Why do we treat strep throats with antibiotics?

Clinicians like making the diagnosis of strep (streptococcal) throat (and patients hate being told they have a viral sore throat) for the same reasons: a belief in the ability of antibiotics to shorten the illness and prevent serious complications That’s what we doctors were taught, and that’s what we’ve taught our patients. How justified is this belief?  Well, there are problems.

First, about half of the patients we diagnose with strep throat do not actually have strep throat. Streptococcal bacteria commonly inhabit the back of the throat (oropharynx) without causing illness, a process called ‘colonization’ which is not associated with symptoms. On average, 12% of healthy children have strep in their throat at any given time, and  many as one in four children may have strep in their throat during the winter. If a patient has a viral sore throat and a rapid strep test or a throat culture is done,  the strep that is found is an innocent bystander, not related to the illness, and antibiotics cannot help (but can cause side effects).

Second, multiple studies show that strep throat usually resolves on its own, without antibiotics, in about 3-5 days, and that treatment with antibiotics shortens the course of a strep throat by an average of less than 24 hours.

Finally, it is not clear that antibiotic treatment is important to prevent complications. Take the case of the most feared complication, rheumatic fever (RF).

During most of human history, diphtheria was the rightly feared throat infection. Before 1900, it was the leading cause of death in children under 5, and the overall mortality rate was 10%. By the 1940s, antitoxin, vaccination, and improved living conditions had largely eliminated this threat, a huge public health victory. (The CDC reports fewer than one case annually in the US since 2000.) Strep throat, on the other hand, was not identified as an important illness needing diagnosis and treatment until after antibiotics were in general use. 

Antibiotics entered medical use with the sulfonamides in the early 1930s. Penicillin became clinically available around 1945. At that time, rheumatic fever (see here andhere) was a common and serious illness with cardiac and joint manifestations. It was known to be caused by infections with certain strains of strep.

During the 1940s, Warren Air Force base near Cheyenne, Wyoming, suffered repeated outbreaks of a particularly virulent strain of strep associated with a high risk of rheumatic fever. A study of recruits at that time showed that the frequency of rheumatic fever as a consequence of strep throat could be reduced from 2% to 1% by a course of penicillin. In this study, for every 60 recruits treated, one case of rheumatic fever was prevented (and one case was not prevented). Publication of this data in the 1950s led to the current standard of care: treat strep throats with 10 days of penicillin (or an equivalent in the case of penicillin allergy) to prevent rheumatic fever. Unfortunately, while the nature of the problem has changed, clinical practice has not.

Rheumatic fever is now incredibly rare in the developed world.  It is not entirely clear why there is so little RF in the US, but it does not appear to be because we are treating strep throats with antibiotics. Interestingly, no patients have developed RF as a result of being treated for strep with placebo. Part of the explanation appears to be that the strains of strep that are prone to cause RF are now uncommon. At any rate, in 1994 the incidence in the US was fewer than 1 case per million population and the CDC stopped tracking it.

Peritonsillar abscess or quinsy (an infection with a collection of pus deep inside the tonsil) is the next most significant complication of strep throat, but preventing one case of peritonsillar abscess requires treatment of 4300 cases of strep throat (NNT = 4300).

Common sense suggests that treating strep throat with antibiotics might decrease infectivity, but this has not been studied, so there is no information about how much impact antibiotics have on spread.

The antibiotics used to treat strep are not without risk. They cause side effects ranging from rash through serious allergic reactions, antibiotic-associated colitis, and even death. The more antibiotics we use, the more we have to deal with resistant bacteria that are hard to treat and therefore require more potent antibiotics, which in turn increases the risk of bacterial resistance and side effects.

Based on current data in the US,  prevention of one case of RF would require treatment of over a million cases of strep, and would result in the following complications of antibiotic use:

  • allergic reactions and anaphylaxis (2,400 cases)
  • diarrhea (50,000 - 100,000 cases), including severe antibiotic associated colitis
  • rash (100,000 cases)

So, what should we do about sore throats? It is still important to have sore throats evaluated by a clinician. Other illnesses, including a wide range of treatable and some serious conditions, can be associated with sore throat: post nasal drip from allergies or sinus disease, mononucleosis, coxsackie virus, GERD, cancers of the throat, HIV, gonorrhea, Lemierre’s syndrome, herpes simplex, candida (yeast), chlamydia, mycoplasma, and peritonsillar abscess.  However, it appears that the benefits of identifying strep throat in order to treat it with antibiotics are quite small compared to the risks, and the data suggests that more patients are harmed than benefited by treatment of strep throat with antibiotics.


Links to more on this topic::