Antibiotics for Strep Pharyngitis: The Pediatric Perspective

Around this time last year, a post on R.E.B.E.L. EM discussed the evidence surrounding the risks and benefits of antibiotics in Strep pharyngitis.  The general take-home was that it should not be done.

Since the article did not specifically discuss differences between children and adults, Brendan Fitzpatrick, one of our 4th year residents sent the article to Indi Trehan, who is boarded in both pediatric emergency medicine and infectious disease.  Thankfully, between saving babies in Malawi, writing grants and submitting papers, he crafted this response which he agreed to let us share on the blog:


>> Fitzpatrick, Brendan wrote:

Let me know what you think:

Brendan M. Fitzpatrick, MD
PGY-4, Division of Emergency Medicine
Washington University School of Medicine in St. Louis


> Trehan, Indi wrote:

Hey man,

I talked about this Strep throat thing with some (admittedly more biased and traditional) folks in ID. You're absolutely right, the incidence of ARF/RHD is going down and there is very little left in rich countries, but I don't think any of us are comfortable yet not treating this.

Some thoughts, going through the article from top to bottom:

Background - 10 million patients treated with antibiotics annually but less than 10% have Strep...

True, but this is 10 million adults and most places don't test them for Strep. One of the outcomes of that study was the re-emphasis on the importance of only testing people with symptoms (Centor score at least 2). In the three big children's hospitals I have worked in, this is standard of care -- we only treat those who have tested positive.

C. diff and allergic reactions

The rate of C. diff is extremely low in children -- for example we use clindamycin like candy. Otherwise healthy ambulatory kids who come in with a sore throat just don't get C. diff from the occasional course of antibiotics.

Yes, allergies are always a risk with antibiotics. Hence an additional reason to limit their usage by only testing and treating appropriately as above.

Argument #1

Agreed, reducing symptoms is uninteresting to most of us.

Argument #2

Agreed, suppurative complications are not why most of us treat.

Argument #3

I think this is probably a place where adult data doesn't apply to kids. The rate of RHD/RF reported in that military study in the 1950s is indeed quite low but that's because that was in military recruits. RHD/RF simply happens at much much lower rates in adults than kids -- it really is a disease that is almost entirely of 5-15 year olds, a few younger and a few older of course, but it doesn't surprise me at all to see only a 1-2% rate among what is probably 17-25 year olds.

But even aside from this fact, these studies from the 50's and 60's are all quite small and given the very low rate of these complications, a meta-analysis is in order in order to be able to have enough power to detect a difference. The Cochrane review on this subject does this meta-analysis for us -- see Analyses 4.1 and 4.2 which found that giving antibiotics reduced ARF -- RR 0.27 (95% CI 0.12-0.60). Giving penicillin specifically was also studied -- RR 0.27 (95% CI 0.14-0.50).

Need to treat 2 million patients in order to prevent a single case of RF...

This number is not exactly right. It is correct in its intention of pointing out that you need to treat a ton of people to prevent one complication, but the denominator is not correct since we really shouldn't be testing or treating adults and we shouldn't be treating those without a positive rapid test or culture. So if you limit the testing and treatment to the high risk population (i.e., children) and then only treat those who have a positive test, this 2 million number falls dramatically.

The point about hygiene as the real cause of the decrease in rates of ARF is exactly correct and I'm glad they are pointing this out. In the same way that polio's eradication in the US is just as much a result of improved hygiene as it is from vaccination, we still need to keep immunizing kids against polio due to the unpredictable risk of it flaring up in an outbreak at any time. (Admittedly, this analogy falls apart when comparing the almost-zero risk of complications due to vaccination when compared to the slightly higher risk of penicillin allergy.)

The thing with Group A Strep is that certain strains are rheumatogenic and far more likely to cause ARF. Which ones these are is quite unpredictable. There also seems to be certain people that are particularly susceptible because of their own genetic predisposition (this comes from twin studies and such). And of course who those people are is unknown with our limited genetic techniques.

When you put these together, sometimes you get the perfect storm of sporadic or widespread cases of ARF. So some fair number of kids will still get ARF/RHD and then there are also big outbreaks that arise periodically and unpredictably -- the most famous was in Utah in the 80's -- and so keeping the burden of this bacteria low and limiting its spread seem to be a worthwhile public health goal.

But overall, your point is well taken -- this is a disease with declining incidence and morbidity and we are getting close to the point of not needing to treat it in the US, with the caveats that the numbers derived from adults don't necessarily apply to children who have a much higher incidence of rheumatic complications and that those who will suffer these is relatively unpredictable, especially given the problem that kids are not the cleanest in the world and quickly spread bacteria around their house and day care centers and schools. Another small practical problem we have to remember is "standard of care" -- so if you don't test or treat a child with Strep throat who goes on to develop ARF/RHD and the community standard of care -- "what any reasonable practitioner would do" -- still says to test and treat, then you are up a creek when it comes to malpractice. But I do hope that when IDSA looks to update and revise their guideline (as they do every few years), they take a serious look at this changing epidemiology and decide to narrow their recommendations for treatment...