
Russell Portenoy grew up in Yonkers. He was the first in his family to go to college. He was, like, brilliant. He met this doctor, Ron Kanner, who did pain management. Portenoy laughed and said, "That's silly, because pain is a symptom, it's not a disease. You can't do that." But Kanner convinced him that you could treat the symptom itself.
Portenoy became convinced that medicine wasn't taking pain seriously enough. Doctors needed to prescribe opioids more. He loved opioids. He called them a "gift from nature." They "can be used for a long time, with few side effects and...addiction and abuse are not a problem." He believed that addiction was very rare, and that a thoughtful physician could tell who would thrive on opioids and who wouldn't.
This was Portenoy's overstory. He argued that Madden was too concerned about a few wayward physicians. And as a result, they made it impossible for doctors to treat pain. "What we're trying to say is that physicians have to feel completely empowered and comfortable that they can use these drugs for legitimate medical purposes." Madden worried about the dangerous few. Portenoy focused on the virtuous many.
Portenoy became a superstar. He was always in the news, giving speeches. He was called the King of Pain. Meanwhile, the Maddenites looked on in horror. The debate raged at meetings.
In 1991, NIDA held a meeting to discuss whether triplicate prescriptions should become a national thing. Russell Portenoy was there, of course. He spoke at length. He was worried about the risk of underprescribing. In the end, the plan went nowhere. Portenoy's ideas gained followers. By the mid-90s, the number of triplicate states was down to five, barely a third of the population. Everyone else went with Portenoy.
And that was it. Another policy difference. But there was this drug company in Connecticut named Purdue Pharma that was paying attention.
Purdue had been in the painkiller business for years, with MS Contin. It was a good business, but small. The Sackler family had grander ambitions. They switched to oxycodone. Oxycodone was usually combined with acetaminophen or aspirin. But Purdue's first innovation was to remove the acetaminophen switch from oxycodone.
Then Purdue raised the drug's dosage and created a special extended-release tablet. They called this new painkiller OxyContin, then set out to market it to everyone. Your back hurts? OxyContin. Wisdom teeth? OxyContin.
At Purdue headquarters, they were excited. "OxyContin," one of the Sackler brothers said, "is our ticket to the moon."
In 1995, Purdue hired a market-research firm. They wanted to plan their marketing strategy. They set up sessions with doctors. Purdue wanted to know what doctors thought of a high-dose, extended-release opioid.
The good news was that non-cancer pain was a big part of the physicians' practices. The doctors wanted more treatment options. But the bad news was that the sessions with doctors from Houston were a disaster. Why? Texas was a triplicate state.
The triplicate laws seemed to dramatically affect the product usage behavior of the physicians. The doctors in Texas used Class II narcotics "less than five times a year...if at all." They didn't want to provide the Government with any ammunition to question their medical protocols relative to pain management. The thought of the government questioning their judgment created a high level of anxiety. The focus-group report said that the triplicate and non-triplicate states were night and day.
Purdue's management team took the report seriously. The launch of OxyContin was targeted at states without triplicate laws. No big push in New York State. Yes to West Virginia. And so on. The opioid epidemic did not hit the entire United States equally. It became an example of small-area variation.
The top five opioid-consumers were all "Portenoy states," without a triplicate program. The per-capita opioid consumption was much lower in the Madden states.
If you looked further into the numbers, the differences grew. Here's the breakdown of how willing orthopedic surgeons were to prescribe opioids. The South, the triplicate-free zone, the land of the Portenoy overstory, made up a huge amount of the total.
Think about how remarkable this is. A drug warrior makes California's doctors use a special painkiller-prescription pad. That evolves into an overstory that says opioids are different, and that overstory is so compelling that when Purdue tests its painkiller in a triplicate state, it runs into a brick wall.
Overstories matter. You can create them. They are powerful. And they can last for decades.
Today, a lot of economists are studying the ways that states with triplicate laws differ from others. One economist estimates that if New York had Massachusetts's opioid-overdose rate, an additional 27,000 New Yorkers would have died of overdoses. The only relevant difference is that New York forced doctors to make carbon copies of every prescription, and Massachusetts didn't. And those carbon copies saved thousands of lives.
Even now, triplicate laws don't apply to fentanyl. So you would think the differences between triplicate and non-triplicate states would have faded away. Wrong! If Purdue sales reps put you on a path, you stayed on that path long after they were gone.
Economic growth has been stronger in triplicate states. The health outcomes of babies were better there. Yongbo Sim’s conclusion, after comparing crime rates in triplicate and non-triplicate states, was that “non-triplicate states at the time of OxyContin’s introduction experienced a relative rise in both property (12%) and violent (25%) crimes compared to states with the triplicate prescription policy (triplicate states).”
Paul Madden is probably looking at us and saying, "I told you so."
So, that's overstories. Let's turn to the second thing: superspreaders.
This article from McKinsey talked about the way pharmaceutical companies sold their products to doctors. They were doing it wrong. They needed to "segment" physicians. Some doctors were worth more than others. Drug companies had to figure out the lifetime value of each doctor.
One company took notice: Purdue Pharmaceutical.
Purdue called McKinsey in 2013. Sales of OxyContin had soared, but the growth engine was stalled. The Department of Justice had just accused Purdue Pharma of misleading doctors, and OxyContin's reputation was suffering. What should they do?
McKinsey came up with a plan called "Evolve to Excellence," or E2E. When Richard Sackler heard the presentation, he said, "The discoveries of McKinsey are astonishing." Purdue would pay McKinsey millions for its advice.
At the heart of E2E was this chart. It showed that OxyContin's success was driven by a tiny fraction of doctors, roughly 2,500, who wrote a staggering number of prescriptions. McKinsey called these doctors the "Core" and "Super Core."
McKinsey's advice was blunt. Purdue needed to focus on the superspreaders. Purdue listened.
Then McKinsey said: Zoom in even further on those Core and Super Core. Figure out the ones who are most receptive to persuasion. That meant younger doctors, or doctors who liked spending time with sales reps.
This is the total number of OxyContin sales calls in Tennessee, a non-triplicate state. Between 2007 and 2016, the number of doctor visits by OxyContin sales reps goes up by nearly a factor of five.
The same was true of the superspreaders: they weren’t wired like most doctors. When the Purdue sales representatives downplayed the risks of addiction, the superspreader believed them. When it became clear that OxyContin was being abused, the superspreader was indifferent. They thought handing out drugs willy-nilly was what a doctor did.
One of Purdue’s targets in Tennessee was a doctor named Michael Rhodes. In 2007, he wrote 297 OxyContin prescriptions. Purdue started calling on him. By 2010, he wrote 1,307 prescriptions. Rhodes blossomed. He became Super Core.
The Super Core wanted love. If you visited them one to four times a year, their prescriptions declined. Even if you visited them eight times a year, twelve times a year, sixteen times a year, they still fell. Twenty-four visits a year was the tipping point. If you held their hand and wined and dined them, they would be your best friend forever.
So the sales reps kept visiting Michael Rhodes. From the Purdue sales notes, it is clear that Rhodes’s practice was in shambles. In May 2014, the sales rep and the Purdue District Manager “called on Dr. Rhodes...and continued to encourage him to write more prescriptions, despite his objection.”
All told, from 2006 to 2015, Rhodes prescribed 319,560 tablets of OxyContin. Michael Rhodes was the Nathan Housman of Tennessee.
Purdue built its strategy around people like this. In another case, a Purdue sales rep spoke to a pharmacist about one of their Super Core prescribers. The pharmacist said she was known as "the candyman" because she put every patient on the highest dose of narcotics.
Once the opioid epidemic was off and running, the top 1 percent of doctors "accounted for 49 percent of all opioid doses." Purdue fueled an epidemic that would end up consuming the lives of hundreds of thousands Americans based on the seduction of no more than a few thousand doctors.
The lesson of the opioid crisis is the same. A tiny fraction of doctors was enough to kick-start the epidemic. This is the Law of the Very, Very, Very Few.
Now let's talk about the third thing: Group Proportions.
The opioid crisis unfolded in three acts. The first was the decision by Purdue to avoid the Madden states. The second act began with McKinsey’s reinterpretation of the Law of the Few. But the third act was the most catastrophic. It was when the group proportions of the crisis changed.
In 2010, Purdue announced it would retire the old OxyContin. It would be replaced with OxyContin OP. OP looked the same, had the same ingredients. But it couldn't be crushed and snorted.
Some addicts would switch to another drug. But many would simply stop. Right? And surely the steady flow of new patients feeding the epidemic would slow down. Those using OxyContin didn't think of themselves as traditional addicts.
Those who welcomed Purdue’s reformulation assumed that people with drug problems had a reason for using whatever drug they did. The assumption was that the group proportions of the opioid crisis were relatively fixed. So if you cracked down on one class of user then the overall size of the problem would shrink.
The group proportions weren’t fixed at all.
Here's a chart showing the overdose-death rates for three classes of opioids. One column is prescription opioids. The second column is heroin. The third is synthetic opioids such as fentanyl.
These are the group proportions for the opioid crisis up until OxyContin's reformulation. More than five times as many people were dying from drugs like OxyContin than were dying from heroin and fentanyl. If you must have an opioid epidemic, these are the group proportions you want.
A prescription-drug epidemic is powered by a company operating within the law. Insurers reimburse users. We know when things go wrong. We have levers to push. But what did reformulation do? It shifted those proportions. Prescription-drug users who couldn't crush their pills simply switched to heroin and fentanyl. Take a look at the statistics for the years following reformulation.
Prescription-drug deaths go up slightly over the next decade. But the number of fatal overdoses from heroin goes up 350 percent by 2017. And the number of people killed by fentanyl goes up 22-fold.
Now addicts had become the customers of criminals. They were buying a product made in some shady factory somewhere, laced with who-knows-what. They weren’t snorting anymore. Now they were injecting, and injecting a drug is a hundred times more dangerous.
Heroin was cheaper for addicts. But it ended up being more expensive. Withdrawal from heroin was worse than going cold turkey from OxyContin. Abuse and neglect of children soared. And in time heroin gave way to fentanyl. Fentanyl was deadlier and more addictive.
The opioid problem is now so bad that the early days of the epidemic look bucolic. We would have been better off if we had said no to Purdue’s reformulation in 2010 and kept things the way they were.
Someone would have had to stand up back in 2010 and say, Look. We have two versions of a highly addictive drug. The original version is easily abused. The new and improved version is not. But we don’t want the new and improved version. We want people to continue to crush their OxyContin and snort it.
They estimate that if Purdue had stuck with its original formulation of OxyContin, by 2017, reformulation increased overdose rates by over 100 percent.
In other words, we were slowly winning the war on opioids. But we never had an honest conversation about how epidemics work. So along came OxyContin OP, and everything went sideways.
So here it is. The forensic analysis of the opioid crisis. A little company in Connecticut decided to reinvigorate a gift from humanity. The US got small-area variation. Then McKinsey came in and refocused Purdue’s marketing toward the superspreaders. The Purdue sales reps told the Core and the Super Core doctors that addiction was rare, and patients could tolerate high doses. It wasn’t true.
OxyContin got an extra decade of life. Many more patients became addicted. Sales hit $3 billion a year. Then came reformulation. The people addicted to OxyContin switched to heroin. Then they switched from heroin to fentanyl.
By the early 2020s, the opioid epidemic that had begun back in 1996 with the introduction of OxyContin was claiming the lives of almost 70,000 Americans a year.
Two decades into a pandemic, that line should be going down, not up.
“I have tried to figure out, was—is there anything that I could have done differently, knowing what I knew then, not what I know now,” Kathe Sackler said. “And I have to say, I can’t.”
That is very hard to accept. But so is the story that we tell ourselves that we bear no responsibility for the epidemics that surround us.
Epidemics have rules. They have boundaries. They are subject to overstories, and we are the ones who create overstories. They change in size and shape when they reach a tipping point, and it is possible to know when and where those tipping points are. They are driven by a number of people, and those people can be identified. The tools necessary to control an epidemic are sitting on the table, right in front of us. We can let the unscrupulous take them. Or we can pick them up ourselves, and use them to build a better world.
That’s it for this chapter. Phew.