One morning late in December—three full months after Diane was Floxed—I received several small brown envelopes in the mail from the U.S. Department of Health and Human Services. Inside were dozens of four-by-six-inch clear plastic cards covered with microscopic blue lettering. They were copies of the FDA’s files on Floxin, on microfiche. The accompanying letter apologized for the agency’s reliance on such an outdated technology and suggested that many public libraries still had their old microfiche readers.
I was startled to say the least. My technophobic mother was using a computer for her holiday recipes, yet the FDA hadn’t computerized its applications for new drugs?
The main public library in Philadelphia—the palatial building just down the parkway from the Art Museum—isn’t far from my office, so I started going over there in the afternoons and sifting through the microfiche on one of the old gray-metal readers in the periodical room. The contraptions had ancient photocopying machines attached to them, and for a quarter you could make a fuzzy copy of a fuzzy page.
The FDA files contained close to a thousand of those fuzzy pages. The first six hundred were the agency’s review of the New Drug Application (NDA), the permanent record of the studies that were the basis of the drug’s approval, including the human studies, the “clinical trials,” which are divided into Phases I, II and III. It’s all there in brain-curdling detail, except for the occasional line that has been deleted—just crossed out, really—because it might give away something the company considers a trade secret. The finished NDA review combines an abridged version of the truckload of information submitted by the drug company with a running commentary from the FDA medical officer assigned to ride herd on the application. The medical officer’s observations are footnoted at the bottom of each page like a teacher’s comments, so you can watch some of the scientific give-and-take between the company and the FDA.
The first page that caught my eye was number forty, detailing the results of a Phase I study of safety and pharmacokinetics in thirty-two healthy male geriatric subjects in Senneville, Quebec. (Pharmacokinetics involves how a drug is absorbed, as opposed to pharmacodynamics, which involves the actions a drug causes after being absorbed.) Study number E84-070 was a “randomized, double-blind, placebo-controlled” study, which meant the group was randomly divided, with half put on the drug and half on a sugar pill, with no one, including the doctors, knowing which was which until the “blind” was broken at the conclusion.
I found myself less interested in the test results, however, than in their analysis. The drug company had told the FDA there were “no clinically significant changes” in vital signs or in any neurological, ophthalmologic, heart, blood or urine tests. The company also told the FDA that the differences in the “subject incidence of adverse experiences” between the two groups was “not significant.”
But the medical officer noted—in boldface underlined type, to drive home the point—that included among these “insignificant” adverse experiences was one subject who had to discontinue taking Floxin because of dizziness, tinnitus (ringing in the ear), nausea, vomiting and faintness. The subject had taken only one ofloxacin pill—at a dose of 100 milligrams higher than the one Diane had taken, but a dose the company does sell—and experienced both a hearing loss in one ear and an ongoing loss of balance that persisted even at the six-month checkup.
In fact, the medical officer noted, the Floxin subjects had “3.4 times more adverse reactions than the placebo group.” Yet the drug company said the difference between the groups was “not significant.”
Except for writing that little underlined commentary, I saw nothing to indicate that the medical officer took any other action to address the gap between the two interpretations of “significant”—a gap that seemed roughly the size of the crack my wife had fallen through.
Suddenly, I felt a little sick. But I wasn’t sure if my nausea was caused by what I was learning about the drug or what I was learning about drug approval.
Until I started reading the application, I had actually always assumed that the FDA itself tested drugs before approving them. Or at the very least, I assumed that drug companies did preliminary testing and then the FDA came and redid the tests in its own state-of-the-art labs. Or at the very, very least, I assumed that the FDA vigorously reinvestigated every facet of the results presented by the drug companies.
Instead, I found myself reeling from the realization that all the agency did was read summaries of the results of drug company tests, offer a few comments and occasionally ask to see the raw data on which the summaries were based. Many of the medical officer’s complaints involved clerical errors. The drug company seemed more likely to get in trouble for sloppy paperwork than for sloppy analysis of drug safety.
Whenever I found something interesting in the New Drug Application, I would run down the library’s broad, worn-slippery marble stairs to the pay phone and call Diane to tell her about it. For example, there was a chart showing that females tended to have a higher incidence rate for some types of adverse effects, including CNS reactions, than males. Or an individual case would catch my eye, like the patient the medical officer singled out because he had been forced to discontinue the drug after severe hallucinations.
Further back in the application, the FDA medical officer was interested in ofloxacin patient 12311, who participated in a study of the drug’s effects on uncomplicated, mixed sexually transmitted diseases. The report said that patient had a drug reaction requiring hospitalization for psychiatric treatment. The company had listed the reaction as severe on one summary chart but not in the results of the study itself.
I was interested in this case for another reason. It was the first reference I had seen to a hospitalization for a CNS reaction to the drug, and the study had taken place in Philadelphia, at Temple University. I thought the principal investigator in the study might know something useful for Diane’s case. Even though the study was several years old, I assumed that a patient who ended up in a psychiatric hospital during a simple clinical trial for an antibiotic would be memorable. So I tracked the doctor down to his private practice in the Philadelphia suburbs. When I finally got him on the phone, he said it was a long time ago and he didn’t recall any such patient in his trial.
But he inadvertently gave me a chilling insight into the drug development process. “I was principal investigator, but I was really just pushing papers,” he said, sounding distracted. “We had a fleet of residents who were prescribing and signing patients up. I was just overseeing the collection of data. I had very little contact with the patients. The bureaucracy at the hospital was such a mess, and it was six or seven years ago. I couldn’t really tell you any of the people who would have been handling the patient care. We were working with four or five chief residents, but I can’t really remember anyone specifically.
“Doing that kind of research really isn’t one of my passions, you know. That’s why I don’t do it anymore.”
That’s hardly the level of enthusiasm and intellectual engagement I’m looking for in the scientists who test the drugs I take.
What was interesting about the New Drug Application was that you could see which cases had led to which warnings on the label. A patient’s experience would be mentioned, the medical officer would take note of it, and that’s all it took: it was on the package insert. If only one patient in the trials reported a certain reaction, it was on the label in the singular: seizure, manic episode. If more than one, the event made the label in plural, and the drug caused seizures. Very literal, very legalistic.
At one level, I found this very encouraging: if a single patient’s experience could change the label, then I felt better about my decision to make such a big deal about my one patient. At another level, however, this was scarier than not knowing how something got on the label. Because I was just beginning to realize how little is generally known about a drug before it comes on the market.
Exactly 3,296 patients took Floxin during the clinical trials leading to the drug’s FDA approval. To me, that seemed like an absurdly, even shockingly low number. But then I paged through the PDR, where some companies (but not Ortho) list the number of patients who initially tested their drugs. It turned out that 3,296 wasn’t unusual at all. In fact, when compared with other antibiotics, that 3,296 is actually a bit above average.
But a drug reaction that doesn’t present itself in one of 3,296 patients could still be pretty devastating when the drug reaches the general population. Just do the math. What if the drug causes one in 3,297 people to just drop dead?
What if it causes one in 10,000 to just drop dead? And what if it is then taken by 10 million people? Well, a thousand people could drop dead. If a thousand people dropped dead from a drug, would it be yanked from the market? Should it be yanked from the market? That is assuming, of course, that doctors ever figured out that the drug was responsible for killing their patients, and then bothered to report the deaths as drug reactions.
Ever since I first read the PDR, I had wondered about this phrase that appears in the Adverse Reactions section of every drug label: “less than one percent.” It seemed to be used rather dismissively, which I found strange considering what an arbitrary threshold it was. In all the Floxin trials, for example, if thirty-two people had a certain reaction, it was “less than one percent.” But if thirty-three had it, it would be viewed differently.
I knew a little bit about how people misuse numbers because I had once done a story on John Allen Paulos, the Philadelphia mathematician who wrote the best-selling book Innumeracy. I called him to see if he could offer any insight into “less than one percent.” He hadn’t really thought about it before, but he pointed out that “ ‘less than one percent’ isn’t that small a number, even as a percentage. It’s probably just a cover-all term that really means at most one percent, but probably one in thousands.” The problem, which he tends to see everywhere, is putting the numbers in some context.
“Generally, the people who come up with these numbers are some obscure scientists in some back corner of the company,” he said, “and the people who defend the numbers are lawyers who couldn’t add two fractions if you paid them. The people who know aren’t brought before the public, because the people in the company are afraid they might say something true.… You know, it’s like ‘our invisible people have told us …’ ” On the other hand, he said, a lot of psychological issues are involved in the reporting of medical complications—and whenever numbers rely on self-reporting, there are always confounding factors and the possibility of hysteria.
Paulos told me one of his favorite stories about math and medicine, involving his wife, Sheila, and a minor surgical procedure she was considering. “We went to the doctor, and he said the procedure had a one-in-a-million risk, it’s nothing,” he recalled. “The next time he said, ‘It’s ninety-nine percent safe.’ And as we went out the door, he said, ‘It usually goes quite well.’ ” The doctor seemed oblivious to the fact that he had just cited three vastly different risk/benefit ratios.
“Presumably doctors know the risks,” Paulos said. “But I’ve found they bandy about numbers without any reference to their literal meaning. They say what sounds good.”
What didn’t sound good to me was that even though most of the patients in the clinical trials had taken Floxin without incident, the drug company seemed to be arguing that almost every adverse reaction was somehow “insignificant” or “not drug related.” The medical officer seemed to be disagreeing in several instances, but it didn’t appear to matter. The drug was approved, and nothing I could see on the label or in the promotional materials would lead a doctor or a patient to wonder if there was anything unique about the kinds of adverse reactions Floxin caused, or the kinds of patients who suffered them. Yet this was a class of drugs with unique CNS side effects: even the Japanese company that invented Floxin said so.
I finally decided to simply call the medical officer myself, to try to understand how the process works. Dave Flockhart helped me track down the doctor in what I was quickly realizing was the very small world of drug approval (and reproval) based in and around Washington.
Along the way, I spoke to several people at CDER, the FDA’s Center for Drug Evaluation and Research. The agency has other divisions for food products, medical devices and vaccines (or “biologics”), but CDER regulates medicinal drugs and is the umbrella over government efforts to rule on drug applications and labeling materials, analyze drug advertisements, keep track of adverse drug reactions and investigate violations of drug laws. The medical divisions are divided by drug type. Floxin had been approved by Anti-Infectives. Many of the drugs Diane was now taking had been approved by Neuropharmacology.
I found myself pleasantly surprised at how nice everyone was whom I spoke with at CDER—secretaries, medical officers—and how curious they seemed about Diane’s situation. I don’t think they were just preening for the press, because I wasn’t really calling them as a reporter; they knew I worked for magazines, but they also knew that I was inquiring about a personal matter. They seemed patient and encouraging but also eager to be understood, or aware that they were misunderstood. They told me about drug reactions in their own families—“One of my kids took a drug,” one explained, “and something I know can happen unfortunately did happen to her”—and tried to make me understand what they grappled with.
Finally, I hooked up with the medical officer who had overseen the Floxin application and whose commentary I had been psychoanalyzing at the library. Her name was Dr. Ana Szarfman, a reviewer in the anti-infectives division of CDER. She spoke hurriedly and in flurries, with a thick Eastern European accent. I had just started telling her a little bit about how I had become interested in Floxin when she asked if my wife had ever had a head trauma of any sort. I explained that Diane had been in two auto accidents when she was younger, one of which resulted in a concussion and a loss of consciousness. Stacy Phillips had told me something similar had once happened to her.
“Well, that’s it, that’s it,” Szarfman said.
“What do you mean, that’s it?” I replied. “Based on your understanding of Floxin, that could be enough to lead to this?”
“That’s why, oh yes.”
Well, hell, I thought to myself, shouldn’t that be on the label? Don’t more than one percent of the population ever hit their heads really hard? Shouldn’t a doctor be told to ask about head trauma before giving the drug?
And what else do medical officers know about the drugs we take that never turns up on the label?
I inquired about some of her printed comments in the Floxin NDA. “In many of my comments,” she said, “you must understand, I was a new reviewer. You need to realize, this was my first review.”
Oh, great. She seemed almost apologetic about some of the comments she had made in the NDA, but in a very curious way. I couldn’t tell if she felt she had been unnecessarily harsh about the inconsistencies in Floxin drug trials or whether she was just remembering how honest and bold she had been when she first came to the agency. Did she now feel her comments were rookie mistakes, or perhaps examples of the admirable youthful exuberance that bureaucracy eventually leaches out of you?
She said she had to run to a meeting. But before she hung up, she left me with a cryptic admonishment. “If you’re thinking about this drug, if you’re going to write about this drug, I am glad,” she said. “Just be careful.”
After surviving the six-hundred-page review of the Floxin application, I was ready to plunge into the second envelope the FDA had sent. It contained microfiches of what the FDA had compiled about the drug during its two years on the market in the United States—the pharmacovigilance on Floxin. The data came from the national Adverse Reactions Reporting System, which was CDER’s early warning system for problems not detected during preapproval testing. When a patient experiences a suspected adverse reaction to a drug, the doctor is supposed to contact the drug company and the FDA with this information in what is called a “spontaneous report.” This is to be filed even if there is no way to prove a causation. The reports then sit in the FDA’s database until someone at the agency gets interested in a particular drug and decides to look for trends among the amassed data.
After I made fuzzy copies of all the microfiches, I had more than 150 pages of printouts, each with eight to ten cases and a warning emblazoned across the top: “Cause-effect relationship between each drug and reaction cannot be established with certainty in all cases.” Each entry listed all the drugs the patient had been taking and the nature of the adverse reaction to Floxin. After that, there were spaces for the dose, route and duration of the Floxin therapy, as well as the patient’s age and sex, and then the outcome of the case. There was also a space for whether the report was domestic or foreign.
What jumped out at me before I could even begin looking for cases like Diane’s was how many of the spaces had been left blank. Most of the reports were ridiculously incomplete. Many didn’t include the age or sex of the patient. On most, the “patient outcome” was “unknown” or “none.” I had to wonder: How seriously were doctors taking the need to report adverse reactions? And how hard were the drug companies or the FDA trying to follow up on the reports, to make sure the database was as complete as possible?
I also noticed that while the domestic reports concerned reactions ranging from rashes and urine retention to sudden death, reactions in the smattering of foreign reports were, for the most part, nothing but deaths. Virtually no other reactions had been reported. This confirmed one of the things both Dave Flockhart and Ana Szarfman had told me: that while adverse reaction reporting was terrible in the United States, it was even worse in most of Europe, Asia and South America.
I was also interested in one reaction that showed up a number of times but hadn’t really occurred to me as being an ADR. The problem: “no drug effect.” Well, I thought, why not? Why shouldn’t it be reported when a drug didn’t work? We bother to maintain records of who writes prescriptions, who fills them and who takes them. We supposedly monitor when drugs harm patients. Doesn’t it make sense to keep track of how well drugs do or don’t perform?
The case reports on Floxin—or Tarivid, as it is called in Europe and Asia—were raw, but some trends did pop out. The adverse effects reported were overwhelmingly central nervous system reactions, the most common being insomnia. In each case report, the symptoms were listed alphabetically, and as I charted them, there were clear elevations at agitation, anxiety, confusion, convulsions, depression, dizziness, dream abnormality, hallucinations, headache … and on and on. A significant number of the CNS effects—significant to me, anyway—were reported in women, many of whom were in their thirties and taking Floxin in pill form, without other medications. This seemed to confirm what had been predicted in that chart from the clinical trials: women appeared to be having more CNS reactions.
I then separated out all the men, all the older women and anyone being treated with multiple medications, not because their reactions didn’t matter but because their medical situations appeared dissimilar to Diane’s. From the remaining cases, I counted sixteen reports of convulsions, twenty-five reports of hallucinations and twenty-six reports of other central nervous system effects: everything from depression and agitation to two cases of psychosis. Among this select group of younger women, there were three reports of deafness and six deaths, including one sudden death and one suicide the doctor associated with an “allergic” reaction to the drug. I even found Stacy Phillips’s case. Her doctor reported she had had a convulsion after taking two 300-milligram Floxin tablets.
I called Diane from the library to let her know that now it wasn’t just she and Stacy who had been Floxed.
The Floxin numbers were creepy but, of course, completely out of context—until I opened the third envelope the FDA had sent and looked at the reports on Omniflox, the quinolone that had been withdrawn. One chart, prepared by the FDA, compared adverse reaction reports for quinolones during their first 120 days on the market. Floxin accounted for almost as many reports as Cipro and Noroxin combined. But Omniflox generated nearly three times as many reports as Floxin—in fact, twice as many serious reports as all the other quinolones combined.
The chart convinced me that Omniflox had many more reported safety problems than Floxin. But it also reminded me that I had become interested in quinolones at what must be a very delicate time. Omniflox had been withdrawn only five months before Diane’s drug reaction. Quinolone manufacturers must have been waiting for the other shoe to drop. The medical product liability lawyers were also closely monitoring the quinolone situation. In fact, they were about to descend on Acapulco to discuss strategies for a number of drugs and sharpen their teeth. A Philadelphia attorney who was planning to send associates to the meeting, Steve Sheller, promised to tell me all about it.
I had a long talk with Sheller, who had first introduced me to Stacy Phillips, after he got his report from Acapulco. He explained to me that the American Trial Lawyers Association (ATLA) sets up what are called litigation support groups so that lawyers around the country with individual product liability cases can share information and strategy. At the same time, he described the competition between the lawyers to become the “lead counsel” on a group of cases that might turn into something big, like a lucrative class action suit. Lead counsel directed the path of the lawsuits and usually got a bigger cut of the final payout.
The attorneys had a rapidly growing interest in medical product liability because they were right in the middle of the massive and controversial class action involving Dow-Corning silicone-gel breast implants, as well as a smaller group of suits concerning L-tryptophan, a natural sedative sold in health food stores. The breast implant suits were about the product itself; the L-tryptophan cases appeared to be focused on tainted batches from a single manufacturer. Sheller had a few clients in each suit.
The silicone-gel breast implant cases had been jump-started by the stunning decision of the FDA commissioner, Dr. David Kessler, to temporarily halt their use the previous winter, in January 1992. The size of the possible settlements in the cases had sent lawyers scrambling to make FOI requests from the FDA. Suddenly, the product liability bar was taking drug and medical device cases more seriously. At the same time, plaintiffs’ lawyers were uncovering potentially damaging information during the “discovery” phase in these cases. During discovery in drug and device cases, lawyers make endless requests for company documents, some of which have never been seen by the FDA; they also often take depositions of key people the FDA has never directly questioned. The agency was taking the by-products of civil litigation very seriously as a major investigative windfall. Several major FDA investigations had been initiated, or broken wide open, by documents from civil cases. It was Dow-Corning internal memos, leaked to the press in 1991 after being sealed by the court during civil suits, that had led directly to Kessler’s later ban on the implants. That same year, memos from a suit over the patent rights to the acne medicine Retin-A—which Johnson & Johnson was trying to have approved as the first prescription antiwrinkle cream—reinvigorated a stalled FDA/Justice Department probe of J&J’s promotional practices.
At the same time, Sheller explained, trial lawyers were terrified by a new federal ruling by the First Circuit Court of Appeals in Boston. A woman who had allegedly developed autoimmune disease from collagen injections had sued the manufacturer. The court had held that since the collagen was an FDA-approved medical device, the case should be thrown out because FDA approval protected the product from being considered faulty.
The concept of FDA approval as a blanket shield against liability claims had been upheld at the state court level before. In fact, when Sheller had contacted ATLA colleagues before the meeting to see what anyone knew about quinolones, he heard about a couple of Floxin cases in the South which had been thrown out for that very reason. But there had never been such a ruling in a federal court, where it could become a national precedent.
The Boston ruling covered only medical devices. But drugs and devices are both regulated by the FDA, and a new legal interpretation of the scope of an FDA approval could eventually affect pharmaceuticals as well. Apparently, drug companies had been lobbying for years for just such a protection—which would put lawyers like Steve Sheller out of business. It would also stop the discovery process on many cases before it ever got started. “This decision,” Sheller lamented, “says that if the FDA approved a product, even if the approval was based on false information, the consumer has no rights to do anything.”
Sheller reported that there was little interest in Floxin cases at the meeting, because everyone was so absorbed by the Omniflox situation. Since Omniflox had been withdrawn from the market, it would be much easier to prove that it was a faulty product. And since it had been on the market for only four months, lawyers assumed they would be able to show there was a specific period when the drug company knew about the problems but continued to sell the product anyway. That would play well with a jury.
As far as the lawyers were concerned, Floxin was still FDA-approved, and only a handful of possible plaintiffs had been heard from. It was too early to know how actionable the seizures and neurotoxicities would be. The “discovery” process could easily run a law firm $50,000 to $100,000 in time, travel expenses and stenography fees. Unless a number of firms and clients wanted to share that cost, it meant that without a mid-six-figure settlement, the lawyers might not even break even.
In the meantime, however, Sheller wanted to know what we were doing about making sure Diane’s case was brought to the attention of the FDA and the drug company. Regardless of whether we wanted to initiate legal proceedings—and we weren’t sure that we wanted to—had we made a spontaneous report yet to the FDA? That’s why the system doesn’t work, he said. Nobody bothers to report adverse reactions.
So I called around to Diane’s doctors to see if any of them had reported the reaction. None had. I spoke about this with Dr. Amy Finkelstein, the fine internist to whom Diane had been sent after the emergency room visit and who had since become our regular doctor. A towering, handsome pterodactyl of a woman with a raucous wit, Finkelstein volunteered to contact the drug company. I contacted them, too. She called the medical division, I called the public relations division. The responses couldn’t have been more different.
Dr. Finkelstein called Ortho to report the incident and ask if the company had had any previous experience with such reactions to Floxin. She reported back that an Ortho physician had seemed genuinely concerned about Diane and had revealed that the company knew of three indexed cases of neurochemical changes in the same medical ballpark as Diane’s reportedly brought on by Floxin. The doctor also said that the company was either contemplating, or was about to implement, wording changes in the Floxin label to reflect those reports.
Armed with this information, I called the company’s publicist to see if she could help me find someone at Ortho who could discuss the adverse effects of Floxin and could address fears about quinolones in general in the aftermath of the Omniflox withdrawal. This seemed like a contentious but fair request: these fears weren’t just mine, but were beginning to show up in the pharmacology literature Dr. Flockhart and I were swapping. And much of the Floxin press material describes the drug’s side effects as “generally mild and comparable to other quinolones.” I also wanted to speak to someone about how the drug was being marketed, and I wanted to verify the coming change in the labeling.
I received instead a carefully prepared statement from the company. It said: “It is important to note that while Floxin is from the same class of drugs, quinolones, it has a different chemical structure [from Omniflox] … a completely different safety profile and has been proven worldwide as a safe and effective drug.… In the United States alone, ofloxacin has been on the market for almost two years and millions of prescriptions have been written since that time.
“We’re not aware of any labeling changes. Floxin continues to be a marketed product and the drug is doing very well.”
It turned out that both the company doctor and the PR statement were wrong. The Floxin label had already been significantly altered. About a month after she told me there was no new label, the publicist sent me a copy of the new label—which had been approved by the FDA two months before I ever asked for it. The paragraph about central nervous system effects included new warnings about “agitation, nervousness/anxiety … paranoia and depression, nightmares, insomnia and rarely suicidal thoughts or acts. These reactions may occur following the first dose.” The company also now warned that ofloxacin “should be used with caution in patients with a known or suspected CNS disorder that may predispose to seizures or lower the seizure threshold.” There were also many additional adverse reactions, and seizures had been added to the list of hypersensitivity reactions. Of special interest to Stacy Phillips was a brand-new warning of an interaction with nonsteroidal anti-inflammatory drugs (NSAIDs), which “may increase the risk of CNS stimulation and convulsive seizures.” That was certainly something Stacy’s physician might have liked to know before prescribing her Floxin and the NSAID Motrin.
When I inquired what the company was doing to make sure prescribing doctors knew about all this new language, I was asked to submit my question in writing. Ortho’s written response was, “We do not believe that the changes made to the package insert reflect any basic change in the well-established overall safety profile.” In short, they hadn’t done anything.
The company pointed out, once again, that Floxin “has been used by over 100 million people worldwide and has been shown to be safe and effective when used in accordance with product labeling recommendations. Like all prescription drugs it has a side effect profile which the doctor must weigh in treating the patient. As described in the labeling, serious side effects with ofloxacin are rare when used according to product labeling recommendations.” The company said it had “evaluated in a number of volunteer studies the potential for CNS toxicity. In all of these trials, the results of which were submitted and reviewed by the FDA, ofloxacin displayed no CNS toxicity [my emphasis].”
By the time the new label showed up, I had decided to write a magazine article about Diane’s situation and the adverse reaction we were having to her adverse drug reaction. As I was working on the story, I received word that Dr. Sidney Wolfe, the medical consumer advocate at Ralph Nader’s watchdog group Public Citizen, was about to unleash an updated version of his controversial best-seller Worst Pills, Best Pills: The Older Adult’s Guide to Avoiding Drug-Induced Death or Illness.
Even though it was quite impossible to get him on the phone, Wolfe did finally agree to fax me an advance galley of what his book would have to say about quinolones. He called them “one of the biggest selling and most overprescribed new classes of drugs in the United States.” He noted that “no quinolone is the drug of choice for … bronchitis or … the commonest … community-acquired pneumonia. With rare exceptions, quinolones are not the drug of choice for other infections.” The seven-day course of quinolone treatment, however, “can cost between seven and twenty-one times more than equally effective (for most infections) treatment with other drugs.”
While writing my story, I became increasingly aware that another tall guy with salt-and-pepper hair was getting very interested in the whole legal drug business. Former Arkansas governor Bill Clinton had developed a passion for bashing drug companies during his campaign for president of the United States. While he probably knew that his hero, John F. Kennedy, had railed against high drug prices during his own presidential campaign, Clinton had inherited the issue more directly from one of his major political mentors, David Pryor, the Democratic U.S. senator from Arkansas. Pryor had become one of the most outspoken critics of the pharmaceutical industry after being given the platform of the chairmanship of the Senate Special Committee on Aging in 1989. He had breathed life into the issue of drug company practices by shifting the focus away from the scientific and regulatory questions I was pursuing, and going directly to dollars. Seen through the prism of price, pharmaceuticals were an industry almost anyone could hate. But it was especially loathsome to senior citizens with fixed incomes, bulging medicine containers and rising drug bills. These were the people who were truly being “pillaged.”
Senator Edward Kennedy had long been holding hearings on a variety of pharmaceutical issues, and in the U.S. House of Representatives, Democratic congressmen Henry Waxman of California and John Dingell of Michigan had also been harsh critics of pharmaceutical pricing. The American Association of Retired Persons (AARP) had been strong supporters of the attack on the drug companies, along with a strident nonprofit group called Families USA, which sponsored studies of drug prices. They were often conducted by Steve Schondelmeyer, a crusading professor of pharmaceutical economics at the University of Minnesota. Along with Dr. Sidney Wolfe at Public Citizen, this group had, for some time, represented the front line in the small but tenacious consumerist assault on the omnipotence of drug companies. But it was Pryor’s interest that had influenced Clinton most strongly, and a handful of Pryor’s legal-drug point people were now working at the White House as advisers to the president.
Taking the drug companies to task appeared to be the centerpiece of Clinton’s monumental health care reform strategy. To that end, Clinton was joining Pryor, Waxman and Dingell in turning up the rhetoric, providing the media with powerful “did you know” factoids that set the whole legal-drug culture in a new, darker light. Everywhere I looked, I began seeing stories about high drug prices and drug company profiteering. One of the best was a series in my hometown Philadelphia Inquirer.
The drug business was, and had been for some time, the single most profitable legal manufacturing industry in the world. Every year when Fortune magazine did its Fortune 500 issue, pharmaceuticals ranked at or near the top in return on revenues, return on assets, return on equity and total return to investors. In worldwide sales, legal drugs were a $250 billion business. But the $55 billion U.S. pharmaceutical market was the one where prices had most spiraled out of control. American drug prices outpaced not only the general inflation rate but even the inflation rate for soaring medical costs.
In the previous five years, the consumer price index had gone up 21 percent. But the price of the top twenty drugs bought by elderly Americans had risen almost 80 percent. Americans paid over 50 percent more than Europeans for identical drugs, and 32 percent more than Canadians. U.S. prescription drug prices seemed almost impervious to the market forces affecting most other products Americans bought; because doctors and patients were so clueless about what drugs cost, there had been little pressure on pharmaceutical companies to compete on price. It was not uncommon for a new me-too drug to enter the market at a price as high as, or even higher than, the drug it had been developed to compete against.
The companies blamed their high prices on unreasonable regulation by the FDA and, more important, on the massive research and development costs associated with bringing a new drug to market. The precise amount of the R&D costs was a source of debate. Individual compounds generally cost less than $50,000 to develop: some cost as little as $3,000. But when the costs of all the compounds that failed along the way to a marketed drug were added up, the average was $194 million per drug, according to the Office of Technology Assessment (OTA), a bipartisan congressional agency. Industry preferred the estimate of $231 million per drug arrived at by researchers at Tufts.
Whatever the figure, drug industry critics were quick to point out that the companies spent much less money creating new drugs than they spent creating new demand for drugs. As an industry, pharmaceutical manufacturers spent $10 billion a year on marketing, but only $9 billion a year on research and development. The $10 billion in marketing money was being spent largely to sway the opinions of America’s 550,000 prescribing doctors. FDA chairman David Kessler joined the growing pharmaceutical feeding frenzy by suggesting that large promotion budgets were leading to inappropriate prescribing. “People will be hurt,” he told the Times, because the “promotion is designed to create a market for a product … [but] not necessarily the market that will benefit from the drug.”
Industry critics also pointed out that while the companies were spending $9 billion on R&D, more than half the medicines they were developing were me-too drugs. This assessment was based on the FDA’s internal system for determining whether a new drug provided “significant improvement” over what was already available, “modest improvement” or “no improvement” at all; essentially, it was an official government “me-too rating.” When the high percentage of me-too drugs became a public embarrassment and a potential liability on Wall Street, the companies took immediate action. They forced the FDA to abolish the me-too ratings.
The articles started hitting fast and furious in late February 1993, when eye-opening reports were released by Pryor’s Senate subcommittee and then by Representatives Waxman and Dingell, who had commissioned a devastating study by the Office of Technology Assessment. The OTA report said, “The market for prescription drugs is broken” and went so far as to calculate how much “excess profit” the drug companies were making, even when compared with other high-tech, science-driven, risk-heavy businesses. OTA put the excess profit margin at a staggering $2 billion a year.
Clinton wanted companies to voluntarily lower their prices. He was threatening mandatory price controls—which most other countries had already enacted (chiefly because they had national health insurance, so their governments were actually paying for the drugs). He was also threatening to withdraw a massive tax loophole that was saving pharmaceutical companies over a billion dollars a year. The Section 936 tax credit was available to any U.S. company that brought jobs to Puerto Rico by manufacturing its products there. But it was the pharmaceutical companies that were benefiting most from the loophole. They had taken advantage of it more than all other industries combined, and they had the most to lose.
Clinton’s threats were the same ones that Pryor had been making for years. But as the senator had made very clear in his committee’s annual reports about pharmaceutical pricing, the drug companies seemed neither scared of nor particularly sympathetic toward those who couldn’t afford the medications they needed.
Occasionally I would wonder if I was being too tough on Ortho, or its parent Johnson & Johnson. When that happened, I had only to listen to the president call the pharmaceutical industry “price-gougers” again. Then I knew my cause was just, or at least popular.
Diane was still not feeling much better, although she was beginning to, as doctors say, “climatize” to her symptoms. And as the holidays came, we got a chance to see if people were climatizing to her. I was curious to discover how our friends and relatives had processed the few tidbits of information we had let out about her condition. When you’re cagey about someone’s health problems, you never know how people are going to fill in the blanks.
We went to an annual holiday party that most of our friends in Philadelphia can be counted on to attend. This one offered a smorgasbord of reactions to Diane’s drug reaction, most of them more reassuring than I had expected. I thought some people might find what had happened hard to believe. There were times when I still found it hard to believe. Diane looked fine, acted fine; in fact, since many of her symptoms were ones that only an attentive husband would notice, she would need to go out of her way to explain how she wasn’t fine. But she wasn’t about to do that at a party.
Actually, she didn’t need to. Many of the guests had their own drug reaction tales to share. Diane was especially moved by a woman who told her a harrowing story about her preteen daughter and an antibiotic. After taking several pills, the young girl, who had never had any emotional problems, told her mother she was suddenly overwhelmed with the feeling that she wanted to kill herself. The frightened mother was even more shocked when her daughter stopped taking the drug, and the feeling simply vanished.
The onset and disappearance of that symptom, called suicidal ideation, was exactly the kind of neuropsychiatric drug effect we were learning about. Diane had experienced it acutely in the weeks right after the drug reaction—this unprecedented feeling that she could just jump out the window. She had quickly recognized it as a symptom and was relieved when medication alleviated it. But she wasn’t sure how to explain this phenomenon to others. Hearing a suburban mom discuss her daughter’s drug-induced suicidal ideation with such clear-eyed compassion—and not a hint of “it’s in her head”—made Diane feel more reassured about how it would go when she explained it to her own mom.
Unfortunately, medications hadn’t been quite so successful in addressing her other problems. The drugs, which the doctors had initially hoped would reverse all her symptoms in three weeks, had yet to do their jobs after three months. We were getting more worried.
Diane was primarily being treated by Dr. Michael Miller, a bright young biological psychiatrist and psychotherapist with an arch manner and impeccable taste in suits and office art. Initially, he had put her on the antianxiety agent Klonopin (clonazepam) to quash the insomnia and lithium to bring her mood down from the thought-racing highs of what he concurred with the internist and the neurologist was “acute hypomania.” The Klonopin had made her too sleepy, zombie-sleepy, so she was quickly switched to a less powerful member of the same class of drugs, called benzodiazepines. The new drug was Ativan (lorazepam), which we called “lorz,” and Diane was encouraged to regulate—or “titrate”—the dose herself. She jokingly referred to this prescription as “lorz, to taste.” Since the normal dosage is between 2 and 6 milligrams a day, she was given 1-milligram pills. But sometimes even one of them was too strong. This is apparently a common problem for petite women, and Diane is tiny, just over five feet tall and just over a hundred pounds. Since the pills couldn’t be cut in half, the doctor rewrote the prescription for the half-milligram tablet, which she called a “mezzo-lorz.”
Lithium was the more disagreeable of her two drugs. Besides being used preventively—or “prophylactically”—as the standard treatment for manic-depressive illness (or what is now often called bipolar disorder), it is also used to counter temporary, severe mood symptoms brought on by drug reactions and other illnesses. I quickly came to understand that the medication carries a stigma: it’s one of those drugs people don’t want to take, simply because they hate the characterization of being “on lithium,” and they don’t appreciate their medication being the punch line of every mental health joke. All current Prozac jokes were originally lithium jokes.
Besides the stigma, lithium is also a very high-maintenance medicine. Diane was required to drink more water every day than she was accustomed to showering with, and she had to have regular blood tests to check her lithium levels. The reason was that lithium is a medication with a very fine line between an effective dose and a toxic one.
All drugs have what is called a therapeutic range—which is the amount you want to have in your bloodstream for maximum effectiveness. For all drugs, the dose itself is less important than how your body utilizes the dose. Doctors can roughly assess how a drug is working by asking how you feel, but it can be accurately determined only by measuring your blood levels. If the dose doesn’t bring your blood level into the therapeutic range, the drug may not be doing anything. If it goes above the therapeutic range, you are technically overdosing and inviting some of the drug’s worst side effects.
If it were practical, ideally doctors would test your blood levels for all the drugs you take; it’s the only way to know whether you’re getting the proper dose. But it is not practical, at least not at present, unless you’re already in the hospital and having blood drawn every five minutes anyway. Outpatients are tested only for medications that, like lithium, have a very narrow therapeutic range and a high risk of toxicity. Even with those drugs, however, doctors aren’t always religious about the testing. Why? Perhaps they assume that the patient can be relied upon to tell them if the drug is making them feel sick. Or perhaps they trust that, as healers, they will Hippocratically sense a brewing medication problem—especially in the case of psychiatrists, when they’re seeing a patient weekly. They are probably also in a catch-22 situation with patients and insurers who could criticize them for ordering too many tests.
Why was I speculating about the reasons a doctor might not order yet another blood lithium-level test? Because the package insert said Diane was supposed to have two tests a week during the “acute phase” and at least one every two months after that. Regardless of my unfitness to determine when the “acute phase” ended, I knew she was getting fewer tests than recommended. Diane knew it, too, and had asked Dr. Miller about it. She was just more accepting than I was of his reassurance that he had treated many patients with lithium and had a more realistic idea than the PDR of how much testing was needed.
Diane and I almost never quarrel, but we did find ourselves arguing several times about her lithium-level testing. Dr. Miller is a fine psychiatrist whom I otherwise like very much and generally credit with saving Diane’s life. But I thought he was being too cavalier about these tests. Diane felt I was being paranoid and trying to meddle in her care. It was hard for me to make her understand the way I felt. I continued to believe that if I had been more paranoid and meddlesome before she took the Floxin, this whole situation might have been avoided.
It was also obvious to me that Diane was having more problems with her current medications than she was letting on. The morning after my birthday—for which we have an annual pizza and bowling party in New Jersey—she admitted that she had had a terrifying experience while preparing for the festivities. After walking over to the bakery to pick up my cake, she got lost while carrying it home. She described standing at a corner a half-block away from our house and not being able to figure out where she was. It was heartbreaking.
We finally got our first relief in months by taking a long romantic weekend of tan lines and shellfish in St. Thomas. Diane started out the trip feeling as well as she had in months (and I have pictures of her in a bikini to prove it). But after five days in the tropical hot sun, she wasn’t able to keep up with the amount of water she needed to drink with the lithium. By the end of the trip, she was dehydrated and a little neurotoxic.
Ultimately, the lithium levels became a moot point, because Diane and her doctor agreed that the drug hadn’t done much to arrest her symptoms anyway. They decided to try another pharmacological approach, replacing lithium with Depakote (valproic acid), an antiseizure medication in a different family from the benzodiazepines. Depakote was being used experimentally as a mood stabilizer.
By this time, Diane’s health problems had lasted too long to be viewed as an emergency by anyone but us. Unlike the first few weeks after she was Floxed, when we encouraged family and friends to believe it was a situation with a short-term “cure,” we were now at a different place. When people asked, “How are you?” we were no longer sure they really wanted to know the answer—nor were we sure we wanted to give it to them. I found myself deflecting questions I had wanted to answer just weeks before, consciously rerouting conversations, especially health-related conversations.
And unfortunately, all of a sudden, there were a few other places for those conversations to be detoured.
In mid-January of 1993, my uncle Michael, my father’s younger brother, died in Harrisburg. He was only forty-four years old, and left a teenage son and an ex-wife who didn’t quite know what had hit them.
I was close to my uncle when I was a kid, and he still lived with my Nana and Pop-pop. I spent many Friday nights sitting in his bedroom being introduced to Bob Dylan and Stevie Wonder on his little record player. Later, when he went off to college, he inadvertently continued my education through the records he left behind. The one that made the greatest impact on me was called Undercurrent, by pianist Bill Evans and guitarist Jim Hall, which featured a mesmerizingly grim cover shot of a dead woman floating beneath the surface of a body of water. The music inside was similarly dark—complex, haunted versions of normally upbeat jazz standards—and I came to associate it with him. Michael was thin and handsome and always wore great clothes. But even though he had been a brilliant law student, he treaded water in the real world, always fighting against some kind of undercurrent. Eventually he was towed down.
As I carried Michael’s casket to his grave, I couldn’t pretend that I didn’t lay some of the blame for his death on the dangers of legal drugs. He had migraines, for which he took too many painkillers. Several years before his death, Michael had had an operation on his stomach, and it was bleeding from his stomach that eventually killed him. I couldn’t help but wonder if the damage hadn’t been caused by some of his prescription drugs—the pain pills he took can gnaw away at the stomach lining.
Not long after Michael’s death, my Pop-pop was admitted to the hospital for his emphysema and found out that he also had inoperable cancer. The diagnosis was irrelevant, because mostly what he had was a broken heart—after spending so long preparing for the death of his wife, he had instead lost his youngest son. Knowing Pop-pop would not last much longer, I came to visit him often. My first book was due to be published very soon, and I hounded my publishers to give me just one copy so he could see it before he died.
I hopped on the train to Harrisburg the day the book arrived so I could show it to him that very afternoon. I kissed him good-bye and felt the tickle of his mustache on my cheek for the last time. He died a few days later, and I was back home as a pallbearer for the second time in six weeks.
Then Diane’s father was rushed to the hospital with a perilously high fever. Stoic that he is, Ed had ignored a bladder infection until it nearly developed into septicemia, a life-threatening condition. The doctor at the hospital wanted to put him on, of all things, Floxin. Ed refused it, so they gave him another strong antibiotic instead.
While his reaction was understandable, it did concern me. Certainly anyone with a daughter who had the kind of reaction to Floxin that Diane did might be at greater genetic risk for similar problems. But as I was beginning to understand more fully, it is not only risk that matters. All drugs have risks. It’s a question of whether the potential benefit is worth the risk and whether the patient understands that risk—which depends on whether the doctor knows enough about the drug to really explain the risk.
From a risk/benefit perspective, Floxin was probably not the drug of choice for what had been wrong with Diane. Not only was it “too big a gun,” but it was never clear that the specific bacteria she was infected with would have responded to Floxin. (Doctors are encouraged right on the Floxin label to administer the drug before results of lab tests are back.) But Floxin may very well have been the drug of choice for my father-in-law’s potentially lethal infection, and he waved it off only because of what he knew had happened to Diane.
I felt much better, as did he, when the replacement drug killed the bacteria and brought his temperature down to double figures.
My article was about to come out in the April 1993 issue of Philadelphia magazine. The piece offered a fairly in-depth view of what had happened to Diane and what we had learned about the system that was supposed to have prevented it. I concluded with a discussion of “black boxes.”
During my conversations with FDA doctors and Dave Flockhart, I would often hear about black boxes: big serious warnings placed at the very top of the package insert for certain drugs and then highlighted with, literally, a thick black box around them. A doctor in the FDA antibiotics division told me I could learn a lot from the story of the aminoglycosides—a class of injectable antibiotics, including gentamicin, that had come on the market in the late 1950s and were hailed as a new way of treating many of the same bacterial infections that quinolones combat. After the drugs had been on the market for some time, it was found that, in some patients, they caused both kidney damage and deafness.
They now carry a black box warning for both these adverse drug reactions. A black box doesn’t mean the medication should be avoided: in fact, gentamicin is still commonly used in hospitals. It simply means that the drug can be dangerous and patients taking it should be closely monitored. Drug companies apparently hate black boxes, especially for products like broad-spectrum antibiotics, which they want physicians to feel comfortable throwing at anything. A black box calls attention to a drug’s risk/benefit ratio. Doctors are always supposed to be taking that ratio into consideration, but too often they don’t. Instead, they prescribe on automatic pilot.
A black box also alters the kind of print advertising a drug company can do. For most drugs, companies have a choice between regular ads—which, since they recommend a course of treatment, must be more than one page in order to include prescribing information from the package insert—and “reminder” ads or “promotional labelings,” which simply call attention to a brand name and can be a page or less. A pen or a notepad that a drug company gives out is, basically, a reminder ad and is regulated as such by the FDA’s Division of Drug Marketing, Advertising and Communications (“Dee Dee Mac”). Technically, the agency can act only if a false or misleading ad is published. While companies aren’t required to submit their promotional materials ahead of time, DDMAC is usually asked to screen ads and press material before publication or distribution to see if it will consider them problematic.
Reminder ads can be cheaper (because they require less space) and can often be more “creative” because they provoke fewer arguments with the FDA about whether the scientific claims are accurate. But if a drug has a black box, its manufacturer cannot use reminder ads.
In my article, I called for a black box warning for CNS side effects to Floxin. I also urged the FDA to call a meeting of its Anti-Infectives Advisory Committee to put the drug’s neurotoxicity on trial.
The piece included as much of Diane’s story as she and I felt was appropriate. The debates we had over the inclusion of personal details were intense. Since we had both written about each other before, we had already navigated some of this territory, but the stakes seemed much higher this time. And it’s always easier to tell a personal story when you know how it ends, which we didn’t. Diane and her doctor were already talking about getting her off the Depakote, which she had been taking for only two months. The drug gave her terrible stomach pains and was making her already fine hair fall out in clumps.
Still, overall she was functioning better—no less symptomatic, but more accustomed to being symptomatic. Once Diane admits something is real or true—which can take a long time—she can adapt to almost anything. She was growing weary, however, of being sick and, worse to her, of being perceived as sick. She was still waiting for that combination of pills that would make her feel not just “better” but like someone she recognized as herself. So was I. Even after all the bitter pills we had swallowed, we were still awaiting the wonder drug.
I was proud that Diane had the courage to let me use her story to illustrate an important public health problem. But we both had serious doubts that anyone besides our family members would pay much attention to such an esoteric piece of pharmaconfessional journalism.
And then the phone started ringing.