An excerpt from Unhinged: The trouble with psychiatry- a doctor’s revelations about a profession in crisis. Copyright © 2010 Daniel Carlat. Excerpted with permission by Free Press, a Division of Simon & Schuster, Inc.
It may be hard for people to imagine that drug reps can get us to prescribe a particular drug by simply giving us a meal or being nice to us. The process is subtle, but it happens all the time. To quote Ahari again: “Drug reps increase drug sales by influencing physicians, and they do so with finely titrated doses of friendship.”
Valerie, for example, was my Ambien rep. Ambien, marketed by Sanofi-Aventis, was the first in a new category of sleeping pill, the “non-benzodiazepines.”
Standard sleeping pills, like Valium or Restoril, are benzodiazepines, and they work by latching onto brain receptors for a naturally occurring neurotransmitter called GABA. This latching on boosts GABA’s effects, and GABA is unique because instead of activating neighboring neurons, it inhibits them. Thus, benzos enhance the quieting properties of GABA, and the result is a sensation of sleepiness and relaxation. Ambien, Sonata, and Lunesta also bind to GABA receptors, but they are more selective in what part of the receptor they bind to. Because of this selectivity, non-benzos cause mainly sleepiness and less relaxation. This, in turn, makes them less addictive than benzos, which is their only genuine advantage.
I first met Valerie about a year before Ambien was slated to go off-patent. Sanofi had generated a new medication to replace it, Ambien CR (short for “Controlled Release”), and was marketing it as an improved version of Ambien. As is common when companies produce a slightly altered version of their old model, Sanofi had dropped all marketing efforts for Ambien, which only several months earlier had been the greatest gift to sleep since the Serta mattress. Paradoxically, Ambien was now the enemy, even more of a threat than other companies’ products. Why? Because if Sanofi did not convince doctors to ditch Ambien and switch patients to Ambien CR soon, Ambien would go generic before this “conversion” could take place. Once Ambien went generic, its price would plummet, and insurance companies would pressure doctors to stick with older pills. Ambien CR would be left in the dust.
“This new formulation has two layers,” explained Valerie. “The first layer has 60 percent of the medication, and it dissolves right after your patients take it, so they can get to sleep right away. Then the second layer dissolves gradually throughout the night, so patients can stay asleep.”
“But is it any better than regular Ambien?” I asked.
“We have the plasma concentration here.” She quickly pulled an article out of her briefcase and pointed out a graph in the paper. It was a comparison of the blood levels of Ambien CR with regular Ambien, based on a study done by Sanofi-Aventis in which subjects were given either one of the pills and then had their blood drawn every hour for the next eight hours. There were two overlapping curves, each shaped a bit like an elephant’s head and trunk. Each of the two curves started with a big hump representing an initial peak in Ambien levels within two hours of ingestion, and then each tapered down through the next several hours, as each drug was gradually metabolized out of the bloodstream. The curves coincided up to the two-hour peak, after which the Ambien CR blood levels were somewhat higher.
“As you can see,” she pointed out, “Ambien CR had a higher concentration later in the night than Ambien.”
Looking at the small print, I could see that this was data from a small “phase 1” study, in which the drug was given to a small group of volunteers to measure how quickly it was metabolized. The study didn’t include any measures of sleep. In fact, the volunteers were barely sleeping at all. They had to be awake throughout the study to have their blood sampled.
“Do you have any data comparing Ambien and Ambien CR on patients with sleep problems?” I asked. “This graph shows that the blood levels are a little higher with Ambien CR, but it doesn’t tell me that real patients actually stay asleep any longer than when they take regular Ambien.”
Valerie furrowed her brow, grimacing a bit. “I don’t think we have any of those studies yet. But this plasma concentration data is pretty persuasive.”
Not to me it wasn’t. I had already seen the main Ambien CR studies and I was skeptical that this represented any great leap forward in sleeping pill science. Yes, the new pill looked promising on paper, and I had prescribed it to a few patients. Some of them said it worked well, some said it didn’t put them to sleep as quickly as other pills, while others said that it worked too well as a sleeping pill, leaving them feeling groggy for several hours in the morning after waking up.
Valerie was persistent, visiting me as often as once a month, because she knew I wasn’t prescribing very much Ambien CR. In fact, she knew exactly which sleeping pills I favored and which ones I ignored, down to the percentage of all pills I prescribed. Since the 1990s, such data has been offered for sale by pharmacies to drug companies. The drug companies pass it all on to their drug reps, who can access it from their laptops in the car before they visit a doctor. Because of this practice, which is called “prescription datamining,” drug reps arrive at doctors’ offices “armed and dangerous,” according to Kathleen Slattery-Moschkau, a former rep for both Bristol-Myers Squibb and Johnson & Johnson. “They know what percentage of Prozac or Paxil a doctor prescribes,” she said in an article in the British Medical Journal. “The doctor often doesn’t know this and it gives [reps] an incredible advantage over the doctor.”
In my case, I was well aware that Valerie had this information, because four years previously, in 2002, I had spent a year giving talks for Wyeth, the drug company that markets Effexor. I would accompany Wyeth reps on their visits to doctors. They would provide the lunch, and I would give a mini-lecture on the virtues of Effexor. I eventually resigned from the speakers’ bureau. I came to believe that I was deceiving doctors, and the experience gave me an unusual peek inside the prescription data-mining racket. Before I gave my talks, the drug reps would fax me prescribing information about particular doctors. One note, for example, informed me that the physician we’d be visiting that day was a “decile 6 doctor and is not prescribing any Effexor XR, so please tailor accordingly. There is also one more doc in the practice that we are not familiar with.” The term “decile 6” is drug-rep jargon for a doctor who prescribes a lot of medications. The higher the “decile” (in a range from 1 to 10), the higher the prescription volume and the more potentially lucrative that doctor could be for the company. A note from another rep reminded me of a scene from Mission: Impossible. “Dr. Carlat: Our main target, Dr. S., is an internist. He spreads his usage among three antidepressants, Celexa, Zoloft, and Paxil, at about 25–30 percent each. He is currently using about 6 percent Effexor XR.”
I also learned that drug reps have a list of slang terms for doctors, all based on how valuable they could be for the company. A “high writer” referred to a doctor who wrote many prescriptions, whereas a “low writer” wrote very few, and was unlikely to be worth visiting. A “spreader” wrote prescriptions equally among all medications in a particular class of drugs, and therefore would need a marketing pitch designed both to push the rep’s drug and undermine the competition. A “no see him” was a doctor who refused to see reps at all, while a “sample grabber” didn’t see reps but still requested that they drop off free samples of medications.
I assumed that Valerie saw me as a “high writer” who needed some consistent prodding to get my Ambien CR numbers up. One day she handed me a brochure entitled “Medical Textbook Program.” It read: “Sanofi-Aventis is committed to your ongoing medical education. Therefore, to assist you, we’re offering the opportunity to select one of the following medical resources, compliments of Sanofi-Aventis.”
I looked up from the brochure, and smiled. “This is great—thank you.”
There were eleven options to choose from, and I ordered a small book on the basics of neuroscience. A few weeks later Valerie personally delivered my book—a modest paperback that sells at Borders for about $25. I thanked her, and Valerie left the office reminding me, “Don’t forget about Ambien CR!”
Later that day, a patient came in complaining of insomnia. As I thought about the half dozen or so medications I could use, the image of Valerie and her gift popped into my mind. And I decided, “Why not try Valerie’s drug for this patient?”
I was not consciously rewarding Valerie; it was more subtle than that. Gradually, Valerie had become a friend of sorts. Her face was familiar, she was always nice to my secretary, her presence was a brief reprieve from my busy days, and she had just given me a gift. It is a natural human desire to reciprocate when somebody does something nice for you. Conscious of it or not, I prescribed Ambien CR for this patient in order to return the favor.
As it happened, he didn’t like the drug because of a hangover side effect that caused him a few difficult days at work. What if he had known that I had prescribed the drug to him as a favor to a drug rep? He would have been irate, and justifiably so. Patients like to think that we prescribe based on a careful review of the medical science, not because a likable rep just visited our office and handed us a book.
But this is how salesmanship in the pharmaceutical industry works. Gifts have become a central part of drug company marketing, so integral to the drug rep–physician relationship that the gifts are often not considered gifts. In a review of all the studies that had been published about the impact of drug company gifts on physician behavior, Dr. Ashley Wazana found that the practice is pervasive and highly effective. It begins in medical school and residency. One study found that trainees meet with drug reps on the average of four times a month, and another study of medical residents found that they are receiving an average of six gifts per year from reps. These gifts are generally quite small, with an average per-gift value of $20, well below the AMA’s current guidelines of a maximum value of $100 per gift.
Does the size of a gift to a doctor even matter? While formal studies of this issue have not been done, the story of the Hare Krishnas’ business plan implies that tiny gifts, even when they are unwelcome, elicit a feeling of obligation to give back. Members of the religious cult used to frequent airports and would thrust trinkets into the hands of passers-by, saying, “This is my gift to you.” Although the gifts were frequently unwanted and thrown away, the strategy encouraged people to make a small monetary donation, yielding enough income to make the group a wealthy organization. When drug reps give any gift, from a pen to a lunch to a luxury vacation, they are instilling a feeling of reciprocity in the doctor. It is natural to want to complete a gift exchange—in fact, social scientists theorize that social rules of reciprocity are an important glue holding societies together. Valerie’s gift of a $25 book was a single strand in her efforts to build a relationship based on giving back, and she got at least one prescription for Ambien CR out of the bargain.
While drug companies may be reluctant to do head-to-head studies of the effectiveness of their drugs versus competitor drugs, they are all too willing to conduct studies to disparage competing drugs.
I became aware of this tactic when I ran across two journal articles focusing on the effectiveness of trazodone as a sleeping pill. Both were published in the widely read Journal of Clinical Psychiatry, both purported to be reviews of the evidence, and both reviews bashed the drug with the energy of a critic panning a lousy Broadway musical. The studies piqued my interest because, despite Valerie’s best efforts, I still commonly prescribed both trazodone and Restoril, two older pills that had always worked well for insomnia, and which cost just pennies per dose. Restoril is a benzodiazepine, a type of tranquilizer in the same drug class as Valium and Xanax, while trazodone is one of the older antidepressants. The effects of trazodone last longer than most other sleeping pills, and the drug has no addictive potential. Also, unlike Ambien, it rarely makes patients feel loopy and high. These qualities make it extremely popular among psychiatrists—it is one of the most widely prescribed sleeping pills in the United States.
Because I prescribed trazodone so much, I read the articles that were so critical of it carefully. Their points struck me as exaggerated and unfair. They said it lacked high-quality research data on its use as a hypnotic. But what they left unmentioned is that trazodone was approved almost thirty years ago as an antidepressant, and since it is no longer patented, no drug company stands to profit from doing large clinical trials.
They dusted off older studies highlighting some of trazodone’s side effects, like cardiac arrhythmias or priapism (prolonged painful erections). But these side effects are extremely rare: Priapism occurs in one in five thousand men, and the incidence of cardiac arrhythmias is even lower. I had never seen either side effect in the hundreds of patients for whom I had prescribed it.
Why did these psychiatrists hate trazodone so much? Once I took a close look at the small print, it became clear. Each of the authors had worked as a consultant for drug companies that were marketing the newer sleeping pills, such as Ambien, Sonata, and Lunesta. In fact, one of the articles had apparently been written by employees of Sepracor, maker of Lunesta, and the doctor may have been paid to be listed as the author. Whether he had done any of the actual writing
was unclear. Incensed by these sleazy practices, I wrote an op-ed piece for the New York Times, which generated an angry response from one of the academics I identified—but no denial of the facts as I laid them out.
Apparently, executives at the drug companies had discovered a new marketing tactic: Pay doctors to trash your generic competitors. I imagined the following conversation at one of the companies:
Executive A: “Trazodone’s a problem. It’s eating into our business.”
Executive B: “So what are we going to do about it?”
Executive A: “We need to get some anti-trazodone articles into the journals.”
Executive B: “Brilliant. I know just the medical ghostwriter who can write the first draft. But we’ll have to find an academic psychiatrist to put his name on it.”
Executive A: “No problem. Dr. X is on our advisory board. He’ll do it for a few thousand. Then we’ll help him submit it to the journal. The journal will be happy to publish it, because we’ll promise to order thousands of reprints so our reps can distribute them to psychiatrists in their territories.”
While that conversation is a fantasy, drug companies’ manipulation of the medical publishing world for marketing has been well documented. For example, in 1999, as part of a lawsuit brought against Pfizer, it was revealed that the company had hired an agency called Current Medical Directions to write scientific papers about their antidepressant Zoloft. As detailed in an article later published in the British Journal of Psychiatry, in 1998, 1999, and 2000 the agency’s ghostwriters prepared fifty-five articles and recruited high-profile psychiatrists as “authors.” Some of the articles were completed and ready for submission to journals well before a willing academic was located. In such cases, the documents initially listed the name of the author as “TBD”—“to be determined,” and once willing prospects were located, they were paid an honorarium to affix their names to them.
The articles prepared by Current Medical Directions were published in the nation’s top journals, including the American Journal of Psychiatry and the Journal of the American Medical Association. Only two of the fifty-five articles disclosed to readers that the alleged doctor-authors were paid by the agency or that medical writers were involved. Most astoundingly, these articles outnumbered those written in the old-fashioned way, comprising over half—57 percent—of all articles published about Zoloft in the entire medical literature from 1998 to 2000. Thus, for at least one antidepressant, the bulk of the medical literature was literally written by the drug company that manufactured the drug, which is about as glaring a manipulation of science as one can imagine.
Unsurprisingly, research sponsored by companies almost always produces positive results for the sponsor’s drug. This has been documented in research on a vast array of treatments, including antidepressants, antipsychotics, birth control pills, arthritis medications, and drugs for Alzheimer’s dementia. A meta-analysis of all such studies found that drug company–sponsored studies were four times more likely to produce a favorable outcome for the sponsor’s drug than studies with other funders. The whole process of pharmaceutical research reminds me of a quote from Otto von Bismarck: “Laws are like sausages, it is better not to see them being made.”
But the story gets worse. Not only do drug companies control how so many studies are designed and written, they also routinely hide studies that don’t look good for their drugs. While this practice, called “publication bias,” has long been suspected, the vast scope of the technique was only uncovered in January of 2008, when an Oregon psychiatrist named Erick Turner and colleagues published an astonishing paper in the New England Journal of Medicine. Turner was uniquely suited to dig up this data because he used to work at the FDA reviewing psychiatric drug trials submitted by companies. In an interview, Turner told me that the idea for his study came to him partly because his psychiatrist colleagues seemed overly convinced of the effectiveness of antidepressants.
“People seem to view drug efficacy as a black-or-white phenomenon and, because these drugs were approved by the FDA, they must be effective, and that’s that. This view gets reinforced when they look in medical journals and see nothing but positive study results,” he said. “I would say they fail to beat placebo in 40 percent to 50 percent of clinical trials, and they would say, ‘What are you talking about? I have never seen a negative study.’”
But Turner knew differently, because during his time at the FDA he had reviewed many clinical trials with negative results, and knew that many were never published. He made it his personal project to try to find out exactly what percentage of these clinical trials were discreetly stuffed into a file drawer rather than published in a medical journal. It wasn’t easy, because he no longer worked at the FDA, and therefore no longer had access to the submitted data. He and his colleagues spent hours searching the bowels of the FDA Web site, submitting requests through the Freedom of Information Act (FOIA), and contacting researchers who had previously gotten FDA data through FOIA requests.
Eventually, Turner and his colleagues tracked down the fate of all the research that had been submitted to the FDA about twelve newer antidepressants approved from 1987 to 2004. This included all the SSRIs, both of the dual reuptake inhibitors (Effexor and Cymbalta), and several other antidepressants, such as Wellbutrin, Remeron, and Serzone. These twelve drugs were approved on the basis of seventy-four clinical studies that the drug companies submitted to the FDA. Thirty-eight of the studies were positive, meaning that the antidepressant was significantly more effective than a placebo. Almost all of these (thirty-seven of thirty-eight) had been published in medical journals—no surprise there. However, thirty-six studies showed either negative or questionable results. Of these studies, twenty-two were never published, eleven were published in a way that inaccurately conveyed a positive outcome (“spun,” if you will), and only three were published accurately as negative studies.
From the standpoint of a psychiatrist like me trying to figure out which drugs work, the bottom line was this: If I relied on the published medical literature for information (and what else can I rely on?), it would appear that 94 percent of all antidepressant trials are positive. But if I had access to all the suppressed data, I would see that the truth is that only about half—51 percent—of trials are positive.
Turner called this the “dirty little secret” of the psychiatric world. If there is only a fifty-fifty chance that an antidepressant drug study shows the drug to be effective, what does this mean for patients taking these drugs? Is everybody taking a placebo, and only fooling themselves into believing these drugs work? Fortunately, the implications of Turner’s findings are not quite this dire, because the fact is that patients recruited into antidepressant studies are different from patients who arrive on the doorstep of my office. To get into a study, patients have to meet a series of criteria more exclusive than those of an Ivy League college.
Companies have found by experience that if they want to be sure their drug outperforms a placebo, they have to be very picky about which patients are allowed into the study. They want patients with “pure” depression, unblemished by messy problems such as alcohol use, anxiety problems, or bipolar disorder. Furthermore, because of reasonable worries about the safety of patients who might be assigned to a sugar pill, these studies exclude patients with suicidal thoughts. Other common exclusion criteria include an active medical illness, depression that is too mild, or conversely, depression that has lasted too long.
As I consider that list of exclusions, I have a hard time picturing a single one of my private practice patients who would qualify. Mark Zimmerman, a psychiatrist at Brown University, had this same thought, and he decided to test it. He identified 346 depressed patients who had shown up for treatment at Rhode Island Hospital’s department of psychiatry. Then he pretended that each one was applying for a spot in a typical antidepressant research study, and applied each of the many exclusion criteria commonly used. Only twenty-nine out of the 346 patients, or 8.3 percent, would have gained entry into this exclusive club. The only Ivy League college that matches this is Harvard, which had a 7.1 percent admission rate in 2008.
To find such unusual patients, researchers who are funded by drug companies have to run a media blitz of newspaper and radio ads to encourage them to come out of the woodwork. The bottom line is that antidepressant research studies are not generalizable to real patients, meaning that few of their results, whether positive or negative, are reliable indicators of what would happen to your mood on antidepressants.
In fact, if you ask any psychiatrist in clinical practice, including me, whether antidepressants work for their patients, you will hear an unambiguous “yes.” We see people getting better all the time. True, much of this response is undoubtedly due to the placebo effect, but it would be deceptive for me to prescribe a sugar pill to my patients while telling them that it is a real medication. So I am stuck with prescribing active psychotropic drugs in order to activate the placebo, with the main disadvantage being that such drugs have far more side effects.
Yes, I believe that psychiatric medications work, but this does not let the drug companies off the hook. In perpetrating deceptive marketing practices that manipulate the very science that doctors depend on to make treatment decisions, drug makers have betrayed us all.
Companies have told us that newer antidepressants with fancier neurochemical profiles are more effective than older medications, but unbiased studies funded by the NIMH have disputed this. They have tried to convince us that the newest and most expensive antipsychotics are an advance in science, when in reality they are no more effective than the first antipsychotic, Thorazine, and in some cases they cause more dangerous side effects. They have pushed the new sleeping pills by paying doctors to write pseudo-academic reviews bashing older medications that work as well and are sold at a fraction of the cost. In all these cases, when their studies have yielded findings at odds with their promotional needs, they have often locked the studies away in the file drawers, hiding crucial data from psychiatrists and patients.
Doctors are now fighting back. Beginning in 2004, the International Committee of Medical Journal Editors announced an ingenious plan to prevent companies from hiding negative findings. They set up an online registry (at www.clinicaltrials.gov) where drug companies are expected to post information about all of their studies, regardless of the anticipated outcome. Any study that was not publicized ahead of time will not even be considered for publication in the top journals. The system appears to be working, as there are currently over eighty-two thousand trials registered.
There are other reforms in the works as well. The system of prescription data-mining is under serious attack, with eighteen states considering laws to ban the practice in 2008. New Hampshire passed such a law in 2007, and in 2008 successfully fought off the efforts of industry attorneys to overturn the new ban. The pharmaceutical industry itself is also taking some initiative. In 2008, they strengthened the code of conduct for pharmaceutical reps, banning small gifts (Valerie’s book is now off limits) and limiting the kinds of meals reps can buy for doctors.
These are positive developments, but there is another type of pharmaceutical representative whose actions remain completely unregulated. These reps have unfettered access to the top academics of all fields of medicine, are invited by medical societies to give keynote addresses, routinely publish articles in the best journals, and offer advice about medications that is accepted as gospel by doctors everywhere.
These reps have medical degrees, and some have become millionaires by taking fat payments from drug companies. These are the hired guns of medicine, and they are the subject of the next chapter.
Daniel Carlat is a psychiatrist and author of Unhinged: The trouble with psychiatry- a doctor’s revelations about a profession in crisis.
Submit a guest post and be heard.