Right after I finished my PhD degree (studying, among other things, the mode of action of the antibiotic nalidixic acid), I started my medical clerkship on the wards of a venerable New York City hospital. This was the kind of hospital that in a movie would be called Fort Apache, and the physical exam of the patient would start with a search for guns and knives. There, with my very first patient, I learned important lessons about drug safety.
The patient was an elderly woman in her seventies who, frail and frightened, was breathing about 40 times a minute. Her chest was full of bubbly noises that were loud and sinister, and her EKG showed ST segment elevations that hit the top of the tracing. The diagnosis was easy—pulmonary edema in the setting of a myocardial infarction—and, with the magic of MOSTAMP (morphine, oxygen, sitting, tourniquets, aminophylline), her breathing calmed and she looked much more comfortable.
Just as we were about to move the woman from the treatment room to a bed on the ward, the resident came in and, after detecting a few bibasilar rales, told the intern to give the patient some intravenous digoxin. “Dig” was then a popular treatment for heart failure, although its use in the setting of an ischemic cardiac event was uncertain at best. The resident and intern argued, their voices rising and their faces turning red. Despite reference to high-class studies (evidence-based medicine was alive and well even in 1972), the intern capitulated and, exasperated, pushed a bolus of dig through the woman’s intravenous line.
The sequence of event remains vivid in my mind today. Within a few minutes of the dig administration, the heart monitor began to beep wildly and the screen exploded with a profusion of mean-looking extra beats, which were obviously premature ventricular contractions (PVCs). The PVCs then burst more frequently—machine-gun blasts of four, five, or six in a run—and degenerated into sustained ventricular tachycardia. The old lady looked as if she were in shock, and her eyes rolled ominously into her head.
Alarmed by impending disaster, we then gave the woman a bolus of lidocaine, which can settle down the sodium channels in her heart. We were happy when the monitor showed sinus rhythm. Our relief was short-lived, however, when, to our chagrin and dismay, the old lady had a grand mal seizure, and her small body flapped uncontrollably.
Out came the Valium to stop the seizure. The seizure ceased, but I am sure that you can figure out what happened next. The woman stopped breathing, total silence following a horrific gasp. Faced with another misery of our making, we had to intubate the woman and perform a complete resuscitation. Fortunately, we pulled the woman through—I was the chest thumper, and there were no cracked ribs—but the intern was now seething over the therapeutic misadventure instigated by our squad of Keystone docs, and he was too angry to talk to the family.
“You tell them,” he said sharply to me, his words bullets, as if I were responsible for the whole mess. Dutifully, I went off to the alcove where the family had assembled anxiously, fearing the worst. On my arrival, they rose as if one, and I mumbled something about how we had handled a succession of problems, not mentioning, of course, our own complicity in the troubles. Nevertheless, the family was overjoyed that old lady had survived. They then praised me as a wonderful doctor, indeed, a miracle worker who had snapped the family matriarch from the jaws of death—even if those jaws were iatrogenic.
Within a few minutes of the digoxin administration, the heart monitor began to beep wildly and the screen exploded with a profusion of mean-looking extra beats, which were obviously premature ventricular contractions.
Search for Balance on Uncertain Ground
Digoxin, ventricular arrhythmia; lidocaine, seizures; and Valium, respiratory depression. These are all well-known side effects of the drugs and, as they say, stuff happens even if this trifecta of badness was a rare occurrence.
Stuff happens. Red cell aplasia with gold; leukopenia with propylthiouracil; cirrhosis with methotrexate. These are all calamities that can be caused by drugs that I prescribed in my practice. It goes without saying that I have seen gastrointestinal bleeds from nonsteroidal antiinflammatory drugs and osteoporosis from prednisone.
As I wrote in my previous column, drug safety is a personal matter. Because I have witnessed in my practice some very serious side effects, my personal reaction to certain drugs is likely to be different than that of my colleagues, and I worry more about writing some prescriptions than others. I am not reassured that certain side effects are rare. My own personal experience has taught me that “rare” means it can happen—a tablet of aspirin can provoke near fatal asthma and allopurinol can make the skin blister and boil.
All providers want to use drugs as safely as possible and therefore constantly and carefully try to balance the risks and benefits of anything they prescribe. This balancing is inherently uncertain, because risks are often unknown or subject to perception, bias, and misinformation. Furthermore, the time dimension is key, and it may take years of observation and very large post-marketing studies before the data are sufficient to truly know what the risks and benefits of any therapy are.
In rheumatology, I doubt that we have the metrics to judge the pluses and minuses of our therapies. Do we really know, for example, the benefit of a difference in a Sharp score of 2 between competing treatments in terms of the risks of immunosuppression that may be needed for its achievement?
Pending more real-world data (hopefully to come from the ACR Rheumatology Clinical Registry), advances in pharmacogenomics and pharmacogenetics, and more robust analytic techniques to assay the benefits of therapy in a more objective way, increasing the safe use of drugs will be the personal matter of each provider.
For my part, I try to follow Pisetsky’s rules, of which, trust me, there are many. For those of you who do not know these rules, here are two more:
- Don’t prescribe a drug if you don’t have to.
- If a drug doesn’t work, stop it.
For those of you who have your own rules, please write in. Hopefully, together, we can prove Peltzman wrong.
Dr. Pisetsky is physician editor of The Rheumatologist and professor of medicine and immunology at Duke University Medical Center in Durham, N.C. Contact him at [email protected].
