This month, Dr. Sydney Spiesel discusses where there's a link between antibiotics and the development of asthma, why the rate of Caesarean deliveries is rising, the latest on bird-flu vaccine, and the importance of malaria medications. (Click here and here for the last two monthly columns.)
Antibiotics and asthma: Is there a connection?
The fear: On the subject of antibiotics, I often encounter parents at one of two extremes. Some regard the medications as miraculously able to fix any problem. These families regard me as mean-spirited when I won't prescribe antibiotics for a child who is suffering from a cold. At the other pole are families who must be persuaded to give the medications, even when their children are seriously ill. These parents usually accept treatment in the end, but they do so worrying that the antibiotics will weaken their children's immune systems and make them susceptible to all manner of illnesses—especially asthma. Some doctors share this fear. The incidence of asthma has been rising steadily for about 50 years, and one explanation is the "hygiene hypothesis"—the idea that asthma and other allergic diseases are on the rise because children are exposed less to dirt and germs and, perhaps, because they take more germ-killing antibiotics.
State of the science: A number of studies have tried to examine the relationship between early exposure to antibiotics and the development of asthma. Recently, Dr. Fawziah Marra and a group of colleagues at the University of British Columbia put together the results of eight of the better published studies. Marra's review concluded that taking antibiotics during the first year of life did increase the risk of asthma—and it seemed as if the increase was substantial. But I am not sure about this conclusion, for two reasons.
Caveat No. 1: My first doubt is methodological. Half the studies Marra surveyed were retrospective. Parents of children with and without asthma were asked to remember if their kids had received antibiotics in the first year. There was a marked difference between the two groups. The parents of ill children remembered antibiotics being given to their child almost three times as frequently than did parents of healthy children. But we need to be careful in interpreting results like this. Memory is tricky, and there is a great risk that expecting antibiotic use to cause asthma might lead parents to remember more antibiotic use than actually took place. Conversely, knowing that your child is asthma-free might lead a parent to forget about some early rounds of antibiotics. In my practice, I am often struck by the discrepancies between parental memory and the medical record of illnesses, medication use, and even allergies to medicines.
Caveat No. 2: The results of the four prospective studies in Marra's review add to my skepticism about a causal link. They started with children who were or were not treated with antibiotics in the first year and then tracked the proportion of each group that eventually developed asthma. These studies are more likely to be valid because they don't depend on remembrance of past events. And they showed a much smaller positive association between early antibiotic use and later development of asthma—a 12 percent increase in risk.
Caveat No. 3: Even if there is a clear, positive association between early antibiotic exposure and asthma, what do we make of it? Association, as I am too fond of pointing out, is not causation. Suppose, for example, that a child diagnosed with an ear infection (as happened in about 17 percent of office visits, according to the old literature) was treated with an antibiotic and later developed asthma. We know that certain viral infections lead to ear infections. And we strongly suspect that these infections predispose children to develop asthma. So, it's hard to sort out whether the antibiotic given for the ear infection, as opposed to the viral disease behind the infection, led to the asthma.
Conclusion: The only solid thing we can take away from Marra's analysis is that the proper experiments for determining a link between antibiotics and asthma haven't been done yet. Personally, I doubt there's a link, but I'm open to being convinced.
Caesarean deliveries: Should mothers be able to choose?
The question: Each year on the maternity floors in the hospital where I practice, I see more pink name labels, for Caesarean deliveries, and fewer white ones, for mothers who deliver vaginally. The trend away from vaginal deliveries is longstanding. In 1965, the Caesarean rate in the United States was 4.5 percent; in 1975 it was 10.4 percent; and in 1985 it was reached 22.7 percent. Now C-sections account for about 30 percent of all deliveries. Why?
The explanation: Some of the increase surely comes from doctors' fear of litigation. Statistically, an obstetrician's defense in a malpractice case that involves a damaged baby is likely to be stronger for a C-section delivery than for a vaginal one. But there is more to it. As physicians do more of these surgical procedures, their skills improve. And if they correspondingly deliver babies vaginally less often, they will likely lose some skills for difficult obstetrical problems. In addition, all surgery has become safer as anesthesia has improved. And the risk for vaginal birth after a previous C-section has been found to be greater than previously believed, which means that one Caesarean almost guarantees future Caesareans. Finally, the increase in C-sections is driven by a phenomenon that used to be quite rare—Caesarean delivery upon a mother's request. A woman might ask because she's fearful of the birth process, or wants to control the timing of delivery, or for many other reasons.
State of the science: Since nobody has been collecting statistics on how often women simply choose to have C-sections, we can only guess at how often it occurs. But anecdotally, Caesarean delivery upon request seems to be on the upswing. The National Institutes of Health just conducted a conference of experts to examine the benefits and risks for mothers making this choice and for their offspring. The report of the conference—still in draft form—is an unusually cogent explication of the risks and benefits of each method of delivery. Let me tell you about what it contains:
Risks: The main perils for Ceasarean delivery are 1) the risks inherent in anesthesia (though these have been dramatically reduced by the pulse oximeter—the little gizmo on your fingertip that tells the anesthesiologist the instant that things might begin to go badly, in plenty of time to fix things); 2) increased length of hospital stay; 3) greater chance of infection; 4) increased risk of problems with subsequent deliveries, especially vaginal ones; 5) perhaps some interference with beginning to breastfeed. Caesarean delivery also clearly predisposes babies to early, usually mild respiratory problems—"transient tachypnea of the newborn," or TTN, which results when amniotic fluid is retained in the baby's lungs.
More risks: But vaginal delivery also has risks of which we've become increasingly aware. It may increase the chance of subsequent urinary incontinence, as well as the danger of complications, when a failed attempted vaginal delivery necessitates emergency surgery. Dangers to the baby include 1) increased risk of stillbirth; 2) perhaps greater probability of oxygen deprivation; 3) greater chance of birth injury; and 4) greater likelihood of infection.
Conclusion: The NIH report finally poses this question: If the dangers are approximately equal in either direction (though that's not yet completely clear), should women have the right to decide whether to deliver vaginally or to deliver by Caesarean section? Historically they have not often been given that choice, and honestly, until now I thought the medical evidence so favored vaginal delivery (except when a C-section is strongly indicated) that I never thought much about patient choice and autonomy in picking a childbirth method. The NIH report challenged and expanded my thinking. Though I would continue almost always to favor vaginal delivery over purely elective C-section, the choice may be a little less obvious than I previously thought.
Bird-flu vaccine: Can we make enough of it in time?
The fear: We wait in America, holding our breath, to see if H5N1 influenza will come to our shores. It will. We also hold our breath, along with the rest of the world, to see if this flu that's so lethal and contagious for birds will change and become similarly lethal and contagious for humans. At least once before, in 1918, a bird flu mutated and killed between 50 million and 100 million people worldwide. What's the latest on trying to prevent another pandemic?
Quarantine: Traditionally, three methods have been used to control epidemics. By imposing quarantine, a society can force physical containment of infectious people. Historically, this has rarely worked. Because influenza is so infectious, it would be hard to control an epidemic even if only a few people slip out of quarantine (or turn out to be infectious before the nature of their illness becomes clear).
Treatment: For flu, the treatment choices are not so good and are getting worse. There are presently two classes of medication. One, amantadine, has become completely worthless for H5N1 flu because the virus is now uniformly resistant to it. The second, Tamiflu, is effective for now. But there is evidence that if the drug were to be widely distributed (as it would be in an epidemic), the virus would likely become resistant to it, too.
Immunization: That leaves vaccines. This method is especially attractive because we have experience with using vaccines to prevent influenza and because immunization is an inherently efficient method of warding off illness. Vaccines transform patients into their own medicine manufacturers by priming them to produce protective antibodies—effective, specific medications, made by the body and generally delivered exactly to the site where they are most needed.
State of the science: So, can we depend on flu vaccines to protect against H5N1 flu? That is the question addressed in a recent report by a team headed by Dr. John J. Treanor of the University of Rochester. Conventional flu vaccines—the kind many of us take annually—are made from influenza virus grown in fertile chicken eggs, starting with the strain of human influenza predicted to be important for each anticipated flu season. For Treanor's study, an influenza vaccine was made using the same technology, with starting material collected during an outbreak of H5N1 in Vietnam. A wide range of doses were given to a group of human subjects to determine 1) whether the vaccine was safe and 2) how much of the vaccine's active ingredient was required to generate production.
Results: Here's the good news: In the right dose range, the experimental vaccine more than half the time stimulated the development of what we assume are protective antibody levels. Side effects were common but generally minor (for instance, local pain and tenderness at the injection site). But here's the bad news: It took about 12 times as much of the experimental vaccine as is usually in an annual flu shot to provide a roughly comparable level of protection. Well, that wouldn't be so bad—except that factories now working at absolutely full capacity simply can't keep up with the current need. If you were one of the many people who couldn't get a flu shot this year, I don't need to point out that even in a relatively mild flu season, demand exceeded supply.
Conclusion: So, we now have a promising but weak H5N1 vaccine. We need to find ways to manufacture it in much larger quantities than are presently possible (click here and here for two proposals) or to dramatically increase its potency. And the clock is ticking.
Malaria medications: Take them.
The choices: This is the season when patients waltz into my office to be prepped for their tropical vacations. I have to overcome my jealousy and pretend to be good-natured as I get them ready. Often, they want to know which malaria medication to take. The oldest (and cheapest) drug is chloroquine. Unfortunately, the organism that causes malaria is wily and has evolved to be resistant to chloroquine in most of the world. (The CDC malaria Web site for travelers will tell you if you are fortunate enough to be headed to a destination where chloroquine still works. If not, you have three major choices: mefloquine (Larium), atovaquone plus proguanil (Malarone), and doxycycline. Each choice has disadvantages, including negative side effects. All have to be taken regularly (chloroquine and Larium weekly; Malarone and doxycycline daily). And each must be taken before you leave and after you return. None are 100 percent effective.
The problem: Since many of my trekking patients are adolescents or young adults who might well blow off even my advice, I worry what might happen if they neglect to take their anti-malarial medication. It turns out that I am right to.
State of the science: A recent report by Dr. Gerard Krause and some colleagues at the Robert Koch Institute in Berlin looked at almost 4,000 cases of malaria reported in Germany by travelers over an 11-year period. Does taking your medicine as directed protect against death, the researchers asked?
Results: Anti-malaria drugs clearly decrease the risk of dying from the disease. The protective effect is substantially more powerful when travelers follow to the letter instructions for taking the drugs. It's impossible to say whether the newer medicines were more or less effective because styles of preventive treatment changed over the course of the study. Malaria, however, remains serious business: 116 people in the study died—3 percent of those who contracted it and came home—despite Germany's excellent medical care. The risk of death was higher for older patients, so pre-travel counseling for elderly travelers should be an especial priority.
Disturbing trend?: In almost every year since it began, the study found, a smaller percentage of patients with malaria reported that they had taken their medication. There are two ways to interpret this information—the study doesn't tell us which is right. One (bad) possibility is that fewer travelers are making use of the available medicines, perhaps because doctors aren't insisting enough on the importance of prophylaxis or because of concerns about side effects. Another (good) possibility is that the present treatments are much more effective than the old ones were, so fewer cases of malaria are occurring in travelers taking prophylactic medicines. I hope it is the latter.