Still a Question of Benefit vs. Harm
In the March issue of EP Monthly I wrote an article, “Antibiotics for Strep Do More Harm Than Good”. The piece prompted some interesting and strong responses on the EP Monthly internet pages, including one from perhaps the most famous name in pharyngitis — Dr. Robert Centor of the Centor criteria. This month we’ll go through some of these responses to answer some questions and to boil the issues down to their essence.
But first, on both sides of the argument we’ll need to agree on an important ground rule: the issue is benefit versus harm. If antibiotics result in demonstrably more benefit than harm, or more harm than benefit, then we have our answer. This seems self-evident, an almost childish point. But it’s not. We have been trained in medicine to believe that bacterial colonization of any anatomic structure should be feared, and that bacterial eradication is a therapeutic goal. But no one dies of strep throat. The human immune system is, in general, quite up to the task of handling this self-remitting condition. Therefore the goals of treatment are not bacterial eradication, they are 1) prevention of complications and, 2) reduction of symptoms. These are the benefits we seek. Now, to the essential business at hand: more benefit or more harm?
One of the first responses to the article asked whether or not post-streptococcal glomerulonephritis is reduced by antibiotics. Important question. In eleven thousand strep throats in trials only two mild cases occurred. There is therefore no evidence to suggest that the condition is preventable. If it is preventable, it appears we would need to treat well over five thousand strep throats just to prevent one case of glomerulonephritis. At that point we’re likely to have caused at least a few cases of fatal or near fatal anaphylaxis. Glomerulonephritis is therefore not considered by any professional groups or experts to be a reason for antibiotic use.
A second batch of questions asked about abscesses. While retropharyngeal abscess has never been reported in strep throat trials, peritonsillar abscess has. Antibiotics appeared to reduce PTA, though roughly 143 cases of strep would have to be treated in order to prevent one PTA. Given the common side effects of antibiotics, including diarrhea in 20%, rash in 5%, and yeast infections in 20% of women, this is a tough number to swallow (no pun intended). The use of antibiotics to protect against PTA becomes even more daunting when one considers that more than 90% of PTA’s are easily managed with outpatient needle drainage. Fewer than one in ten requires admission or a second procedure, which means we can multiply our initial number (143) by ten to prevent a complicated PTA. This is over a thousand prescriptions, and it explains why experts and professional guidelines also don’t generally recommend the use of antibiotics to protect against abscess.
A few smart comments noted that rheumatic fever, while rare, is dangerous, and the questioner wondered if rheumatic fever following pharyngitis might therefore lead to a malpractice claim. I consulted the good Dr. Greg Henry, one of the most experienced emergency medicine legal experts in the country. He knew of no cases. Dr. Henry pointed me to the Physician Insurers Association of America, an organization that insures 60% of the privately practicing physicians in the country and maintains the largest national database of malpractice claims, dating back more than 25 years. They haven’t recorded any such claims either.
Some respondents wondered if perhaps widespread antibiotic use is the reason for the decline in rheumatic fever. Bacteriologists and epidemiologists note, however, that rheumatic fever began its precipitous decline before antibiotics were introduced, and that population crowding, hygiene, access to care, and other living condition issues have led to rheumatic fever’s decline. This explains why the disease is still a serious problem in developing, but not industrialized, nations.
At least one web posting pointed out that many professional guidelines recommend antibiotics. This is true. Many guidelines in industrialized nations other than ours, however, specifically recommend against antibiotics for simple strep throat. And there is a growing consensus, as discussed by Dr. Robert Centor in his recent editorial published in the Journal of the American Medical Association, that our guidelines are often based more on opinion than evidence. The guideline development process, Dr. Centor notes, is in serious need of reform.
Speaking of Dr. Centor, I’m proud to say that we spoke. I was a bit star-struck, and our conversation evolved slowly. But it evolved. He and I agreed that sore throat presentations, particularly those that are not due to simple pharyngitis, are an important source of morbidity and even mortality, and should be taken seriously. We also agreed that rheumatic fever, glomerulonephritis, and abscess are all flimsy reasons for antibiotic use. We also agreed that antibiotics are over-prescribed.
We disagree on surprisingly minor points. Dr. Centor believes that symptom reduction is a primary and adequate reason to treat strep throat with antibiotics. He believes that patients who are ill (generally with at least three Centor criteria) benefit adequately to make them worthwhile. I, on the other hand, believe that anti-inflammatory agents, steroids, and narcotics, are all better. I also believe they have fewer adverse effects. Dr. Centor notes that while there may be little data to show it, antibiotics also probably decrease spread of the disease. And while I have seen no data to support the claim, I concede that it is possible. Perhaps future studies will tell. Dr. Centor, I will add, is a gentleman and a scholar. And for now, we’ll agree to disagree—but only a little.
Thanks everyone for your spirited commentary and discussion. More to come….
But first, on both sides of the argument we’ll need to agree on an important ground rule: the issue is benefit versus harm. If antibiotics result in demonstrably more benefit than harm, or more harm than benefit, then we have our answer. This seems self-evident, an almost childish point. But it’s not. We have been trained in medicine to believe that bacterial colonization of any anatomic structure should be feared, and that bacterial eradication is a therapeutic goal. But no one dies of strep throat. The human immune system is, in general, quite up to the task of handling this self-remitting condition. Therefore the goals of treatment are not bacterial eradication, they are 1) prevention of complications and, 2) reduction of symptoms. These are the benefits we seek. Now, to the essential business at hand: more benefit or more harm?
One of the first responses to the article asked whether or not post-streptococcal glomerulonephritis is reduced by antibiotics. Important question. In eleven thousand strep throats in trials only two mild cases occurred. There is therefore no evidence to suggest that the condition is preventable. If it is preventable, it appears we would need to treat well over five thousand strep throats just to prevent one case of glomerulonephritis. At that point we’re likely to have caused at least a few cases of fatal or near fatal anaphylaxis. Glomerulonephritis is therefore not considered by any professional groups or experts to be a reason for antibiotic use.
A second batch of questions asked about abscesses. While retropharyngeal abscess has never been reported in strep throat trials, peritonsillar abscess has. Antibiotics appeared to reduce PTA, though roughly 143 cases of strep would have to be treated in order to prevent one PTA. Given the common side effects of antibiotics, including diarrhea in 20%, rash in 5%, and yeast infections in 20% of women, this is a tough number to swallow (no pun intended). The use of antibiotics to protect against PTA becomes even more daunting when one considers that more than 90% of PTA’s are easily managed with outpatient needle drainage. Fewer than one in ten requires admission or a second procedure, which means we can multiply our initial number (143) by ten to prevent a complicated PTA. This is over a thousand prescriptions, and it explains why experts and professional guidelines also don’t generally recommend the use of antibiotics to protect against abscess.
A few smart comments noted that rheumatic fever, while rare, is dangerous, and the questioner wondered if rheumatic fever following pharyngitis might therefore lead to a malpractice claim. I consulted the good Dr. Greg Henry, one of the most experienced emergency medicine legal experts in the country. He knew of no cases. Dr. Henry pointed me to the Physician Insurers Association of America, an organization that insures 60% of the privately practicing physicians in the country and maintains the largest national database of malpractice claims, dating back more than 25 years. They haven’t recorded any such claims either.
Some respondents wondered if perhaps widespread antibiotic use is the reason for the decline in rheumatic fever. Bacteriologists and epidemiologists note, however, that rheumatic fever began its precipitous decline before antibiotics were introduced, and that population crowding, hygiene, access to care, and other living condition issues have led to rheumatic fever’s decline. This explains why the disease is still a serious problem in developing, but not industrialized, nations.
At least one web posting pointed out that many professional guidelines recommend antibiotics. This is true. Many guidelines in industrialized nations other than ours, however, specifically recommend against antibiotics for simple strep throat. And there is a growing consensus, as discussed by Dr. Robert Centor in his recent editorial published in the Journal of the American Medical Association, that our guidelines are often based more on opinion than evidence. The guideline development process, Dr. Centor notes, is in serious need of reform.
Speaking of Dr. Centor, I’m proud to say that we spoke. I was a bit star-struck, and our conversation evolved slowly. But it evolved. He and I agreed that sore throat presentations, particularly those that are not due to simple pharyngitis, are an important source of morbidity and even mortality, and should be taken seriously. We also agreed that rheumatic fever, glomerulonephritis, and abscess are all flimsy reasons for antibiotic use. We also agreed that antibiotics are over-prescribed.
We disagree on surprisingly minor points. Dr. Centor believes that symptom reduction is a primary and adequate reason to treat strep throat with antibiotics. He believes that patients who are ill (generally with at least three Centor criteria) benefit adequately to make them worthwhile. I, on the other hand, believe that anti-inflammatory agents, steroids, and narcotics, are all better. I also believe they have fewer adverse effects. Dr. Centor notes that while there may be little data to show it, antibiotics also probably decrease spread of the disease. And while I have seen no data to support the claim, I concede that it is possible. Perhaps future studies will tell. Dr. Centor, I will add, is a gentleman and a scholar. And for now, we’ll agree to disagree—but only a little.
Thanks everyone for your spirited commentary and discussion. More to come….