A cross-national study of acute otitis media: risk factors, severity, and treatment at initial visit. Report from the International Primary Care Network (IPCN) and the Ambulatory Sentinel Practice Network (ASPN).

BACKGROUND Treatment of acute otitis media (AOM) differs worldwide. The Dutch avoid antimicrobials unless fever and pain persist; the British use them for 5 to 7 days, and Americans use them for 10 days. If effects of therapies are to be compared, it is necessary to evaluate rates of risk factors, severity of attacks, and their influence on treatment decisions. We wanted to compare the prevalence of risk factors for AOM and evaluate their association with severity of attacks and of severity with antimicrobial treatment.

METHODS We undertook a prospective cohort study of 2,165 patients with AOM enrolled by primary care physicians; 895 were enrolled from North America, 571 were enrolled from the United Kingdom, and 699 were enrolled from The Netherlands. The literature was searched using the key words "acute otitis media," "severity," and "international comparisons."

RESULTS The prevalence of several AOM risk factors differs significantly among patients from the three country networks; these factors include race, parent smoking habits, previous episodes, previous episodes without a physician visit, tonsillectomy or adenoidectomy, frequency of upper respiratory tract infections, day care, and recumbent bottle-feeding. Dutch children have the most severe attacks as defined by fever, ear discharge, decreased hearing during the previous week, and moderate or severe ear pain. In country-adjusted univariate analyses, increasing age, exposure to tobacco smoke, day care, previous attacks of AOM, previous attacks without physician care, past prophylactic antimicrobials, ear tubes, adenoidectomy, and tonsillectomy all contribute to severity. Only country network, age, history of AOM, previous episode without physician care, and history of adenoidectomy and tympanostomy tubes are independently related to increased severity, while current breast-feeding is protective. Severity of attacks influences treatment decisions. Dutch children are least likely to receive antimicrobials, and even for severe attacks the British and Dutch physicians usually use amoxicillin or trimethoprim-sulfa; North American children with severe attacks are more likely to receive a broad-spectrum second-line antimicrobial.

CONCLUSION Dutch children have the highest ratings in all severity measures, possibly reflecting parental decisions about care seeking for earaches. When comparing groups of patients with AOM, it is necessary to adjust for baseline characteristics. Severity of episode affects physician treatment decisions. Adoption of Dutch guidelines restricting use of antimicrobials for AOM in the United States could result in annual savings of about $185 million.