A cognitive error when treating adolescent pharyngitis

Since prior to my entrance to medical school, common wisdom for treating sore throats involved the prevention of rheumatic fever.  Since group A strep pharyngitis is the cause of most acute rheumatic fever, all efforts have focused on treating group A strep.   Studies in the 1950s showed that penicillin treatment decreased the probability of patients developing rheumatic fever.

The prevailing theory in the 50s and 60s, that we should diagnose group A strep and ignore all other pharyngitis etiologies, persists to the present.  Guidelines focus on the group A strep diagnostic problem.  Articles investigating inappropriate use of antibiotics for upper respiratory infections lump pharyngitis in with bronchitis and other upper respiratory complaints.  Those studies criticize physicians who provide antibiotics to patients who do not have a positive test for group A strep (either a rapid test or a culture).

The prevailing construct assumes that sore throats are sore throats regardless of age.  Many pharyngitis experts are actually streptococcal experts who focus on group A strep.  Many pharyngitis experts are pediatricians who focus on pre-adolescent pharyngitis.  They hold any other potential bacterial pathogen to a higher standard of proof that group A strep.

When one carefully reads the pharyngitis literature, especially the literature on complications and other bacterial pathogens, one finds an interesting observation.  Pre-adolescent pharyngitis actually is predominantly group A strep or viral, but adolescent (here I will use a loose definition of ages 15-30) pharyngitis has more varied bacterial etiologies.

We published the first article describing group C streptococcal pharyngitis as endemic and symptomatic in adolescents.  Multiple other articles, primarily from college student health, have also documented this finding.  The best randomized controlled trial of penicillin in adults found that group C pharyngitis patients had 1 day less symptoms when treated with penicillin.

More recently Fusobacterium necrophorum is gaining attention, primarily because of outstanding work in Denmark and England.  I recently published an article that uses published data to estimate that the complications (peritonsillar abscesses and Lemierre’s syndrome) from this organism are more dangerous and more devastating than those from group A strep in this country.

Lemierre syndrome is much more common as a complication of adolescent pharyngitis than is acute rheumatic fever.  It is deadlier and causes significant disabilities in some patients.

Yet current guidelines lump all pharyngitis into one bucket and only address group A streptococcal pharyngitis.  Rapid tests only detect group A pharyngitis.  These guidelines now mention group C and Fusobacterium but opine that we do not have enough evidence to worry about them.

Obviously I argue that we cannot wait for complete evidence in addressing adolescent Fusobacterium pharyngitis.  Lack of evidence for addressing adolescent pharyngitis differently is not equivalent to evidence that we should not treat adolescent pharyngitis differently.

Not all medical questions have sufficient evidence to make good decisions.  We physicians must purposefully extrapolate from what we know about pathophysiology and then make rational decisions.  I submit that if we knew that an adolescent had Fusobacterium pharyngitis we would have a logical and defensible reason to treat with antibiotics, specifically penicillin. But we do not have a readily available diagnostic test for Fusobacterium necrophorum because it is a difficult to grow anaerobic bacteria.

So we must ask ourselves how much “overtreatment” with penicillin is tolerable to prevent 1 patient developing a peritonsillar abscess or worse yet Lemierre’s syndrome.  Very few organisms show sensitivity to penicillin in 2014.  Fortunately strep (both group A and group C) and Fusobacterium remain sensitive.

Current criticism of treating adolescent pharyngitis based on signs and symptoms alone represents (in my mind) a cognitive error.  Our lack of prospective studies definitively answering this question should not exclude a logical approach to trying to prevent a rare but extraordinarily serious infection.

Robert Centor is an internal medicine physician who blogs at DB’s Medical Rants.

Comments are moderated before they are published. Please read the comment policy.

  • Ron Smith

    Hi, Robert.

    Great article. Thought I might throw in some perspective from my 30 years.

    Most docs have never seen rheumatic fever. I have seen one case. The little girl was not even the patient I was examining. As I was looking at her sister, mom did a ‘say doc’ and asked about her left hand that was tremoring.

    Sure enough the ASO was high and I and the Pediatric cardiologist and Pediatric neurologist independently diagnoses rheumatic fever with neuropathy as the primary manifestations. This young girl is now moved on to adult medicine, but on one of her last visits she told me that she if she goes 2 or 3 months without the injections, the tremors return.

    My uncle died of rheumatic carditis before age 30 some 60 years ago now also. I’m tuned to watch for this.

    The limited diagnostics except for group A beta hemolytic strep is something that I’ve thought about much. There is a clinical gray area that sometimes begs treatment even if the strep test is negative.

    I generally consider treatment in the face of a negative strep test if the symptoms are present for three or more days. The negative strep test only has an accuracy of 93% so I know that by the nature of swab collection, I’m going to miss some true streps.

    Compliance and drug effectiveness are things I weigh as well. I generally am the one that gets contacted when a particular antibiotic fails, so it is not scientific when I say that I think amoxil fails with true strep some 20 to 30 percent of the time. I generally use cephalexin instead, but find equal failures if dosed less than the qid recommendations. It is hard to recommend 40 doses of medicine though with the negative strep test, so in those cases I have been using azithromycin (sp) instead.

    Strep in particular and pharyngitis in general is something where medicine has been a casualty of its success. I try to explain this each time to parents so that they will understand my dilemma when treating. Preventing rheumatic fever still requires vigilance or we will see resurgence just like we have measles and Pertussis.

    Thanks for the great article and further instructive comments would be appreciated.

    Warmest regards,

    Ron Smith, MD
    www (adot) ronsmithmd (adot) com

  • Steve

    Here is the million dollar question- we know that 1 in every 4,000 to 5,000 patients will have fatal anaphylaxis from peniclllin. That’s FATAL anaphylaxis. The question remains is whether the risk of developing peritonsillar abscess (PTA) and Lemierre’s disease is more or less than that. I could find no studies that have shown that antibiotics reduce the rate of PTA. Even if the rate of PTA formation is more frequent than that- is it worth fatal anaphylaxis for a disease that is usually treated with a simple I and D? Lemierre’s is much more serious but does it occur at greater frequency of 1 in 4,000? If the answer is no then we are literally killing more patients than we are helping by treating all pharyngitis with antibiotics.

    • Ron Smith


      I have a hard time with that number based on my experience.

      What is generally overlooked is that every dose of a medication runs the same risk of developing a drug allergy reaction. So there should be no safe harbor in thinking that you have had a particular antibiotic many times before without a reaction. Penicillin seems to have been the primary incriminate in a dominant misconception.

      In thirty years of primary Pediatrics, I’ve seen less than a handful of peritonsillar abscesses. Maybe I’m just an outlier here?

      Warmest regards,

      Ron Smith, MD
      www (adot) ronsmithmd (adot) com

    • Robert Centor

      I know of no data on such a high risk of anaphylaxis.

    • Robert Centor

      According to a reference from Wikipedia – Anaphylaxis to penicillin occurs once in every 2,000 to 10,000 courses
      of treatment, with death occurring in less than one in every 50,000
      courses of treatment. I believe it is much less with oral penicillins than with parenteral penicillins. In one study the army gave 500k recruits oral penicillin without a death.

  • http://briarcroft.wordpress.com/ Emily Gibson

    Another great post about a very common and potentially life threatening clinical dilemma, Dr. Centor. As a college health physician who has to make these decisions several times a day with my patients, I appreciate your advice to actually use clinical judgment based on population vulnerability and a patient signs and symptoms.

    The college health population is particularly prone to tonsillitis complications and warrants special consideration. We use your Centor Criteria routinely in our decision-making about diagnosis and treatment along with use of a complete blood count and mono testing in rapid strep negative individuals with prolonged and/or severe symptoms. White cells with left shifted elevations or the presence of atypical lymphocytes are very helpful in appropriate treatment decision-making.

    At our public regional university of 15,000 students, we see 12-15 peritonsillar abscesses per academic year urgently referred to ENT for drainage and treatment with broad spectrum antibiotics. We have had two cases of Lemierre’s Syndrome in the last two years resulting in ICU admissions, both surviving. Infectious mononucleosis pharyngitis is almost as common as Group A strep, both running 2-3 cases per clinic day out of 130 patient visits.

    We have seen a hand full of acute rheumatic fever cases and post-strep glomerulonephritis in our population during my 25 years doing college health.

    I appreciate your clinical caution that not all protocols fit all populations.

Most Popular