Certain medical problems are a pleasure to treat--when the diagnosis can be made with confidence and no major inconvenience to the patient, when the treatment is effective and safe, when the cure is complete. By all rights, strep throat should be one of these diagnoses.
In the 50s, when penicillin was still a miracle drug, strep throat must have been fun. Imagine Gregory Peck in a grade B black-and-white film, frowning slightly as he palpates his young patient's neck for lymph nodes. The lovely adolescent opens her mouth and he carefully swabs her throat. No gagging disturbs the solemnity of the moment. Gravely, he passes the swab over the culture plate and sets the plate in the incubator. Turning back to the young woman he confides, "I believe you have an infection with the streptococcus. If the bacteria multiply on the plate, penicillin will cure you."
"Thank you, doctor," she whispers hoarsely.
Sure enough, the suckers grow. And thanks to the wonders of antibiotics, Missy is able to return to choir practice. In a voice-over, Peck muses: "The specter of rheumatic fever, which used to lurk behind every sore throat, no longer haunts our children. Penicillin has brought us into a new age." Perhaps Missy isn't aware that the reason Dr. Peck is so satisfied with himself is that he has prevented a case of rheumatic fever. But she doesn't bother the doctor with questions: she's better and she's grateful. At the concert, she sings out to him in the audience, and he smiles wisely back.
Forty years later, the medical parts of the story haven't changed much. We still use throat cultures to diagnose streptococcal infection, although "rapid diagnosis" tests based on the chemical properties of the bacteria are almost as accurate and much faster. We still treat with penicillin, and our primary intent is still not to make the symptoms disappear but to prevent rheumatic fever--however quiet that intent is kept. It so happens, as studies have demonstrated only in the last ten years, that patients with streptococcal sore throats do improve more quickly when they take penicillin, but the sore throat will go away even without treatment in a week or so.
Rheumatic fever, on the other hand, is serious business. Many of us who grew up with antibiotics have never seen a case of the disease, but the "major criteria" and "minor criteria" for its diagnosis are part of the lore of the old days. About 2 percent of patients with untreated strep throats go on to develop an illness of three to six months' duration, characterized by fever, involuntary movements, heart damage, arthritis, and skin rashes. No one knows for sure why this happens, but the leading theory is that the people who succumb to rheumatic fever have immune systems that go haywire fighting the streptococcal bacteria and end up fighting their own tissue. Some patients make a complete recovery, but others are left with permanent heart damage, which was a common cause of premature death in the preantibiotic era. Even now, in third-world countries, experts estimate that rheumatic heart disease causes 25 to 40 percent of all cardiovascular disease. The peak age of strep infections--and thus of rheumatic fever--is between five and fifteen, so these are patients in the prime of life suffering from an eminently preventable disease.
Besides decreasing the incidence of rheumatic fever, prompt antibiotic treatment helps contain the spread of streptococcal infection, since the patient is not contagious as long. Streptococci are transmitted from person to person in the saliva droplets from a sneeze or a cough. This transmission requires relatively close contact, so closed populations of young people, like schoolchildren or military recruits, are the most susceptible to outbreaks of rheumatic fever. These still occur sporadically in the United States. The Centers for Disease Control reported ten cases of rheumatic fever at the Naval Training Center in San Diego, California, between December 15, 1986, and July 15, 1987. Interestingly, until 1980 all incoming recruits received penicillin shots to prevent strep infection. The injections were discontinued because the risk of acute rheumatic fever appeared so low, but they were reinstituted after the outbreak. One of the patients who developed rheumatic fever had received oral penicillin for his sore throat but had not finished the prescription. Cecil's Textbook of Medicine reminds us, "Because signs and symptoms frequently subside in a few days, there is a tendency to shorten the time antibiotics are given. Brief periods of antibiotic therapy do not eliminate the streptococci from the pharynx. Patients treated briefly have a greater risk of developing acute rheumatic fever than those who are adequately treated for at least ten days."
The conquest of rheumatic fever is a tale that stirs the heart of every red-blooded health worker. Even the staid journal of the American Medical Association exulted in an editorial in 1986, "The remarkable change in the natural history of group A beta-hemolytic streptococcal pharyngitis and its sequelae is one of the great success stories of the 20th century." But the exquisite irony of the clinical situation is that 90 percent of sore throats are caused by viruses, which don't respond to antibiotic treatment. Yet patients with sore throats seek antibiotics for the relief of symptoms, and are less than impressed with the physician who simply informs them they don't have strep.
As an internist, I know that many patients are so accustomed to antibiotic treatment they perceive the health-care worker as an obstacle in the way of their prescription. They know they recovered when they took penicillin before--and all sore throats are the same, and of course it was the drug that cured them, not time. They don't see the point in waiting 24 to 48 hours for a culture. If the doctor is lucky, he may have one of the more rapid tests for identifying strep, so that he can decide to treat based on that result rather than his assessment of the clinical symptoms. At least then he can hide behind the test, because patients do not appreciate hearing that they probably have a virus. They feel that the practitioner has not recognized the severity of their symptoms and the urgency with which they need to recover.
After a few of these confrontations, the typical earnest young practitioner is so deflated that he or she asks, "Why not just give everyone penicillin?" Aside from the cost, which would be prohibitive, and the substantial risk of allergic reactions and side effects, there is a compelling public-health reason not to hand out antibiotics to all comers: they stop working. The streptococcus is one of the few bacteria that have not developed resistance to penicillin over the last 50 years. Antibiotics are not general poisons; they depend on the specific vulnerability of bacteria to be effective. Penicillin, for instance, prevents the streptococcal bacterium from making its cell wall by interfering with an important chemical step. Other bacteria with more complicated cell walls are unaffected by the drug: some bacteria that used to be more susceptible have developed the ability to inactivate the penicillin molecule, so they have acquired resistance. Viruses do not have their own cell walls but are parasites that set up shop within our own cells. A chemical that indiscriminately killed bacteria and viruses would probably kill us as well.
So we don't want to give everyone penicillin. Why not at least culture everyone? The major objection to this course of action is cold cash. Each culture costs money, and even in a setting where the patient pays out of pocket for the test, someone has to absorb a minimum of about $4. No one can afford to practice today without trying to control unnecessary cost, and a test that will be negative 90 percent of the time is just too expensive when multiplied by all the people with sore throats who seek medical attention. Because of this problem, there have been many attempts in the medical literature to identify the clinical characteristics that separate those with strep throat from those without. For instance, a 1986 article in the Journal of the American Medical Association compared four statistical models for predicting the presence of the bacteria based on clinical signs: discriminant analysis, branching algorithm, logistic regression, and adjusted logistic regression. When it comes to cost-effectiveness studies, the statisticians go hog-wild.
Unfortunately, there are no symptoms or signs that invariably announce the streptococcus. The four most reliable indications of bacterial infection are a fever of 101 degrees or greater, pus on the tonsils, swollen lymph nodes in the neck ("swollen glands"), and the absence of a cough. (Cough accompanying a sore throat suggests a virus.) A patient with a cough and none of the other signs would have a probability of less than 10 percent; a patient with no cough and all of the other signs would have a probability approaching 50 percent.
Most physicians are familiar with this data, but there is no consensus on how it should be used. A survey in The Annals of Internal Medicine showed that 23 percent of primary-care physicians never use throat cultures, 25 percent always do, and 52 percent selectively culture. This is not as crazy as it sounds, since different patients require different strategies. In a transient emergency-room population, where many patients have no telephone, it doesn't make sense to culture much, since it will be impossible to reach the patient with the results in 24 hours. Better to treat those with two or more signs and forget it. In a stable suburban practice, with a lower prevalence of strep anyway because of less crowding, it makes sense to rely more on cultures, since in this era of the answering machine, car phone, and beeper, those folks are not hard to reach.
Doctors are not heroes today, and Gregory Peck has retired. The scene opens on a beleaguered Meryl Streep on the phone. The patient, calling from his car, wants her to call in an antibiotic prescription to a pharmacy. He has laryngitis, a cough, and a runny nose. No fever. She explains that it sounds like a virus, particularly the laryngitis, but he interrupts to say that he has to give a presentation tomorrow and he has to be better by then. She insists that she cannot prescribe antibiotics without seeing him; he insists that she fit him in today. "OK. Come at 5:30." She knows that she has only postponed the showdown, and sure enough, he is not pleased when she examines him and says she thinks it's a virus. The offer of a throat culture doesn't mollify him; it takes too long.
What would Gregory Peck have done? In the old movies, patients didn't disagree with the doctor. Of course, it's the end of a long day, and sometimes she just prescribes the penicillin, what the hell. But then she doesn't respect herself in the morning. She says her final "No, I'm sorry. I don't think you need antibiotics," and walks out. To Dr. Streep, it is difficult to withhold a drug that a patient is convinced will cure him, even though she knows she's right. Such power probably felt fine in the 50s, when the doctor, like the government official and the clergyman, was a father figure. Father knew best, after all. But in the 90s, the white coat does not automatically confer moral authority. Physicians must earn it anew with each patient.
The whole scenario demonstrates how out of synch doctors and patients can be. Physicians aren't putting down patients when they tell them it's a virus: after all, AIDS is a virus, and no one would call it a trivial problem. Patients aren't trying to exasperate physicians, but the pace of their lives doesn't allow for rest and chicken soup. They are passing on their stress to the physician. The well-meaning practitioner assumes that the patient came to see him in order to ascertain the presence or absence of strep. The patient, who hurts, is disappointed with the doctor who "didn't do anything" and took his money. Two separate realities create an unsatisfying encounter. The patient is frustrated by the physician's dismissal; the doctor is frustrated by the patient's clear vote of no confidence.
Joe Girard, the champion car salesman who wrote How to Sell Anything to Anybody, has a "rule of 250." He says that no salesperson can ever afford to alienate even one customer, because the customer has contacts with about 250 other people who will certainly hear about a negative experience (but not necessarily a positive one). The rule of 250 explains why polls show people like their own doctor, but think the rest of them are jerks. It also implies that a problem like a sore throat, where most of the time the doctor's treatment will (quite properly) run counter to the patient's preference, can wreak as much havoc with public confidence as actual negligence.
In the future, there may be no need for doctor-patient interaction to treat strep throat. Home diagnostic kits are already on the market, and a system to automatically dispense antibiotics to those who test positive undoubtedly is being considered somewhere. If that happens, physicians will lose the satisfaction of making the diagnosis, but they will be relieved of the burden of rationing the technology. Computer-generated decision trees will take the place of clinical judgment in many areas, at once dictating practice and freeing the individual doctor from responsibility.
It's probably progress. Yet in a society where no one, from day-care age to retirement, can afford to be sick, the 90 percent of sore-throat patients who don't have strep will still be disgruntled. With more research, we may conquer the common cold. But for now, only a reactionary would consider the possibility of accepting minor illness as a part of life, and allowing each other the luxury of time to recover.
Art accompanying story in printed newspaper (not available in this archive): illustration/John Figler.