It started out with just one pill. My wife's gynecologist had given her samples of a new antibiotic to treat a minor urinary tract infection. The doctor told her to take this wonder drug twice a day for three days. When I left for work the next morning, I said goodbye to Diane as she took the first pale yellow, oval tablet with breakfast. Six hours later I was bringing her, delirious, to the emergency room. Her life hasn't been the same since.

In the first hours after Diane took the pill, her left arm went numb and her vision became cloudy. She got lost in her home office, and when she found her desk, she couldn't figure out how to turn off the computer she writes on every day. When she went to lie down, she started shaking uncontrollably and then saw white. She felt like she was dying.

I tend to call home several times a day to touch base with my wife. Luckily, Diane was able to pick up the phone and tell me a few of her symptoms. I called her gynecologist, who told me to take her to the hospital. When the cab got me home from the office, I found Diane lost in her closet. She stammered that she had wanted to change clothes to go to the hospital, but couldn't find her white shirt. I looked down and saw that it was an inch from her hand.

Married people can only afford to panic one at a time, so I pretended I was not scared as I helped her on with her shirt and took her to the hospital. As she spoke -- haltingly, elliptically -- to the emergency room doctors, more symptoms emerged. Her jaw was terribly sore from clenching against what we assumed was a seizure, and her pupils were fixed and dilated. She felt like something was "melting" behind her eyes.

After several hours of neurological exams, the word came back from the poison control center that all Diane's symptoms had been previously reported as reactions to the antibiotic she took. The drug is called Floxin. She had, as we now say, been "Floxed."

The doctors explained that Floxin had side effects different from those caused by other types of antibiotics prescribed for urinary tract infections. When it caused problems, Floxin was more likely to disturb the central nervous system, where adverse reactions range from headaches, dizziness, confusion and sleep disorders to seizures, hallucinations and major psychiatric episodes.

We were then told what I now realize is the prevailing wisdom about all adverse drug reactions -- that the effects would subside when the medication left her system. But Diane's symptoms did not disappear as promised. Some waned, but new ones developed. Besides the "melting" and the fixed pupils, she had aggressive insomnia, visual distortions and aphasia: She would get halfway through a sentence and not be able to get the rest of the words out.

Over the next two weeks, she endured a CAT scan, an MRI, a spinal tap and an EEG, besides all the blood work, to rule out any possible explanation for her symptoms other than the reaction to the drug -- which was supposed to be long gone from her system.

When all the tests came back negative, the doctors concluded that she was still experiencing "acute hypomania." The drug reaction had triggered some genetic predisposition. And since her body hadn't been able to put on the brakes naturally, she would need a combination of heavy-duty drugs, with their own possible side effects, to do it.

But at least that urinary tract infection had cleared up.

IT HAS BEEN well over a year since Diane got Floxed. In that time, we've learned more than we wanted to know about America's other drug problem: adverse reactions to prescription and over-the-counter medications.

While most people take most drugs without problems, no medication is risk-free. In the United States, more people die each year from adverse reactions to legal medications than succumb to all illegal drug use (which claims about 10,000 lives annually). ADRs -- the industry shorthand for adverse drug reactions -- are believed to affect 1 million to 2 million Americans a year. Many just get skin eruptions, but it has been estimated that from 3 to 11 percent of all U.S. hospitalizations are caused by ADRs. And about one-quarter of all inpatients have adverse reactions to the drugs they get in the hospital. There is little public awareness of the problem: It is estimated that fewer than 10 percent of America's drug reactions ever get reported. According to one study, only 1 percent of the most serious reactions get reported. Nobody knows what percentage get properly diagnosed and treated.

Before Diane's frightening experience, I had always thought of prescription drugs as pretty much idiot-proof. Your doctor tells you to take them, so you do, assuming that the worst that can happen is they won't work. It turns out the worst that can happen is that you drop dead. The next worst is that your body is permanently damaged. Less worse, but still not very good, is that you suffer for hours, days or weeks with something your doctor may or may not recognize as a drug reaction -- from one drug or an interaction. It may or may not go away by itself.

Idiot-proof? Well, maybe the idiot part was right. The proof is right on the package insert, which most people don't bother to read -- including, it turns out, far too many prescribing doctors. That's one of the things I learned while trying to find out if my wife is something more than just a statistically acceptable casualty of "friendly fire" in the war on disease.

The search for answers to her problems has taken me deep inside the Washington-based world of drug approval. It also led me to a Georgetown University pharmacologist with a bold proposal that some consider utopian -- but others believe is a crucial missing link in the Clinton health care plan.

AFTER DIANE'S condition was diagnosed, I started digging up everything I could on Floxin and the general subject of ADRs. I made a Freedom of Information request to the Food and Drug Administration for anything that had to do with the drug. I searched computerized newspaper databases for anything with a "flox" in it, and called for a Floxin press kit from the manufacturer, Ortho Pharmaceuticals.

I quickly found out that Floxin is one of a family of new antibiotics called "quinolones," all of which have a "flox" in their generic names (Floxin is ofloxacin). They are primarily prescribed for infections of the skin and the urinary and respiratory tracts, and for sexually transmitted diseases. The family includes Cipro (the largest seller), Noroxin, Penetrex and Maxaquin. But, in June of 1992, it tragically lost a relative: Omniflox, which was voluntarily withdrawn after only four months on the market because of reports to the FDA of severe adverse reactions, including three deaths. The news stories about the Omniflox fiasco -- which prompted a shareholders' suit and a slew of liability suits, many of them already settled, against Abbott Laboratories -- were both horrifying and calming to Diane and me. They begged the question of how the FDA approval process worked, but also reminded us that if the worst was over, it could have been a lot worse.

In the meantime, I bought a new Physicians' Desk Reference -- the compendium of paid drug company labeling information that Diane usually consulted before taking medicine. Floxin was too new to be included in our old dogeared PDR, which was why she had uncharacteristically taken the drug without knowing much about it: The samples, as is typically the case, were given to her without any instructions or warnings at all.

We looked up all her new drugs -- which she swallowed with considerably more courage than I think I could have mustered under the circumstances -- and then looked up Floxin. To our astonishment, a lot of what Diane had experienced was listed among the dozens of reactions that had been reported by "less than 1 percent" of patients out there. These reported reactions included seizures, hallucinations, cognitive changes, psychotic reactions and a manic reaction. As I reread the PDR listing, I searched for words that might have tipped us off that Diane would be one of the susceptible ones, something that could have alerted her doctor. Diane looked at the adverse reactions listings and found them strangely comforting. They meant that she wasn't alone, and that nobody could accuse her of making it all up, of being "a hysterical woman."

As I researched Floxin, I told my wife's story to anyone who would listen. Many responded with a tale about an unexpected drug reaction they or someone close to them had had. One person I talked to -- who had taken Floxin with no problem, like millions of other people, apparently -- mentioned Diane's story to a lawyer who couldn't believe what he was hearing. An attorney in his office had told him confidentially about an alarmingly similar experience.

I called her up. Caroline Reeves was a 31-year-old litigator with a newborn and a 3-year-old. She had taken two Floxin pills, for abdominal pain she and her doctor assumed was the return of a post-childbirth pelvic infection. She'd had a grand mal seizure in her sleep, hadn't recognized her kids when her husband finally nursed her back to consciousness, and even though her memory had returned, she hadn't been "right" since. She was especially angry about the drug reaction because she had asked her doctor for the antibiotic Keflex, which she had used successfully before, and he had persuaded her to try the stronger, and more expensive, Floxin. It also turned out that she didn't have the infection.

Reeves's cognitive and memory problems, insomnia and visual distortions (she also had the "melting") had her thinking about leaving law. She lived in constant fear of having another seizure, especially while driving. She was also worried about the stigma she might carry if she told other colleagues or clients about her condition. But she was more afraid of taking another pill, so she couldn't bring herself to try the anti-seizure medication that had been prescribed for her.

ALTHOUGH DIANE'S DOCTORS seemed to be doing a good job with her symptoms, only our pharmacists -- our new best friends -- shared our fascination with the drug reaction itself. So I tracked down a clinical pharmacologist, Dave Flockhart, at the Georgetown University School of Medicine. One of his areas of research is pharmacogenetics: trying to determine why certain types of people might have problems with drugs that are "safe" by FDA standards, so ADRs and drug interactions might be predicted and avoided. Only 10 percent of drug reactions are truly allergic. The rest are technically intolerances, overdoses or interactions, although the mechanisms aren't always clear. A genetic inability to properly metabolize a drug, for example, can technically cause an "overdose" from a normally safe amount; the same thing can happen if a doctor doesn't adjust dosages for body weight.

When it came to central nervous system reactions -- to Floxin or anything else -- Flockhart was, unfortunately, mostly telling me what wasn't known or knowable yet. But he did know that quinolones attach to a binding site in the brain that has been identified as having something to do with seizure thresholds. He explained that some people are unaware that they have a genetic predisposition to seizures or a "seizure focus," perhaps created by a head injury.

Flockhart was sympathetic to my complaints about how hard it was to get information about drug reactions, especially to a new drug. Whom do you call, what do you read, when the drug company offers nothing but package inserts and press releases, and most of the professional journal papers are based on studies done by the manufacturer to get the drug approved by the FDA? And since I was reading the same information doctors got, I had to wonder -- weren't they concerned about how little unbiased information there was about the drugs they prescribed?

Flockhart offered to check the quinolone literature himself and gave me the overall impression that pursuing this was important. "My suspicion," he said on the phone, "is that this drug is okay for most people and dangerous for some people. And we need to let the people in danger know that."

He explained that the pharmacology department at Georgetown is one of a handful of places that receive grants from the FDA to do research on post-marketing problems and drug interactions. It is research that drug companies do not go out of their way to sponsor because it might limit, rather than expand, use of their products. He also said that the chairman of his department, Raymond Woosley, was beginning to "get political" about the shortage of unbiased information about medications. Woosley, he said, would soon be unveiling a radical proposal to address some of the problems inherent in the FDA approval system, which offers a few checks along the way, but no real balances to the $10 billion a year that drug companies spend to promote and market some $55 billion in drugs via America's 550,000 prescribing physicians. Doctors get most of their information through FDA-approved printed material and word of mouth from company sales reps -- "detail people" -- fewer than 5 percent of whom have any formal training in pharmacy.

Flockhart suggested that if I ever finished avenging my wife, I should come down and talk to Woosley. Soon Flockhart and I were faxing things back and forth. He helped us focus our panic, and kept reminding us about the difference between emotion and science. The most important thing, he said, was to get hold of the clinical data. In a best-case scenario, my snooping around would cause the FDA to call an "advisory committee meeting" to look into the type of adverse reaction Diane had and make the drug company study who was at risk and why. Then the company would have to alert doctors as to which patients should avoid the drug unless it was the only choice in a life-threatening situation and the benefits outweighed the specific risks.

In fact, he wondered why a doctor was using something as strong as Floxin as a first line of defense against a common urinary tract infection anyway. "Most people would have used a smaller gun," he said. But he wasn't terribly surprised, and I later learned why. Antibiotics, among the most commonly used classes of drugs, are improperly prescribed 40 to 50 percent of the time. Either they aren't necessary at all, or the wrong drug, wrong dose or wrong duration is prescribed.

Apparently, everyone in medicine knows this but the patients.

OVER THE NEXT FEW MONTHS, I dove into floxinalia. I devoured the drug application documents, observing the process of FDA approval through the charts and graphs from study after boring study. I second-guessed every time a subject's adverse reaction in a trial was deemed "not drug related," because I could see how one such decision could change the perception of a drug's safety or efficacy. I had long phone conversations with several people in the anti-infectives division of the FDA, including the doctor who had been the medical officer overseeing the Floxin application.

Less productive were my phone communications with the PR department at Ortho Pharmaceuticals. For some reason, the company kept denying there was any new information on central nervous system (CNS) reactions to the drug, even though a new label for Floxin had been approved several months earlier with many new warnings about adverse reactions, including "paranoia and depression, nightmares, insomnia and rarely suicidal thoughts or acts." These reactions, the label added, "may occur following the first dose."

When I later asked what the company did to make sure prescribing doctors knew about all this new language, a spokesman responded, "We do not believe that the changes made to the package insert reflect any basic change in the well-established overall safety profile." The company pointed out 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 . . . Serious side effects with ofloxacin are rare when used according to product labeling recommendations." The company also 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."

Last April, I wrote an article for Philadelphia magazine chronicling Diane's painful experience and my search for some answers to the practical and philosophical questions raised by her drug reaction. Although I was proud that Diane had the courage to let me use her story to illustrate a problem, I doubted that anyone but our family members would pay much attention to this esoteric piece of pharmaconfessional journalism.

Instead, the phone started ringing off the hook; suddenly, I knew several dozen people who had had reactions to Floxin. Then the folks at "Good Morning America" decided to use the article for a women's health segment on adverse drug reactions; they had Diane and me on, along with Murray Lumpkin, then director of the division of anti-infectives of the FDA's Center for Drug Evaluation and Research. The show generated more calls and letters. Some were from obvious hypersensitivity junkies, but many were just from ordinary people who had become symptomatic after taking Floxin. Some didn't know there were such things as adverse drug reactions. Or their doctors had assured them that this drug couldn't cause such reactions. Some had even been encouraged by their doctors to keep taking the medication, even though the package insert clearly states that it should be discontinued immediately at the first sign of "restlessness, lightheadedness {or} confusion."

I spent hours on the phone with people. While I didn't mind explaining what little I knew, one idea kept occurring to me. If these people were so desperate for help and information that they were calling me, how good a job could the FDA, the drug companies and prescribing physicians be doing on adverse drug reactions?

IN JUNE, my suspicions were confirmed when FDA Commissioner David Kessler announced that he was scrapping the national Adverse Reactions Reporting System (ARRS) and creating a new mechanism. For years, ARRS had been the cornerstone of Phase IV -- the part of the FDA approval process that was supposed to identify problems a drug causes after it has been marketed to the real world. Only 3,000 to 4,000 people are studied in the three-phase FDA approval process; besides the obvious statistical limitations of the sample, women, children and the elderly are often grossly underrepresented among the subjects. Post-marketing reports from doctors and patients (which don't require proof of causation, just suspicion) are supposed to send early warning "signals" to the FDA and drug companies about unforeseen problems. It is then up to the FDA to interpret these signals and force labeling changes or even formal Phase IV studies.

In announcing his new MEDWatch system, Kessler was acknowledging what many people had known for some time -- especially since the embarrassing report on breast implants issued by the House Government Operations subcommittee on human resources in 1992, which said that only 5 percent of adverse reactions to drugs (and medical devices) get reported, and that the gravity of some of these reports is ignored.

The reason Kessler gave for the underreporting wasn't comforting either. In an article published in the Journal of the American Medical Association, he said that physicians, when confronted with an unexpected outcome of treatment, might not consider drug-induced disease -- because medical students receive limited training in clinical pharmacology and therapeutics. He cited a survey of U.S. medical schools that found that only 14 percent had required courses in "core skills and principles of therapeutic decision making."

The MEDWatch program offers a newly simplified form, a new public awareness campaign and a new toll-free number (800-FDA-1088) for reporting drug reactions. It could slowly change physicians' attitudes about ADRs. But "physicians are caught in a double bind," says Joe Graedon, a pharmacist who does a nationally syndicated column and radio show about prescription drugs with his physician-wife. Doctors are taught that one of their first obligations is to "do no harm," and yet, Graedon argues, every time they write a prescription they risk violating that fundamental tenet. "It creates an environment in which it is better for the physician to become defensive and try to deny there was a drug reaction.

"It's hard for a doctor to say, 'I hurt you. I didn't mean to, but you were that 1 in 50 or whatever who developed a bad reaction to the drug. And I feel terrible.' So, up come the defenses and the fear of litigation, and it becomes better to stonewall."

BY EARLY SUMMER, my wife and I decided to return to our regularly scheduled lives, already in progress. Diane went to a neuro-ophthalmologist to see if anything could be done about her visual distortions, the only symptom that had never responded to treatment. She finally decided that she would just have to get used to the visuals, just as she was getting accustomed to the effects of the anti-seizure medication. She started writing again, and we began to accept that whatever had happened was not completely reversible.

I went back to writing about things other than Floxin. But I did finally call Ray Woosley to ask for a copy of the proposal that Flockhart had mentioned. When it arrived, I did what I'm sure many people in much higher positions than mine have done: I threw it onto a pile of things I should read.

But, several months later, I was drawn back into the ADR fray. The FDA called a meeting of its anti-infectives advisory committee to discuss the exact topic I had raised in my article: CNS toxicity of quinolones, and how well their labels warned doctors and patients. I was invited to testify.

It was a fascinating day at the Silver Spring Holiday Inn. An FDA epidemiologist confirmed what I had noticed about the adverse reaction reports on Floxin -- that there seemed to be a significant number of seizures, major psychiatric episodes and sleep disturbances. But he was even more concerned about a newer quinolone, Maxaquin, which had a higher number of seizure reports. The meeting quickly degenerated into an existential debate between drug company docs and FDA committee members about what the adverse reaction numbers really meant. Ultimately, the committee voted to recommend moving the warnings about CNS reactions to the very top of the package inserts for all the quinolones. The specific language about seizures was put into boldface, and some other wordings were changed to -- in the words of one advisory committee member -- encourage doctors to "think twice . . . about the risk-benefit ratio" before casually prescribing quinolones, which some doctors have reported using to prevent travelers' diarrhea.

I had hoped for much more far-reaching changes. As the meeting ended and the participants shuffled out, I looked back over the notes I had taken on my laptop. One line jumped out at me. One drug company doctor had mentioned, matter-of-factly, "As you know, physicians will not even look at a package insert. If they do, it's for seconds . . ."

It was a good thing I had arranged to meet Dave Flockhart and Ray Woosley for beers after the meeting. I needed someone to tell me that the process of warning doctors and patients about reactions like Diane's wasn't as futile as this day in Silver Spring made it seem.

Flockhart, with whom I had a phone friendship but whom I had never met, turned out to be an animated, almost cartoonish character: short and balding with glasses, a bushy mustache and a mischievous grin. Woosley was also compact, but with dark hair, chiseled features and a more serious demeanor. They finished each other's sentences like a clinical pharmacology comedy team -- Flockhart excitedly yapping with his heavy Scottish accent, Woosley calmly explaining in his slight Kentucky drawl.

"We can tell you a story," Flockhart began, and they proceeded to tag team through their research into the antihistamine Seldane and its occasional association with heart arrhythmias. (Woosley's father died of heart disease, so it's not surprising that his department tends to be interested in heart-related drug problems.)

"Seldane was the 10th most prescribed drug in the country," Woosley cut in, "and it was just about to go over-the-counter," meaning that the drug was about to become available without prescription. "Buried in the experiments were some arrhythmias, patients blacking out . . . So, there was a case at Bethesda Naval Hospital where a woman was on ketoconazole for a fungal infection and took a friend's Seldane and started blacking out. They called in clin pharm, looked into it as a drug interaction, and ended up presenting it to Carl Peck {who was then director of the FDA's Center for Drug Evaluation and Research}. It was an arrhythmia. It turns out the company had a bunch of these cases and called them overdoses.

"The company stand is, 'We thought it was an overdose. The parent compound {in the drug} was found in high levels.' What happens, we now know, is, when you take it with this other drug, the parent compound builds up -- and that's what causes the trouble . . . Carl Peck was telling me that the company wanted to come up and talk to him about going OTC. Carl said, 'Why don't you come up, but the question is going to be if it stays on the market at all.' . . . It didn't go over-the-counter, and Seldane now carries this big boxed warning {about possible interactions with ketoconazole as well as erythromycin.} So you can see why there is a disincentive to the drug company to find out what was wrong with this drug."

Flockhart jumped in. "This is representative of the kind of research that never gets done . . . Very little is put into Phase IV kinds of questions -- what are the long-term side effects of drugs, how do they affect different genders, different races?"

Then Woosley: "I think we can predict ahead of time, with lots of drugs, who will have those bad reactions like your wife did. But a company would never fund the research, because they are afraid it would limit the way doctors would have to use their drug. They want a completely nonthinking doctor. Sometimes research like this means that a doctor can't prescribe a drug until he checks a blood test. And they don't want a drug that has that problem.

"There was a drug that caused lung scarring. One percent of the people who took it died. The company attitude was, 'We don't think it's such a big problem.' Deep down they think, 'If you study this side effect, it just magnifies it. We're going to go out and tell them it's not a problem. Hey, don't worry about this lung scarring bit!' And that's what their salesmen did."

"That's criminal behavior!" Flockhart said.

The point of the Ray and Dave show is to sell Woosley's plan to save America from drug ignorance. The plan, commonly called CERT, calls for the creation of "centers for education and research in therapeutics" -- in other words, a national chain of research centers in major universities that have existing clinical pharmacology departments. The centers would be funded by the federal government and would be equipped to take on a variety of FDA-assigned research, independent of drug company economics. The CERTs would also serve as education and information centers so physicians, pharmacists and patients could get economically unbiased facts about drugs -- including comparisons of price, efficacy and safety.

CERT is an updated version of a plan first suggested in the '70s. The idea must be in the air again, because a similar but somewhat less ambitious proposal was floated in the January issue of the New England Journal of Medicine. That proposal called for one national center just to independently evaluate the post-marketing data. But the concept has, so far, found no place in the Clinton health care reform plan -- even though, in its current form, the Clinton plan would likely exacerbate the problem by increasing prescription coverage for the elderly.

Woosley has encountered mixed responses to CERT. "I did a presentation to the drug forum at the National Academy of Sciences," he recalls. "The NIH said, 'Great idea, but we don't have money for that.' FDA said, 'Great idea, but we don't have the money for that, you'll need legislation.' The drug industry said, 'That's the worst idea we ever heard.' One guy said, 'Ray, you and I used to be friends. That was offensive.'

"That's because the drug companies honestly don't think they're doing anything wrong -- or that there's anything wrong with the system. There is. But it's going to take recognition that this country is spending billions on drugs -- and if the only source of information about those drugs is the marketing that drug companies do, then drugs aren't going to be used in the most cost-effective way. It's hard to even call this a problem. It's a built-in part of our economic system that creates a problem."

Woosley also thinks the CERT plan would address the malaise in the world of Washington drug doctors. "Morale is low at FDA," he says, "and I'll tell you why. There's a tremendous amount of expertise being crimped by only being able to review what drug companies give them. They have to take the labeling proposed by the company and work with that, and they can't dictate changes in therapy. They may know -- 'I reviewed three drugs, this one is better' -- but they can't tell doctors that. They're restricted by law not to recommend treatments. Somebody besides drug reps should be recommending treatments!"

Carl Peck, who left his FDA post in November to take an academic position in the Netherlands, is an enthusiastic supporter of Woosley's plan. "The CERT proposal is extraordinarily on the mark," he says. "The knowledge gap it would fill has significant health implications, as well as economic ones. With all the people hospitalized for adverse drug reactions, it will save many hundreds of thousands of dollars. But the research will also find out the kind of stuff that you, me, anyone would want to know before taking the risk of dying.

"CERT is inherently meritorious. The question is if it can be packaged and compete well with other issues in health care."

Sidney Wolfe, of Public Citizen's Health Research Group, says that CERT would "increase awareness of adverse drug reactions and lead to faster, safer care. Rather than depending on the current FDA system, it would be better to do what Woosley suggests. Will CERT ever get the support and funding to see reality? It depends on how smart Woosley is. But the proposal could prevent a massive amount of damage."

Woosley still needs to build legislative support. He got an encouraging letter from Arkansas Sen. David Pryor, who has been a thorn in the side of the drug industry from his position as chairman of the Special Committee on Aging -- and whose proposed Pharmaceutical Access and Prudent Purchasing Act of 1990 sought to address some of the same issues. But Pryor's drug point man, John Coster, cautions that CERT is "an interesting concept, but the senator isn't endorsing it as of yet. It's not a really hot issue right now for health reform because the macro issues are taking precedence. If CERT does spark national interest, however, it's likely to primarily be a Medicare drug benefit issue, because the elderly seem to suffer most of the adverse drug reactions."

Syndicated drug columnist Joe Graedon supports CERT but acknowledges that it will be an uphill battle. "The problem is there's no constituency for it," he says. "The public doesn't realize how critical it is. Drug companies would much rather control the research. The FDA, which should be advocating this, can barely keep up with what it's doing now."

While nobody would go on the record denouncing the plan, there was a deafening "no comment" from the offices of Illinois Sen. Paul Simon and California Rep. Henry Waxman, legislators heavily involved in health care issues, as well as from the Pharmaceutical Manufacturers Association, the National Institutes of Health and the congressional Office of Technology Assessment. The FDA's Kessler and Lumpkin, after requesting copies of the proposal, would not comment, although an FDA spokesman reemphasized the importance the agency places on reporting adverse drug reactions under the new MEDWatch program. Requests for comment from several drug companies went unanswered.

I KEEP WANTING this story to end, but it never does. Late last year I got a call from a producer of Oprah Winfrey's show. She wanted to do a program on adverse drug reactions because she had just had one -- to Floxin. Diane and I appeared on the program, along with several other people we had met through the original article, and since then I've gotten a steady stream of calls. Many of them are from people who had almost the same reaction Diane did, but weren't as lucky to have doctors who at least recognized a drug reaction and were willing to learn what they didn't know about how to treat it. I've talked to people whose spouses have lost their careers in the aftermath of drug reactions, people whose fathers attempted suicide because of depression that seemed to have been triggered by quinolones.

It breaks my heart to tell these people that there are no easy answers to their specific problems, or the general problem of adverse drug reactions. But most of them are just happy to hear that their pain isn't so esoteric -- that just because their doctors never heard of it doesn't mean they are making it up.

I just finished reading advance galleys of Adverse Reactions, a book being published this month. It's by Tom Maeder, a writer who has pieced together the history of one of the first "wonder drugs" done in by what he referred to as "promiscuous prescribing" and ignorance of adverse reactions; the drug, chloramphenicol, is a famous antibiotic from the '50s that was linked to deadly aplastic anemia in a certain number of cases.

Maeder's book begins in 1951 with the tale of a doctor from Southern California, Albe Watkins, who prescribed the drug to his 8-year-old son to prevent infection after surgery in his urinary tract. The boy later died of aplastic anemia, which was so rare that the doctor was certain it had been caused by the drug -- and he decided to take his case to the FDA. He drove cross-country, stopping in towns to call local doctors to see if they had any similar nightmare experiences with the drug. By the time he arrived in Washington, he had 45 cases of deaths he believed were linked to the drug. His work set off an FDA investigation that led to dramatic and extensively publicized label changes, which should have caused doctors to limit its use to only the most serious cases in which less risky drugs had already failed. But the drug continued to be used as a broad-spectrum antibiotic for decades, with the drug company successfully convincing doctors that the risks were minuscule.

Albe Watkins was a brave man but a man with nothing to lose; his son was dead. My wife is very much alive. I think it's time for me to give the attention I've been paying to Floxin back to her.

Stephen Fried's book Thing of Beauty: The Tragedy of Supermodel Gia will be published in paperback this June.