Sickening

How Big Pharma Broke American Health Care and How We Can Repair It

John Abramson

• Publisher: Mariner Books
• Year: 2022
• ISBN:
  9781328956989
• Topic: Medicine
• Subtopic: Health Care Delivery

The inside story of how Big Pharma's relentless pursuit of ever-higher profits corrupts medical knowledge—misleading doctors, misdirecting American health care, and harming our health.

The United States spends an excess $1.5 trillion annually on health care compared to other wealthy countries—yet the amount of time that Americans live in good health ranks a lowly 68th in the world. At the heart of the problem is Big Pharma, which funds most clinical trials and therefore controls the research agenda, withholds the real data from those trials as corporate secrets, and shapes most of the information relied upon by health care professionals.

In this no-holds-barred exposé, Dr. John Abramson—one of the foremost experts on the drug industry's deceptive tactics—combines patient stories with what he learned during many years of serving as an expert in national drug litigation to reveal the tangled web of financial interests at the heart of the dysfunction in our health-care system. For example, one of pharma's best-kept secrets is that the peer reviewers charged with ensuring the accuracy and completeness of the clinical trial reports published in medical journals do not even have access to complete data and must rely on manufacturer-influenced summaries. Likewise for the experts who write the clinical practice guidelines that define our standards of care.

The result of years of research and privileged access to the inner workings of the U.S. medical-industrial complex, Sickening shines a light on the dark underbelly of American health care—and presents a path toward genuine reform.

Part I

Health Care American-Style

1

Vioxx: An American Tragedy

Don’t believe anything, not one thing, put out by a pharmaceutical company. Just don’t believe it. You start from there.

— CATHERINE DEANGELIS, editor in chief, Journal of the American Medical Association

Late in 2004, I received a heartbreaking note from a woman who had seen me on television discussing safety issues associated with Vioxx and Celebrex, two widely used pain relievers. Mrs. Susan Palmer wrote to ask if I would be willing to review her daughter’s medical records: “It would be a blessing to us. Please help us. My daughter, Stacey, was 17 years old.”*

Several months earlier, Stacey, a healthy honors student and budding gymnastics star, fell and hit her head during practice. She blacked out briefly, and when she came to, she had pain in her right wrist. Stacey’s mother picked her up at school and brought her to the emergency room. There was no evidence of serious head trauma, but wrist X-rays revealed a fracture. A cast was put on, and she was instructed to take ibuprofen (Motrin or Advil) as needed for pain and to follow up with a pediatric orthopedist.

About a week later Stacey began having headaches. Ibuprofen provided little relief, so Stacey’s mother brought her to their family doctor. The doctor found she had some tenderness in the back of her neck and the right side of her head, but Stacey’s examination was otherwise normal. A CT scan of her head was also normal. Because ibuprofen was not working, the doctor gave Stacey eight Vioxx 25 mg tablets, samples provided by the manufacturer, to be taken once daily, and made a follow-up visit for one week later. Vioxx, a pain reliever marketed as a safer version of ibuprofen and similar drugs, had recently been approved to treat migraine headaches.

That Stacey’s family doctor had samples of Vioxx meant a Merck drug rep had “detailed” him (that is, given him information about the company’s new drug, trying to persuade him to prescribe it). The rep probably had visited the office, brought food and small gifts (like pens and notepads emblazoned with VIOXX) for the doctor and his staff, and “educated” them about the benefits of Vioxx. And the rep probably left marketing materials, possibly including a reprint of the VIGOR (short for Vioxx Gastrointestinal Outcomes Research) article that had been published in one of the most influential medical journals, the New England Journal of Medicine (NEJM), on November 23, 2000. The article concluded that Vioxx had a safety advantage over an older anti-inflammatory drug, naproxen (Aleve), because it caused fewer serious stomach complications.

But when Stacey returned for a recheck, she reported that her headaches were occurring daily and had become so severe that she was going to the school nurse’s office two or three times a week to lie down. She also reported having blacked out while walking to school. Again, Stacey’s exam was normal. The doctor suspected post-concussion syndrome.* Because the Vioxx had not helped, he advised Stacey to go back on ibuprofen at a higher dose and arranged for her to see a neurologist. Stacey never got to that appointment. Four days later, at 4:30 a.m., Mrs. Palmer found her daughter dead in her bed, the victim of a massive stroke.

This needless tragedy almost certainly would not have happened if it weren’t for the failure of three guardians of public safety: the manufacturer, which concealed the risk of Vioxx; the New England Journal of Medicine, which published the study claiming the drug was safe and failed to correct the article after its editors became aware of the undisclosed risk; and the FDA, which failed to act effectively as soon as it became aware of that risk.

DOCTORS WERE FALSELY REASSURED

As the Palmer family struggled with their loss, one of the world’s largest drug companies was dealing with a crisis of its own. On September 30, 2004, less than three weeks after Stacey died, Merck pulled its blockbuster Vioxx off the market. A just-completed Merck-sponsored study had shown 25 milligrams of Vioxx per day — the dose Stacey had taken — doubled the risk of heart attack and stroke. According to Dr. Peter Kim, the president of Merck Research Labs, the study showed an “unexpected increase in cardiovascular disease rates.” He told the New York Times, “What we saw was stunning.”

As a Merck executive, Dr. Kim had good reason to be unhappy with the latest study results — but he should not have been surprised. More than four years earlier, as shown by internal documents, Merck executives lamented similar results from the VIGOR study. On March 9, 2000, when the VIGOR data first became available to Merck scientists, Merck’s then head of research, Dr. Edward Scolnick, wrote in an internal e-mail that the cardiovascular dangers of Vioxx (meaning increased risk of heart attack, stroke, and blood clots) were “clearly there” and “a shame.” Nonetheless, his e-mail concluded reassuringly: the class of drugs of which Vioxx was a member (the COX-2 inhibitors) “will do well and so will we” (italics mine).

The VIGOR data Merck sent to the FDA in August 2000 — by law, this had to be a full accounting of the trial’s results — stand in stark contrast to the reassurance about the safety of Vioxx claimed in the NEJM article published on November 23, 2000. The article did report that people who were treated with Vioxx had four times as many heart attacks as those treated with another anti-inflammatory drug (naproxen) but then stated that the increase occurred in people who should have been taking aspirin to prevent heart attack and stroke but were not allowed to by the study design. Importantly, of the thirteen authors, eleven were researchers with financial ties to Merck, and two were Merck employees, both of whom had received Dr. Scolnick’s e-mail of March 9, 2000. The Merck-employed authors knew, or should have known, that the NEJM article they coauthored had underreported the increased risk of heart attack associated with Vioxx.

When the 2000 article presenting the results of the VIGOR trial was first published, I read it with interest. However, it was a second NEJM article, published the second week of August 2001, that created a watershed moment in my career. I was eating lunch and reading medical journals in my office when I came upon an NEJM overview of the COX-2 inhibitors Vioxx and Celebrex. The article claimed (incorrectly it turned out) that these drugs caused “significantly fewer serious gastrointestinal adverse events” than older anti-inflammatory drugs. The review did, however, include more detail about the cardiovascular events in the VIGOR trial than the initial NEJM article had, reporting that people treated with Vioxx experienced significantly more heart attacks and strokes than those treated with naproxen.* But the article then dismissed this increased danger with a peculiar argument: “The difference in major cardiovascular events in the VIGOR trial may reflect the play of chance. The end point was prespecified, and the difference in the frequency of events was statistically significant, but the absolute number of cardiovascular events was small (less than 70)” (italics mine).

The moment I read those two sentences, I knew something was dreadfully wrong with the information presented about Vioxx in the world’s most influential medical journal. First, even a beginner knows statistical significance is based on the number of events that occurred, so dismissal of statistical significance because of a small number of events made no sense and reeked of bias. Second, what a monumental double standard — dismissing the cardiovascular harm because “less than 70” events had occurred but at the same time touting the reduction in serious gastrointestinal events associated with Vioxx, though there were even fewer of those. Nobody on the Merck side of this issue was blowing off the gastrointestinal (GI) benefit of Vioxx because only 53 serious GI events had occurred in the VIGOR trial.

The inexplicably favorable treatment of Vioxx in this second article made me wonder if either of its two authors had financial ties to Merck. I wasn’t surprised when I read the fine print: Both did. But I was puzzled because at the time, the NEJM had a clear policy that authors of review articles and editorials were expected not to “have any financial interest” in the companies that made the products discussed. That the NEJM editors changed this policy in June 2002 to allow such conflicts of interest does not minimize their wanton disregard of their own rules in 2001.

The data presented in the review article showed that the authors had access to more VIGOR data than had been published in the 2000 NEJM article. I tried and failed to find the source of their information. But just two weeks later, an article in the Journal of the American Medical Association (JAMA) reported even more of the VIGOR data; it showed that the risk of suffering a cardiovascular event was 2.38 times greater with Vioxx than with naproxen. I pasted the web address from footnote 20 of the JAMA article into a web browser and was suddenly looking at the FDA’s assessment of the data Merck had submitted from the VIGOR study.

As I read through the FDA reports dated February 2001, it became patently clear that the NEJM articles’ assurances about the safety of Vioxx were based on three brazenly misleading claims. First, the authors simply omitted three heart attacks that occurred in people treated with Vioxx. They cooked the books, lowering the number just enough to avoid a statistical threshold that would have shown that, aspirin or no aspirin, Vioxx was triggering heart attacks.* Second, the original NEJM article reported heart attacks alone, rather than all cardiovascular events (heart attacks, strokes, and blood clots); if it had included all cardiovascular events, as called for in the study protocol, it would have shown a more than double (and statistically significant) increase in cardiovascular risk associated with Vioxx. And in a third, equally audacious breach of scientific integrity, both NEJM articles simply failed to report that people treated with Vioxx developed 21 percent more serious illnesses than those treated with naproxen. In other words, the overarching finding from the VIGOR study was that Vioxx was significantly more dangerous than naproxen and provided no better pain relief.

A fundamental principle drummed into me during my medical training was that a doctor’s practice should be guided by the findings of high-quality studies published in peer-reviewed medical journals, but I suddenly understood that at least some information in the NEJM could not be trusted. Now I wondered if this was also true for other journals. I returned to the FDA website, searching for the data behind an article I’d read in the Washington Post. The newspaper had reported that a JAMA article claiming to show that Celebrex caused fewer serious GI complications than older over-the-counter anti-inflammatory drugs was also based on a misrepresentation. It turned out the JAMA article had relied on only six months of data from the large Pfizer-sponsored Celecoxib Long-Term Arthritis Safety Study (CLASS). However, CLASS was a twelve-month study. At the end of twelve months, patients who took Celebrex did not have significantly fewer serious GI complications than those who took older anti-inflammatory drugs.

The two most trusted medical journals in the United States had published incorrect, manufacturer-biased reports about major drugs. The FDA knew that both articles had misrepresented the data but did not correct the misleading information.

I was stunned.

THE NEJM WAS PART OF THE PROBLEM

A Seattle pharmacist, Jennifer Hrachovec, PharmD, had found the FDA data about Vioxx and the VIGOR trial at least two months before I did. She and a physician colleague submitted a letter to the NEJM in June 2001 pointing out that the journal’s report had not included all the heart attacks that occurred among people treated with Vioxx, and thus the published results made Vioxx look less dangerous than it really was. The NEJM responded that space was limited and did not publish the letter.

Two months later, in August 2001, Dr. Hrachovec called in to a radio show on which Dr. Jeffrey M. Drazen, the editor in chief of NEJM, was a guest. According to the Wall Street Journal, Dr. Hrachovec “begged” Dr. Drazen to update the article: “My concern is that doctors are still using [Vioxx] and exposing their patients to higher risks of heart problems and they just don’t even know that that’s the case.” (Remember, this was three years before Stacey Palmer died.) Dr. Drazen responded dismissively, “We can’t be in the business of policing every bit of data we put out.” Indeed, as we will see, the NEJM was not in that business.

On September 17, 2001, one month after Dr. Hrachovec’s call to the radio show, the FDA sent an official warning letter to Merck, stating the company’s marketing of Vioxx minimized “the potentially serious cardiovascular findings” observed in the VIGOR trial and was “false, lacking in fair balance, or otherwise misleading.” In uncharacteristically forthright language, the FDA letter called a Merck press release that claimed Vioxx had a favorable cardiovascular safety profile “simply incomprehensible given the rate of [heart attacks] and serious cardiovascular events compared to naproxen.” Finally, based on “the fact that your violative promotion of Vioxx has continued despite our prior written notification,” the FDA demanded Merck “immediately [cease] all violative promotional activities and dissemination of violative promotional materials for Vioxx.” Strong words from the FDA. The letter received some coverage in the press and was available on the FDA website (if you knew where to look). But still, this critically important information did not come from a source that doctors considered more reliable than their medical journals, and an FDA official later testified to the Senate that the FDA did not do enough to make sure doctors understood the cardiovascular risk of Vioxx.

Even though the FDA officials grasped the seriousness of this problem, they left a loophole big enough to drive tens of thousands of funeral processions through: They did not stop Merck from continuing to purchase hundreds of thousands of reprints of the incomplete and falsely reassuring NEJM article and hand them out to doctors. These reprints repeated the very same marketing claims that the FDA warning letter had expressly forbidden Merck to make, but the company let the NEJM get paid to do the talking. And the NEJM was apparently happy to do so, selling more than 929,000 reprints of its misleading article (mostly to Merck) — far more than enough to give one to every prescription-writing doctor in the United States — and bringing between $697,000 and $836,000 into the journal’s coffers.

When Vioxx was pulled...