Denialism: How Irrational Thinking Hinders Scientific Progress, Harms the Planet, and Threatens Our Lives - Michael Specter (2009)

Chapter 1. Vioxx and the Fear of Science

The daily work of science can be repetitive and dreary. Even the most talented researchers spend the bulk of their lives bathed in the fluorescent light of the laboratory, hovering over a bench, staring at slides, and hunting for meaningful patterns in strings of numbers. Still, like many of his colleagues, the cardiologist Eric Topol had always dreamed that one day he might have his “eureka” moment—a flash of insight that would permit him to see clearly what others couldn’t see at all. In 2001, Topol got his wish—but not in a way he had ever imagined; there were no shouts of joy, no elation or champagne, nothing of the kind. “I was just sad,” he said, remembering the moment when he realized that one of the nation’s most popular new medications was killing people. “Then I was angry, and eventually I became outraged.”

At the time, Topol was chairman of the Cleveland Clinic’s cardiology department, which he more than any other physician had transformed into one of the finest in American medicine. His research into how to prevent and treat heart attacks was highly valued and constantly cited. As perhaps the clinic’s most visible face, Topol was already prominent. But it was his role in helping to expose the grave risks posed by the anti-inflammatory drug Vioxx that turned him into one of the country’s best-known doctors. It also made him one of the most controversial, in part because he repeatedly stressed how little regard the Food and Drug Administration seemed to have for the hundreds of millions of people it had been created to protect.

The cloak of invincibility had long before been stripped from any government agency, replaced by the constants of doubt and denial; politicians, scientists, doctors, and lawyers are held in lower esteem today than at any time in decades. Yet no previous incident—not the explosion of the space shuttle Challenger, Ford’s willingness to dump a death trap called the Pinto on the American public, not even the nuclear accident at Three Mile Island—demonstrates more vividly why that mistrust has become so pervasive.

Vioxx was introduced by Merck with great enthusiasm in 1999, one of a new class of drugs called cox-2 inhibitors, which were designed to interfere with an enzyme called cyclooxygenase-2, which, among more beneficial duties, produces chemicals that cause inflammation (and pain). Before Vioxx appeared, hundreds of thousands of people who suffered the debilitating effects of arthritis and other chronic ailments faced an unpleasant choice every day: they could take drugs like aspirin or Advil, or they could endure agony in order to avoid the bleeding ulcers and other serious stomach complications those drugs can cause. Vioxx was referred to as “super aspirin,” which didn’t seem like much of an exaggeration: in early studies it offered better pain relief than any traditional remedy, and was far less likely to disturb the stomach. The drug quickly came to be seen by those who needed it most as a kind of magic potion, one that only the tools of modern medicine could have produced. Driven in part by one of the most aggressive advertising campaigns in medical history, more than twenty million Americans took Vioxx at one time or another. In 2003 alone, Merck sold more than $2.5 billion worth of the drug.

Topol, who suffered from an arthritic knee, loved Vioxx. Even now he readily attests to its effectiveness. “Nothing worked as well for me before or since,” he said. “Vioxx truly dulled the pain.” One February morning in 2001, though, he noticed a report that struck him as odd. Topol had been invited to deliver a lecture about the future of cardiac care to a gathering at the Medical College of Georgia in Augusta. Over breakfast at his hotel, he started to page through the copy of USA Today that he found on his doorstep. One particular story leapt out at him. “It was about Vioxx,” he said, “and this study,” called VIGOR—Vioxx Gastrointestinal Outcomes Research—“which was intended to determine whether Vioxx really was easier on the stomach than other, less powerful nonsteroidal anti-inflammatory medication.”

Between January and July 1999, researchers had followed eight thousand patients with rheumatoid arthritis. Half took Vioxx to control their pain; the other half took naproxen, which is sold over the counter as Aleve. (It was a large and fairly conventional randomly assigned, double-blind study—which meant that the patients had no idea which of the two drugs they were taking, and neither did their doctors.) The first time the safety committee assigned to monitor the study looked at the data, it found exactly what one might have expected: people in the Vioxx group were less likely to experience significant stomach distress than those who took Aleve.

The trial also showed something that had not been anticipated, and the news there was disturbing: participants who already suffered from heart disease were far more likely to have heart attacks if taking Vioxx than if they had been given Aleve. Nobody was sure why, and because Merck had never expressed concern about the drug’s effect on the heart, there were no cardiologists on the safety committee (which was not unusual since that wasn’t the purpose of the trial). Scientists wondered if the difference might have been due to the fact that people in the trial were required to stop taking aspirin, since it can lower the risk of heart attack or stroke. It was also possible that something previously unrecognized about the chemical composition of Aleve itself helped protect the cardiovascular system. (That would have provided a benign explanation for the differing rate of heart attacks, and Merck endorsed the hypothesis with great enthusiasm.)

“I was not a drug safety expert and I never even had any interest in the issue,” Topol said. “My principal research was in heart disease and heart attacks, and that dates back more than twenty years.” Topol had made a name for himself as a postdoctoral researcher at the University of California at San Francisco. He then took a job at Johns Hopkins University, where he became the first physician to treat heart attacks with the powerful clot-busting agent known as tPA; he also directed a pivotal study that compared the efficacy of that drug with an older treatment, streptokinase, in saving lives. In 1991, Topol moved to the Cleveland Clinic, where for the next fifteen years he served as chairman of the department of cardiovascular medicine.

What Topol saw in USA Today that morning in Augusta made no sense to him. “Why would a new anti-inflammatory agent prove less protective against heart attacks than the one you can buy at a pharmacy with no prescription?” he wondered. Yet, the newspaper report suggested that patients taking Vioxx were more than twice as likely to have heart attacks as those taking Aleve. For people with a history of heart disease, the risk was far higher. (Risk numbers don’t mean much unless they are accompanied by some assessment of the statistical probability that those risks could occur by chance. In this study, which at the time was the largest ever conducted using Vioxx, that number was .002—two in one thousand. In other words, if the study were repeated a thousand times, results like those would appear by chance twice.) “I thought, that is interesting, they are saying a highly touted experimental drug was not as good as the one you buy in the drugstore,” Topol said. “They weren’t saying anything about Vioxx causing heart attacks, just that Aleve seemed better at preventing them.”

The distinction was crucial because after the data was made public, Merck asserted, as it would for the following three years, that Vioxx posed no increased risk of heart attack or stroke. “It did seem strange,” Topol said, “but I didn’t give it a lot of thought. After all, that’s why you do clinical studies—and so it looked like maybe Aleve had some protective effect we didn’t know about. Surprising as that would have been, it certainly wasn’t beyond the realm of possibility. Still, it wasn’t my thing and I didn’t dwell on it.”

Topol delivered his address and returned to Cleveland, where Debabrata Mukherjee—“one of my fellows and a brilliant scientist”—had also seen the report which showed that people using Vioxx were far more likely to suffer from heart attacks than those taking over-the-counter pain medication. Mukherjee became intensely curious about the reasons for such surprising results. He dove into the data that Merck had been required to provide the FDA, and soon realized that the initial report failed to include all the essential information at Merck’s disposal. (Nor could he find any scientific support for the company’s suggestion that the results reflected Aleve’s previously unrecognized protective powers rather than the dangers of Vioxx.)

“Deb had gone onto the FDA Web site to look at all the data presented in the advisory committee meetings—something, by the way, that I had never done,” Topol said, shaking his head with a wan smile of admiration for the diligence of his young colleague. Vioxx was not the only cox-2 inhibitor on the market; Celebrex, made by Pfizer, was introduced that same year, and Bextra had also recently been approved by the FDA. But each drug functioned in a slightly different way and the numbers for Vioxx were more disturbing than those for either of the others. Mukherjee told Topol that there was a “real problem with Vioxx in particular,” he recalled. “I had many other things on my mind so I said, ‘Nahhh, it’s not that big a deal. We have different work to do. Let’s not waste our time on this.’ Deb wouldn’t have it. He insisted that I look at the data with him, absolutely insisted. So I did.”

Once Topol had the statistics in front of him, he saw why Mukherjee had become so agitated. “The evidence was right there,” he said. “I still cannot believe that nobody else had seen it. That’s when I began to understand what was really going on in that USA Today story. It just clicked: the company was attributing these miraculous functions to Aleve instead of investigating the potential dangers of their new drug. They were playing games: what they said at the time didn’t seem to be an outright lie, but it also wasn’t the truth that people needed to know. I said let’s write this up. After all, these data mattered. It wasn’t even a heart study, it was supposed to assess stomach complications, but you can’t just shy away from information like that. There were too many lives at stake.”

Topol and Mukherjee quickly put a paper together, along with Steven Nissen, another prominent cardiologist at the Cleveland Clinic, who had attended the advisory meeting where Vioxx was approved. “Deb drove the research and I gave it a framework,” Topol said. The paper was the first independent analysis to include all the data the FDA had obtained from the VIGOR study, and it cast serious doubt on the supposition that naproxen offered special cardiovascular protection. The study was published in the Journal of the American Medical Association later that year. The three stopped short of calling for a moratorium on the use of Vioxx—the data at their disposal were not conclusive enough to warrant such a suggestion. However, they did warn doctors to take special care when prescribing the drug to people with heart disease. In their review, the authors stressed that Vioxx and other cox-2 inhibitors could cause serious side effects, and that a broader examination of their impact would be essential. “Given the remarkable exposure and popularity of this new class of medications,” they wrote, “we believe that it is mandatory to conduct a trial specifically assessing cardiovascular risk and benefit of these agents. Until then, we urge caution in prescribing these agents to patients at risk for cardiovascular morbidity.”

ERIC TOPOL IS TANNED and trim, a gangly man in his fifties with an oval face, graying hair that has begun to thin, and the type of relaxed affect that only someone who moves from a climate like Cleveland’s to balmy San Diego could cultivate. Today, Topol has an entirely new kind of job: he is professor of genomics and the director of the Scripps Translational Science Institute in La Jolla. Scripps, one of the nation’s largest biomedical research organizations, is eager to apply the emerging science of genomics—the information contained within our genes—to clinical medicine. Topol believes genetics will soon provide the knowledge we need to make substantial reductions in the incidence of heart disease. And that knowledge, of genetic predispositions and their implications for individuals, is increasing rapidly. Naturally, any reduction in the rate of heart disease, which kills at least a million Americans every year, would have a profound impact on public health. The field is young, and it hasn’t been long since Scripps decided to invest heavily in it. When I visited La Jolla in the spring of 2008, the institute’s building was only partially finished. Several floors consisted of little more than concrete shells and plastic sheeting. The setting, though, was spectacular: Topol’s office looks out onto the Torrey Pines Golf Course, and beyond that, the Pacific Ocean. As I watched from his office window, dozens of people floated by on parasails before gently setting down in the shimmering green sea.

If Topol’s life seems enviable, it hasn’t been that way for long. His persistent criticism of Merck and, by implication, of the FDA, lasted three years, during which time Vioxx killed thousands of people. Topol found himself an outcast in his own profession, shunned for his warnings and eventually driven from the department he made famous. “There were years of sleepless nights, of bitterness,” he said, speaking almost as if he were describing the ordeal of another person. “Years during which I allowed myself to wonder what on earth were we doing to people in the name of science.” He came to realize that Merck possessed data that should have led the company to question Vioxx’s safety long before it was approved.

Research published in the British medical journal the Lancet has estimated that 88,000 Americans had heart attacks after taking Vioxx, and that 38,000 of them died. In testimony before Congress, David Graham, the FDA’s senior drug safety researcher, said the death toll may have reached as high as 55,000—almost exactly the number of American soldiers killed in the Vietnam War. (We will never know for sure; there is no precise way to account for the number of deaths caused by a drug like Vioxx, which was taken by millions of people. It’s easy to notice an increased rate of heart attacks in a large group. Proving with certainty the specific cause of any one of them is nearly impossible.)

“This was one of the most remarkable breaches of trust in American scientific history,” Topol said. He stood up and poured coffee. “It has been years now and even today nothing fundamental has occurred to ensure that this won’t happen again. That is what amazes me. It is as if we took nothing from this tragedy but fear and a sense that those scientists were liars. One of the most important drugs of our time was released without the proper safety checks, and after the company’s own scientists wondered in writing whether it would kill people. When Americans say they are skittish about the drug system and about conventional medicine itself, can anyone really be surprised?”

There have been a raft of studies released about Vioxx—from the Government Accountability Office, the Institute of Medicine, and many private organizations; eventually both the Senate Finance Committee and the House Committee on Government Reform held hearings. Every report, and much of the testimony, described the FDA’s bureaucratic inefficiencies, its reluctance to take controversial positions, and Merck’s willingness to exploit those weaknesses. “The FDA set up an internal safety auditing group for drugs going onto the market,” Topol said. “They have responded by approving fewer new drugs than at any time in their history. The organization is just paralyzed.”

By the end of the millennium, Americans had come to assume that a drug approved by the federal government was a drug they could swallow without wondering whether it would kill them. Vioxx changed all that. Thousands of lawsuits later, in 2007, the company placed nearly $5 billion into a settlement fund. That deposit permitted Merck to avoid nearly fifty thousand lawsuits. It also brought an end to hundreds of class-action cases filed on behalf of dead or injured Vioxx users, which, had they succeeded, could well have put Merck out of business. The settlement was the largest ever made by a pharmaceutical company. As part of the deal, Merck was never forced to admit fault in a single one of those deaths.

The Vioxx episode wove strands of fear and uncertainty together with an inchoate sense, shared by large swaths of American society, that we are ceding control of our lives to technology, particularly to highly sophisticated technology we can barely understand, and that we are doing so at a speed that seems to accelerate every year. Denialism is at least partly a defense against that sense of helplessness. What person, after watching Vioxx kill her husband, wouldn’t say no to the next wonder drug? The story line—a predatory drug company lusting for profits—is not entirely new. The notion of technology as a force that does more harm than good, and of scientists toying with human life, dates back at least to Shelley and Goethe. Rousseau, the first Romantic, longed for the innocence and supposed simplicity of nature. He was convinced that science would have a pernicious effect on society, promising more than it could possibly deliver. G. K. Chesterton, in his book Eugenics and Other Evils, was even more direct, referring to organized science as “government tyranny.”

It would be hard to find many examples in the past four centuries of scientists acting together to threaten humanity. Only a few are necessary, though, and there have been enough dark moments along the remarkable march of progress to generate anxiety and feed denial. Most of those moments were caused by error, not evil. Yes, in 1970 the Ford Motor Company, probably the definitive industrial symbol of twentieth-century America, introduced a car, the Pinto, that its engineers knew was likely to kill passengers. (Before the Pinto was introduced, in what may well stand out as the most remarkable memo in the history of engineering, Ford statisticians argued that the $11 cost of fixing each car added up to more than twice the amount of money—at $200,000 per burn death and $67,000 for each serious injury—that they would have had to pay in lawsuits or settlements.)

More often, there are simpler reasons to question the primacy of science and technology. In the name of improving our lives, some of the smartest people on earth have managed to ruin quite a few of them. DES, or diethylstilbestrol, was the first synthetic estrogen. It was cheap, easy to produce, and unusually potent. First prescribed in 1938, DES was given to women who had experienced miscarriages or premature deliveries. Despite mixed results in the laboratory, the drug was considered safe and effective both for a pregnant woman and her developing fetus. It wasn’t. In the United States, as many as ten million people were exposed to DES by 1971, when it was pulled from the market. “DES Daughters,” as they came to be known, are at increased risk for several types of cancer, as well as structural abnormalities of the reproductive tract, complications with pregnancy, and infertility.

Fear builds far more easily than it dissipates. Ronald Reagan once famously claimed that the “nine most terrifying words in the English language are: ‘I’m from the government and I’m here to help.’ ” If anyone needed a reminder of how far our faith in science had fallen by the end of the twentieth century, Vioxx demonstrated that five other words could prove just as frightening: “Trust me, I’m a scientist.” It was quite a crash. Pharmaceutical companies were among the most highly valued institutions in America after World War II, and it’s not hard to see why. They introduced the core values of consumer culture to American medicine. Drugs became like everything else Made in America: products meant to ease life and solve problems.

The flood of new antibiotics and the rapid development of vaccinations for everything from diphtheria to polio helped define the spirit of the country with a single word: optimism. America was a can-do nation and it possessed technology that would solve the world’s problems. From infectious disease to cancer, and from pollution to hunger, we would overcome it all. Nylon, Lycra, Teflon, Kevlar, and Mylar, for example—all made by DuPont—were all triumphs of ease and modernity. We could fix whatever was broken, cure whatever ailed us, and make life easier for everyone along the way. Today the words “DuPont,” “Merck,” and “Monsanto” are often used as epithets, and they compete with tobacco companies for the role of the most loathed American corporations. Conventional medicine and technology itself, despite their clear rates of success, seem to many people as likely to cause danger as to enhance our lives.

In a Harris poll of attitudes toward corporate America published in 2008, just 27 percent of respondents said they “somewhat or strongly” trusted the pharmaceutical industry. More than half described their views as firmly negative, which places Big Pharma slightly below Big Oil, and a bit above tobacco companies, in the esteem of the average American. The figures for the FDA were only marginally higher. What was written with sincerity only a few decades ago is now played strictly for laughs: in 2006, the satirical newspaper the Onion ran a story in its “Science and Technology” section under this headline: “Wonder Drug Inspires Deep, Unwavering Love of Pharmaceutical Companies.” The year before, a film version of John le Carré’s novel The Constant Gardener was released. Like the book, it portrayed an international pharmaceutical conglomerate as avaricious and cartoonishly evil. The plot was ludicrous and appealed to the worst possible stereotypes of mindless capitalism. But people ate it up.

Vioxx and other preventable catastrophes ensured that they would. Corporations, wrapping themselves in the mantle of progress but all too often propelled by greed, have done more than religion or even Luddism to inflame denialists and raise doubts about the objectivity of science. In 2008, reports surfaced that for more than a year, Merck and Schering-Plough concealed the fact that their jointly marketed cholesterol drug, Vytorin, was no more effective than generic statins costing less than half as much. Nonetheless, the companies still spent more than $100 million during that time to advertise the drug’s special qualities.

News like that has become routine. In early 2009, the British company AstraZeneca was found to have somehow mislaid unfavorable studies about its antipsychotic drug Seroquel. At about the same time, more than one hundred students at the Harvard Medical School publicly questioned the ethics of their professors, some of whom are frequently paid as consultants by the pharmaceutical companies whose products they are supposed to judge. The situation got so bad that the Institute of Medicine, the branch of the National Academy of Sciences charged with providing advice on critical issues of medical research, denounced physicians who accepted money from drug companies. “It is time for medical schools to end a number of long-accepted relationships and practices that create conflicts of interest, threaten the integrity of their missions and their reputations, and put public trust in jeopardy,” the IOM concluded in a report. Just two weeks later, the Scientist revealed that Merck had published a journal filled with favorable articles about more than one of the company’s drugs, without ever bothering to disclose that the publication, the Australasian Journal of Bone and Joint Medicine, was sponsored by the company itself. “To the jaundiced eye, [the journal] might be detected for what it is: marketing,” Public Citizen’s Peter Lurie said. “Many doctors would fail to identify that and might be influenced by what they read.”

If all that wasn’t sufficiently damning, the Baystate Medical Center in Springfield, Massachusetts, revealed that Scott S. Reuben, its highly influential former physician in charge of acute pain treatment, fabricated data from twenty-one medical studies that claimed to show the benefits of painkillers like Vioxx and Celebrex. “The pharmas are in big trouble in terms of credibility,” said Rob Frankel, a brand consultant who focuses on medical industries. “They’re just above Congress and used-car salesmen.”

THIRTY YEARS AGO NOBODY discussed the principal motive behind scientific research: nobody needed to. It was a quest for knowledge. Today, the default assumption is that money matters most of all, and people tend to see science through the prism of commerce. At least until Viagra was introduced, and endorsed on television by Bob Dole, a former candidate for the presidency, no drug had been marketed more successfully than Vioxx. In 2000, the year after it first appeared, Merck spent $160 million advertising their painkiller. They were able to do that thanks to the advent, just three years before, of direct-to-consumer advertising. Only two countries allowed pharmaceutical companies to advertise prescription drugs directly to consumers: New Zealand and the United States.

In America, such ads virtually always consist of glossy promotional materials used to announce major medical advances. (The federal government requires that they include tiny print “information” presented in medical jargon, the meaning of which, for most consumers, is nearly impossible to understand.) These advertisements are not really intended to educate patients, nor to help them become more sophisticated about their own health. They are purely an attempt to get doctors to fill more prescriptions, and they work with stunning regularity. “Blockbuster” drugs like Vioxx, Viagra, and the cholesterol medicine Lipitor can become a multinational corporation’s central source of income.

Our regulatory system encourages companies to invest in marketing, not in research: in the United States, a new drug typically takes a decade to develop and costs hundreds of millions of dollars. With stakes that high, and lawsuits waiting for any company that commits even the smallest error, pharmaceutical firms are far more likely to profit from aggressive sales of products that are already available than from introducing anything new. One reason the ads succeed is that it is nearly impossible to spend a day without seeing one or more major drugs advertised on television. (Another reason is the amount of money companies spend. The marketing budget for AstraZeneca’s heartburn pill Nexium, a sure moneymaker, is bigger than the comparable budget for Budweiser beer.)

Rheumatoid arthritis afflicts more than two million Americans, and it can be devastating. Many of those people were overjoyed when advertisements for Vioxx began blanketing the airwaves in 1999. Suddenly, men who couldn’t bend over were tying their shoes again, walking dogs, and regaining lives that slowly had been consumed by pain. Dorothy Hamill skated across millions of television screens on behalf of Vioxx, overcoming arthritis and moving with the nimble certainty of a teenager at the Olympics. “People dancing in the streets, twirling their partners in joy,” Topol said. “That’s all anybody ever saw.” Go talk to your doctor about Vioxx, the ad would say. And people did, by the millions. In 1996, American pharmaceutical companies spent $11.4 billion on direct advertisements; by 2005 the figure was more than $29 billion. Doctors were overwhelmed with requests, and for the most part were only too happy to comply, writing nearly one hundred million prescriptions for Vioxx alone between 1999 and 2004.

The words “ask your doctor” have become code for “change your prescription.” Often, what people ended up with was no better than the cheaper and more readily available drugs they were taking in the first place. And just as often, the drugs for sale were so new to the market that their safety was hard to gauge. “Americans must face an inconvenient truth about drug safety,” Henry Waxman, the veteran California congressman, said when asked his position on the impact of these advertisements. Waxman is perhaps the most astute congressional observer of American medicine, and since the election of Barack Obama, in his new role as chairman of the House Energy and Commerce Committee, he may also be the most powerful. “The truth is that we inevitably allow drugs on the market whose risks are not fully known,” he said. In 2006, the Institute of Medicine suggested a moratorium on such advertising, a brief pause before permitting companies like Merck to hawk powerful chemicals as if they were Cheerios or vacuum cleaners. That would certainly have saved many of the lives lost to Vioxx.

People tend to see what they are looking for, however, and to millions, pain relief was all that mattered. When it comes to getting a message across, the Super Bowl, where ads for drugs like Vioxx and Viagra have been ubiquitous, will always matter more than the New England Journal of Medicine. In this atmosphere, Eric Topol was largely on his own, and he became the target of Merck’s fullest fury. “Merck went after me with all they had,” he told me. Topol had collaborated with Merck often—at the time of his clash with the company over Vioxx he was actually running one of its trials, for the anti-platelet drug Aggrastat. Like most people in his profession, Topol considered Merck a remarkable place. During one stretch beginning in 1987, it was named by Fortune magazine for seven straight years as the most admired corporation in America, a record that remains unmatched. Merck seemed to prove that profits and decency were not incompatible. “I was in no way predisposed to have negative feelings toward Merck,” Topol said. “In fact it was quite the contrary.” None of that mattered, though, because they came after him with a cleaver. “They said my first paper was ‘data dredging,’” by which they meant a pedantic report full of numbers that proved nothing. “They told anyone who would listen that I had been a fine researcher until the day that paper was published, but at that moment I had suddenly lost it. And at the time all I had said was, ‘Wait a minute, there needs to be some more study here.’ ”

By the end of 2001, however, Topol had moved on. He had begun to focus primarily on genomics and his research switched from treating heart attacks to preventing them. Others were studying whether Vioxx increased the risk of heart attacks, and he felt he had done all he could to address the issue. At the time, Topol had no idea how divisive Vioxx had become at Merck itself. It turned out that scientists there had worried as early as 1996 about the effect the drug would have on the cardiovascular system. The reason for that concern was clear. Vioxx altered the ratio of two crucial substances, the hormone prostacyclin and a molecule called thromboxane, which together help balance blood flow and its ability to clot properly. Suppressing prostacyclin reduces inflammation and pain, and that made Vioxx work. Suppress it too powerfully, however, and thromboxane can cause increased blood pressure and too much clotting, either of which can lead to heart attacks. By 2002, Merck decided to embark on a major study of the cardiovascular risks caused by Vioxx—just as Topol and his colleagues had suggested. The trial would have produced useful data fairly rapidly, but just before it began, the company abruptly scuttled the project. In the end, Merck never made any significant effort to assess the cardiovascular risk posed by its most successful product.

Instead, the company issued a “Cardiovascular Card” to sales representatives. More than three thousand members of the sales force were instructed to refer doctors with questions to the card, which claimed—falsely—that Vioxx was eight to eleven times safer than other similar painkillers. The sales reps were told to produce the card only when all else had failed, to help “physicians in response to their questions regarding the cardiovascular effects of Vioxx.” The FDA, realizing that doctors needed to understand the gravity of the findings from the VIGOR trial, issued a strongly worded letter instructing Merck to correct the record. “Your claim in the press release that Vioxx has a ‘favorable cardiovascular safety profile,’” the letter read in part, “is simply incomprehensible, given the rate of heart attack and serious cardiovascular events compared to naproxen.” The company reacted swiftly: “Do not initiate discussions of the FDA arthritis committee … or the results of the … VIGOR study,” the sales force was told. If doctors proved too querulous, the Merck representatives were instructed to respond by saying, “I cannot discuss the study with you.”

In the summer of 2004, the Lancet asked Topol and the gastroenterologist Gary W. Falk, a colleague of Topol’s from the Cleveland Clinic, to sum up the state of knowledge about Vioxx and similar drugs known, as a class, by the name coxibs. Their editorial was published that August under the title “A Coxib a Day Won’t Keep the Doctor Away. Topol was taken aback when he realized how little had changed. “It was amazing to see that nothing had been done in three years,” he recalled. “It was not even clear that Vioxx protected the stomach. It cost four dollars a day for these darn pills.” On September 29 of that year, Topol happened to dine with Roy Vagelos, Merck’s much-admired former chief executive, who had been retired for nearly a decade. Topol was visiting a New York-based biopharmaceutical company called Regeneron whose board Vagelos chairs. There were few people in medicine for whom Topol had as much respect. “We started talking about Vioxx,” he said. “It was the first time I ever spoke to Roy about it. I remember that conversation well: it was at Ruth’s Chris Steak House in Westchester. Roy went on for a while. He was entirely opposed to the Merck approach to Vioxx. And he didn’t mince words.”

The following morning a Merck cardiologist called Topol and told him the company was removing Vioxx from the market. Another trial had shown that patients taking the drug were at increased risk of heart attack and stroke. That study, APPROVe, began in 2000 as an attempt to discover whether Vioxx helped prevent the recurrence of colon polyps. It didn’t. “I was shocked,” Topol said. “But I thought that it was responsible for them to pull it. And Steve Nissen came down to my office, also very pleased, and said, ‘Isn’t that great? They are pulling the drug.’ We both thought it was the right thing to do.”

For Topol, that could well have been how the story ended. But Merck began to mount a press offensive. The message never varied: Merck put patients first. “Everything they had ever done in the course of Vioxx was putting patients first. All the data was out there,” Topol said, still stunned by the brazen public lies. “This just wasn’t true. It wasn’t right. I called and tried to speak to Ray Gilmartin”—Merck’s chief executive. “Neither he nor anyone else returned my calls.” (That itself was significant: after all, Topol ran one of the most important cardiology departments in the country; he was also the director of a Merck drug trial.) “This was a breach of trust that really rocked the faith people have in institutions like those,” Topol said. “We are talking about thousands of heart attacks. There were simply gross discrepancies in what they presented to the FDA and what was published in journals. I took them on. I had to.”

That week, Topol wrote an op-ed piece for the New York Times. He called it “Vioxx Vanished.” The Times had a better idea: “Good Riddance to a Bad Drug.” Noting that Vioxx increased the risk of heart attacks and strokes, Topol wrote that “our two most common deadly diseases should not be caused by a drug.” He also published a column in the New England Journal of Medicine, called “Failing the Public Health”: “The senior executives at Merck and the leadership at the FDA,” he wrote, “share responsibility for not having taken appropriate action and not recognizing that they are accountable for the public health.”

On December 3, 2005, in a videotaped deposition presented under subpeona at one of the many trials following the recall, Topol argued that Vioxx posed an “extraordinary risk.” A colleague from the Cleveland Clinic, Richard Rudick, told him that Gilmartin, the Merck CEO, had become infuriated by Topol’s public attacks and had complained bitterly to the clinic’s board about the articles in the Times and the New England Journal of Medicine. “What has Merck ever done to the clinic to warrant this?” Gilmartin asked.

Two days after that testimony, Topol received an early call telling him not to attend an 8 a.m. meeting of the board of governors. “My position—chief academic officer—had been abolished. I was also removed as provost of the medical school I founded.” The clinic released a statement saying that there was no connection between Topol’s Vioxx testimony and his sudden demotion, after fifteen years, from one of medicine’s most prominent positions. A spokeswoman for the clinic called it a simple reorganization. The timing, she assured reporters, was a coincidence.

DID THE RECALL of Vioxx, or any other single event, cause millions of Americans to question the value of science as reflexively as they had once embraced it? Of course not. Over the decades, as our knowledge of the physical world has grown, we have also endured the steady drip of doubt—about both the definition of progress and whether the pursuit of science will always drive us in the direction we want to go. A market disaster like Vioxx, whether through malice, greed, or simply error, presented denialists with a rare opportunity: their claims of conspiracy actually came true. More than that, in pursuit of profits, it seemed as if a much-admired corporation had completely ignored the interests of its customers.

It is also true, however, that spectacular technology can backfire spectacularly—and science doesn’t always live up to its expectations. When we see something fail that we had assumed would work, whether it’s a “miracle” drug or a powerful machine, we respond with fear and anger. People often point to the atomic bomb as the most telling evidence of that phenomenon. That’s not entirely fair: however much we may regret it, the bomb did what it was invented to do.

That wasn’t the case in 1984, when a Union Carbide pesticide factory in Bhopal, India, released forty-two tons of toxic methyl isocyanate gas into the atmosphere, exposing more than half a million people to deadly fumes. The immediate death toll was 2,259; within two weeks that number grew to more than eight thousand. Nor was it true two years later, when an explosion at Unit 4 of the V. I. Lenin Atomic Power Station transformed a place called Chernobyl into a synonym for technological disaster. They were the worst industrial accidents in history—one inflicting immense casualties and the other a worldwide sense of dread. The message was hard to misinterpret: “Our lives depend on decisions made by other people; we have no control over these decisions and usually we do not even know the people who make them,” wrote Ted Kaczynski, better known as the Unabomber, in his essay “Industrial Society and Its Future”—the Unabomber Manifesto. “Our lives depend on whether safety standards at a nuclear power plant are properly maintained; on how much pesticide is allowed to get into our food or how much pollution into our air; on how skillful (or incompetent) our doctor is… . The individual’s search for security is therefore frustrated, which leads to a sense of powerlessness.”

Kaczynski’s actions were violent, inexcusable, and antithetical to the spirit of humanity he professed to revere. But who hasn’t felt that sense of powerlessness or frustration? Reaping the benefits of technology often means giving up control. That only matters, of course, when something goes wrong. Few of us know how to fix our carburetors, or understand the mechanism that permits telephone calls to bounce instantly off satellites orbiting twenty-eight thousand miles above the earth only to land a split second later in somebody else’s phone on the other side of the world.

That’s okay; we don’t need to know how they function, as long as they do. Two hundred or even fifty years ago, most people understood their material possessions—in many cases they created them. That is no longer the case. Who can explain how their computer receives its constant stream of data from the Internet? Or understands the fundamental physics of a microwave? When you swallow antibiotics, or give them to your children, do you have any idea how they work? Or how preservatives are mixed into many of the foods we eat or why? The proportion of our surroundings that any ordinary person can explain today is minute—and it keeps getting smaller.

This growing gap between what we do every day and what we know how to do only makes us more desperate to find an easy explanation when something goes wrong. Denialism provides a way to cope with medical mistakes like Vioxx and to explain the technological errors of Chernobyl or Bhopal. There are no reassuring safety statistics during disasters and nobody wants to hear about the tens of thousands of factories that function flawlessly, because triumphs are expected, whereas calamities are unforgettable. That’s why anyone alive on January 28, 1986, is likely to remember that clear, cold day in central Florida, when the space shuttle Challenger lifted off from the Kennedy Space Center, only to explode seventy-three seconds later, then disintegrate in a dense white plume over the Atlantic. It would be hard to overstate the impact of that accident. The space program was the signature accomplishment of American technology: it took us to the moon, helped hold back the Russians, and made millions believe there was nothing we couldn’t accomplish. Even our most compelling disaster—the Apollo 13 mission—was a successful failure, ending with the triumph of technological mastery needed to bring the astronauts safely back to earth.

By 1986, America had become so confident in its ability to control the rockets we routinely sent into space that on that particular January morning, along with its regular crew, NASA strapped a thirty-seven-year-old high school teacher named Christa McAuliffe from Concord, New Hampshire, onto what essentially was a giant bomb. She was the first participant in the new Teacher in Space program. And the last.

The catastrophe was examined in merciless detail at many nationally televised hearings. During the most remarkable of them, Richard Feynman stunned the nation with a simple display of show-and-tell. Feynman, a no-nonsense man and one of the twentieth century’s greatest physicists, dropped a rubber O-ring into a glass of ice water, where it quickly lost resilience and cracked. The ring, used as a flexible buffer, couldn’t take the stress of the cold, and it turned out neither could one just like it on the shuttle booster rocket that unusually icy day in January. Like so many of our technological catastrophes, this was not wholly unforeseen. “My God, Thiokol, when do you want me to launch, next April?” Lawrence Mulloy, manager of the Solid Rocket Booster Project at NASA’s Marshall Space Flight Center, complained to the manufacturer, Morton Thiokol, when engineers from the company warned him the temperature was too low to guarantee their product would function properly.

SCIENTISTS HAVE NEVER BEEN good about explaining what they do or how they do it. Like all human beings, though, they make mistakes, and sometimes abuse their power. The most cited of those abuses are the twins studies and other atrocities carried out by Nazi doctors under the supervision of Josef Mengele. While not as purely evil (because almost nothing could be), the most notorious event in American medical history occurred not long ago: from 1932 to 1972, in what became known as the Tuskegee Experiment, U.S. Public Health Service researchers refused to treat hundreds of poor, mostly illiterate African American share-croppers for syphilis in order to get a better understanding of the natural progression of their disease. Acts of purposeful malevolence like those have been rare; the more subtle scientific tyranny of the elite has not been.

In 1883, Charles Darwin’s cousin Francis Galton coined the term “eugenics,” which would turn out to be one of the most heavily freighted words in the history of science. Taken from a Greek word meaning “good in birth,” eugenics, as Galton defined it, simply meant improving the stock of humanity through breeding. Galton was convinced that positive characteristics like intelligence and beauty, as well as less desirable attributes like criminality and feeblemindedness, were wholly inherited and that a society could breed for them (or get rid of them) as they would, say, a Lipizzaner stallion or a tangerine. The idea was that with proper selection of mates we could dispense with many of the ills that kill us—high blood pressure, for instance, or obesity, as well as many types of cancer.

Galton saw this as natural selection with a twist, and felt it would provide “the more suitable races or strains of blood a better chance of prevailing speedily over the less suitable.” Galton posed the fundemental question in his 1869 book, Hereditary Genius : would it not be “quite practicable to produce a highly gifted race of men by judicious marriages during consecutive generations?” As the Yale historian Daniel J. Kevles points out in his definitive 1985 study In the Name of Eugenics, geneticists loved the idea and eagerly attempted to put it into action. Eugenics found particular favor in the United States.

In 1936, the American Neurological Association published a thick volume titled Eugenical Sterilization, a report issued by many of the leading doctors in the United States and funded by the Carnegie Foundation. There were chapters on who should be sterilized and who shouldn’t, and it was chock full of charts and scientific data—about who entered the New York City hospital system, for example, at what age and for what purpose. The board of the ANA noted in a preface that “the report was presented to and approved by the American Neurological Association at its annual meeting in 1935. It had such an enthusiastic reception that it was felt advisable to publish it in a more permanent form and make it available to the general public.”

Had their first recommendation appeared in a novel, no reader would have taken it seriously: “Our knowledge of human genetics has not the precision nor amplitude which would warrant the sterilization of people who themselves are normal [italics in the original] in order to prevent the appearance, in their descendants, of manic-depressive psychosis, dementia praecox, feeblemindedness, epilepsy, criminal conduct or any of the conditions which we have had under consideration. An exception may exist in the case of normal parents of one or more children suffering from certain familial diseases, such as Tay-Sachs’ amaurotic idiocy.” Of course, for people who were not considered normal, eugenics had already arrived. Between 1907 and 1928, nearly ten thousand Americans were sterilized on the general grounds that they were feebleminded. Some lawmakers even tried to make welfare and unemployment relief contingent upon sterilization.

Today, our knowledge of genetics has both the precision and the amplitude it lacked seventy years ago. The Nazis helped us bury thoughts of eugenics, at least for a while. The subject remains hard to contemplate—but eventually, in the world of genomics, impossible to ignore. Nobody likes to dwell on evil. Yet there has never been a worse time for myopia or forgetfulness. By forgetting the Vioxxes, Vytorins, the nuclear accidents, and constant flirtation with eugenics, and instead speaking only of science as a vehicle for miracles, we dismiss an important aspect of who we are. We need to remember both sides of any equation or we risk acting as if no mistakes are possible, no grievances just. This is an aspect of denialism shared broadly throughout society; we tend to consider only what matters to us now, and we create expectations for all kinds of technology that are simply impossible to meet. That always makes it easier for people, already skittish about their place in a complex world, to question whether vaccines work, or AIDS is caused by HIV, or why they ought to take prescribed pain medication instead of chondroitin or some other useless remedy recommended wholeheartedly by alternative healers throughout the nation.

IF YOU LIVED with intractable pain, would you risk a heart attack to stop it? What chances would be acceptable? One in ten? One in ten thousand? “These questions are impossible to answer completely,” Eric Topol told me when I asked him about it one day as we walked along the beach in California. “Merck sold Vioxx in an unacceptable and unethical way. But I would be perfectly happy if it was back on the market.”

Huh? Eric Topol endorsing Vioxx seemed to make as much sense as Alice Waters campaigning for Monsanto and genetically modified food. “I can’t stress strongly enough how deplorable this catastrophe has been,” he said. “But you have to judge risk properly and almost nobody does. For one thing, you rarely see a discussion of the effect of not having drugs available.” Risk always has a numerator and a denominator. People tend to look at only one of those numbers, though, and they are far more likely to remember the bad than the good. That’s why we can fear flying although it is hundreds of times safer than almost any other form of transportation. When a plane crashes we seeit. Nobody comes on television to announce the tens of thousands of safe landings that occur throughout the world each day.

We make similar mistakes when judging our risks of illness. Disease risks are almost invariably presented as statistics, and what does it mean to have a lifetime heart attack risk 1.75 times greater than average? Or four times the risk of developing a certain kind of cancer? That depends: four times the risk of developing a cancer that affects 1 percent of the population isn’t terrible news. On the other hand, a heart attack risk 75 percent greater than average, in a nation where heart attacks are epidemic, presents a real problem. Few people, however, see graphic reality in numbers. We are simply not good at processing probabilistic information. Even something as straightforward as the relationship between cigarette smoking and cancer isn’t all that straightforward. When you tell a smoker he has a 25 percent chance of dying from cancer, the natural response is to wonder, “From this cigarette? And how likely is that really?” It is genuinely hard to know, so all too often we let emotion take over, both as individuals and as a culture.

The week in 2003 that SARS swept through Hong Kong, the territory’s vast new airport was deserted, and so were the city’s usually impassable streets. Terrified merchants sold face masks and hand sanitizer to anyone foolish enough to go out in public. SARS was a serious disease, the first easily transmitted virus to emerge in the new millennium. Still, it killed fewer than a thousand people, according to World Health Organization statistics. Nevertheless, “it has been calculated that the SARS panic cost more than $37 billion globally,” Lars Svendsen wrote in A Philosophy of Fear. “For such a sum one probably could have eradicated tuberculosis, which costs several million people’s lives every year.”

Harm isn’t simply a philosophical concept; it can be quantified. When Merck, or any another company, withholds information that would have explained why a drug might “fail,” people have a right to their anger. Nonetheless, the bigger problem has little to do with any particular product or industry, but with the way we look at risk. America takes the Hollywood approach, going to extremes to avoid the rare but dramatic risk—the chance that minute residues of pesticide applied to our food will kill us, or that we will die in a plane crash. (There is no bigger scam than those insurance machines near airport gates, which urge passengers to buy a policy just in case the worst happens. A traveler is more likely to win the lottery than die in an airplane. According to Federal Aviation Administration statistics, scheduled American flights spent nearly nineteen million hours in the air in 2008. There wasn’t one fatality.)

On the other hand, we constantly expose ourselves to the likely risks of daily life, riding bicycles (and even motorcycles) without helmets, for example. We think nothing of exceeding the speed limit, and rarely worry about the safety features of the cars we drive. The dramatic rarities, like plane crashes, don’t kill us. The banalities of everyday life do.

We certainly know how to count the number of people who died while taking a particular medication, but we also ought to measure the deaths and injuries caused when certain drugs are not brought to market; that figure would almost always dwarf the harm caused by the drugs we actually use. That’s even true with Vioxx. Aspirin, ibuprofen, and similar medications, when used regularly for chronic pain, cause gastrointestinal bleeding that contributes to the death of more than fifteen thousand people in the United States each year. Another hundred thousand are hospitalized. The injuries—including heart attacks and strokes—caused by Vioxx do not compare in volume. In one study of twenty-six hundred patients, Vioxx, when taken regularly for longer than eighteen months, caused fifteen heart attacks or strokes per every one thousand patients. The comparable figure for those who received a placebo was seven and a half per thousand. There was no increased cardiovascular risk reported for people who took Vioxx for less than eighteen months. In other words, Vioxx increased the risk of having a stroke or heart attack by less than 1 percent. Those are odds that many people might well have been happy to take.

“All Merck had to do was acknowledge the risk, and they fought that to the end,” Topol said. “After fifteen months of haggling with the FDA they put a tiny label on the package that you would need a microscope to find. If they had done it properly and prominently, Vioxx would still be on the market. But doctors and patients would know that if they had heart issues they shouldn’t take it.”

Most human beings don’t walk out the door trying to hurt other people. So if you are not deliberately trying to do harm, what are the rules for using medicine supposed to be? What level of risk would be unacceptable? A better question might be, Is any risk acceptable? Unfortunately, we have permitted the development of unrealistic standards that are almost impossible to attain. The pharmaceutical industry, in part through its own greed (but only in part), has placed itself in a position where the public expects it never to cause harm. Yet, drugs are chemicals we put into our body, and that entails risks. No matter how well they work, however, if one person in five thousand is injured, he could sue and have no trouble finding dozens of lawyers eager to represent him. People never measure the risk of keeping the drug off the market, though, and that is the problem. If you applied FDA phase I or II or III criteria—all required for drug approval—to driving an automobile in nearly any American city, nobody would be allowed to enter one. When we compare the risk of taking Vioxx to the risk of getting behind the wheel of a car, it’s not at all clear which is more dangerous.