Reflections on the AAP/AAFP clinical practice guidelines

April 2005

The AAP and the American Academy of Family Physicians Clinical Practice Guidelines on the Diagnosis and Management of Acute Otitis Media (AOM) have been published and widely disseminated by means of outpatient lectures, pharmaceutical company-sponsored grand rounds, dinner lectures, and other methods.

The guidelines were intended to assist the primary care clinicians by providing a framework for clinical decision-making in the diagnosis and management of AOM. Seven hundred and sixty articles were reviewed and rated for quality of methodology and importance of results (evidence-based medicine).

The guidelines were not intended to establish a protocol for all children with AOM. Excluded from the guidelines were children with AOM who have signs or symptoms of systemic illnesses unrelated to the middle ear disease, those with cleft palate, those with genetic conditions, such as Down syndrome, and those with immunodeficiencies. Also excluded are children with a clinical recurrence of AOM within a 30-day period and children with underlying chronic otitis media with effusion (OME).

A work in progress?

I hope that the guidelines are in evolution and will undergo changes in criteria for diagnosis and recommendations for management within the next few years. My reflections on the guidelines are not meant to be critical of the enormous amount of thought and labor that were expended in this project; they only reflect my current thinking which will be provocative.

The guidelines require the definition of AOM to have signs and symptoms of recent onset middle-ear inflammation. However, by the use of wiggle words such as “any of the following” and “either/or,” in my opinion, the stated goal of precision in diagnosis is rendered indistinct and imprecise. For example, a child with some ear pain who has a retracted tympanic membrane (TM) that is “distinctly” erythematous can fit the definition of AOM. The TM in this case will have limited or no mobility to positive pressure with the pneumatic otoscope. This may cause a feeling of fullness in the middle ear and discomfort or rubbing the auricle or sticking a finger in the ear to try to relieve the negative middle ear pressure. Air-serous fluid levels are frequently found with viral upper respiratory infections. They appear as “hairlines or as thin fluid with bubbles of air or fluid with a meniscus between the air and the serous fluid.” Sometimes the thin fluid has a serum color. I am unaware of any studies that show that this type of air-fluid levels contain pathogenic bacteria, and I am very doubtful that they do. Air-fluid levels with apparent purulent material in the middle ear liquid may contain bacterial pathogens. Watchful waiting for children with this finding should be an acceptable alternative to antibiotic treatment.

I am sad that the committee did not insist that an opacified TM showing fullness (incomplete bulging) or total bulging as a sine qua non of AOM. Acute otorrhea in warmer seasons can be caused by otitis externa and this should be differentiated from middle ear otorrhea. The committee specified that there be symptoms of distinct otalgia that preclude normal activity or sleep, but all of us in primary pediatric practice have seen young children with precise and unmistakable signs of AOM without much pain or any fever.

The diagnosis of AOM should depend on specific tympanic signs, not on symptoms. The decision to prescribe antibiotics should consider the degree of pain and or fever, not the diagnosis of AOM. Although recommendation #2 states that it is based on randomized clinical trials with limitations, I know of no unique study of painless AOM that included tympanocentesis. I am sure all in primary care pediatrics have seen children with opacified bulging tympanic membranes who have minimal pain or even with no apparent pain that still contain middle ear pathogens. To treat or not to treat with antibiotics is a different issue than the correct diagnosis of the middle ear condition.

Although the committee correctly states that bulging of an opacified poorly mobile TM is the best predictor of AOM, the either/or choice weakens the accuracy and precision of the diagnosis. Why is it so difficult to pin the tail on the donkey in the correct location? Uncertainty in the diagnosis received undue prominence. Those who have performed 1,000 or more tympanocentesis and cultures of middle ear fluid find few tympanic membranes that fall under the heading of “uncertain diagnosis.”

Uncertainty should be unusual in the diagnosis of AOM, if the otoscope emits brilliant halogen light, the otoscope speculum is designed with a 5 mm or greater straight portion from the tip, earwax is meticulously removed from the ear canal, the child is cooperative or restrained, and the examiner takes the necessary time to perform a careful examination.

The observation recommendation

Committee Recommendation #3A permits 48to 72-hour observation without use of antibacterial agents in selected children with uncomplicated AOM and prompt re-evaluation or initiation of antibiotic drugs if pain or fever persist or worsen.

I am a proponent of observation or parental option regardless of the age of the child, including children younger than 1 year. I have followed this personal recommendation for several years without mishap, even when the diagnosis of AOM was certain, as long as the infant did not experience severe pain or high fever and the TM was not dark purple or fiery red and bulging. Because of the probability of virulent middle ear pathogens in the above scenarios, I always institute antibiotic therapy promptly.

The 1999 revision of the Dutch College of General Practitioners guidelines for management of AOM permits 48- to 72-hour observation of children of all ages with minimally or moderately symptomatic AOM. Why are we so worried that bad things will occur because the child is younger than 2 years or even younger than 1 and AOM is diagnosed with certainty?

Committee Recommendation #3B narrows the initial antibacterial agent to high-dose amoxicillin if the decision is made to treat AOM with an antibiotic. Because the bacteriology of AOM seems to be evolving away from Streptococcus pneumoniae as the primary middle ear pathogen to Haemophilus influenzae, the initial choice of amoxicillin may be changed within the next few years.

Investigators in Israel, Kentucky, Virginia and possibly other places have noted that 50% of cases of AOM are now caused by H. influenzae, and about 50% of those bacteria are resistant to amoxicillin. Therefore, 25% of all cases of AOM in those locations and probably many others are likely to be resistant to amoxicillin. I still use amoxicillin and I use high-dose (80mg/kg), but I have noted an increasing number of symptomatic failures or recurrences of AOM within the 20-day post-treatment follow-up period. Time will tell.

The diagnosis and management of AOM and OME continue to fascinate me and keep me thinking about these common diseases and how little we really know about permissible management strategies. Comments about this column are solicited and welcomed.

Editorial note: As pointed out by one of the consultants on the guideline panel, Michael Marcy, MD, in the December issue of Infectious Diseases in Children, the guidelines provide options for therapeutic treatment, including the more controversial option to withhold antibiotics in some children and are by no means “a mandate.” Dr. Schwartz’s column reflects a wide-spread split in opinions on these evidence-based guidelines, and as more data emerge on this issue, even more opinions are likely to appear on the pages of this publication. Stay tuned for more. — Richard F. Jacobs, MD

For more information:

• http://aappolicy.aappublications.org/cgi/content/full/pediatrics;113/5/1451?fulltext=otitis&searchid=QID_NOT_SET

• Richard H. Schwartz, MD, is from the department of pediatrics at Inova Fairfax Hospital for Children, Falls Church, Va.