McConaghy and Smith are to be congratulated for their carefully reasoned review of the importance of reducing inappropriate and broad-spectrum antimicrobial treatment of acute otitis media, which concludes that “we should be focusing efforts on reducing indiscriminate use and identifying subsets of children who truly need antibiotics.” We agree wholeheartedly with these sentiments and have published principles of judicious use of antimicrobial agents for otitis media, as they cite.¹

The task of the Drug-Resistant Streptococcus pneumoniae (DRSP) Therapeutic Working Group, however, was different. This group was charged with providing advice for the management of acute otitis media, given an identified subset of children for whom antibiotics were truly needed. In 1999, resistant pneumococci were of primary concern in this subgroup of children, and few clinical, patient-oriented outcome studies were available to guide clinicians in making the best treatment decisions for their patients. In fact, none of the 18 drugs currently labeled for treatment of acute otitis media has gained U.S. Food and Drug Administration approval for use against resistant pneumococci, primarily because the clinical evidence for efficacy against these pathogens is so hard to come by.

Nevertheless, ample evidence shows that resistant pneumococci are of real clinical concern and that adequate treatment is important for good clinical outcomes. Of the major pathogens causing acute otitis media, pneumococci are not only the most common but are the least likely to resolve in the absence of appropriate therapy. For example, 50 percent of Haemophilus influenzae infections will resolve even if the patient is treated with a placebo, but only 20 percent of Streptococcus pneumoniae infections will resolve.² Clinical outcome is in fact well correlated with bacteriologic eradication of pathogens of the middle ear fluid,³ and pneumococcal resistance is directly correlated with clinical failure⁴ and bacteriologic persistence.⁵

What, then, is the family physician to do after careful examination of the child and determination that all criteria for acute otitis media requiring antimicrobial therapy have unequivocally been met? If resistant pneumococci are not a concern, any of the 18 approved drugs might appropriately be selected, but this is no longer the case for most areas of the United States. For pneumococci that are not susceptible to treatment with penicillin, drugs such as cefaclor, loracarbef, cefixime and ceftibuten are inactive and more likely to fail, clinically and bacteriologically.^5,6 Therefore, in the absence of controlled, patient-oriented outcome studies, the DRSP Therapeutic Working Group document was an attempt to provide some guidance for antimicrobial treatment in the era of resistant pneumococci.

Fortunately, amoxicillin remains an excellent first-line choice because it is effective, safe, inexpensive and convenient to administer. The higher dosage provides expanded coverage of resistant pneumococci. McConaghy and Smith ask if this will increase the risk of adverse drug reactions, and the answer is no, so far as we are aware. For those children with clinically documented treatment failures, the most likely pathogens are beta-lactamase– producing H. influenzae and drug-resistant S. pneumoniae. The alternative agents identified are all effective against drug-resistant S. pneumoniae and are beta-lactamase stable. It is true that each has shortcomings for some patients (e.g., amoxicillin-clavulanate is expensive, cefuroxime axetil has a bitter taste and ceftriaxone must be injected). Most of the alternatives, however, are slightly yet measurably more likely to lead to treatment failure.

In summary, we are pleased to see that the DRSP Therapeutic Working Group document has generated such careful thought and criticism and heartily endorse the call for more patient-oriented outcomes studies and publications. We do not agree that “there is little correlation between the Petri dish and how a child responds to antibiotic treatment.” Rather, we hope that the available bacteriologic and clinical efficacy data summarized in the document provide a rational interim approach for treating the subset of children with true acute otitis media.