Context: There is limited information on the identity and antibiotic susceptibility of bacterial pathogens in children with chronic otitis media whose repeated antibiotic use may place them at increased risk of antibiotic-resistant bacteria.
Objective: To determine, at myringotomy for tympanostomy tube placement, (1) the prevalence of bacteria, (2) the extent and patterns of antibiotic resistance, and (3) the risk factors associated with the presence and resistant status of pathogens.
Design: Prospective, multi-site, cohort study.
Setting and patients: Children undergoing myringotomy for tympanostomy tube placement between November 1, 1999 and March 31, 2000 in seven hospitals in Toronto, Ontario, were identified. If fluid was present, aspirates were submitted for bacteriologic testing. A follow-up telephone questionnaire was administered to patient caregivers in order to identify risk factors for the presence of (1) culturable pathogens and (2) resistant pathogens.
Main outcome measures: The identification and prevalence of bacteria cultured from the middle ears of subjects, and the degree of nonsusceptibility to commonly prescribed antibiotics.
Results: Among 601 patients (mean age 3.9 years, 60.7% male), both a telephone interview (n=544) and an ear specimen (n=527) were obtained for 478. Pathogens were found in middle ear effusions of 37% of the children in the study; including at least one 'definite' pathogen in 189 children (31.4%), and a further 32 children (5.3%) with at least one 'possible' pathogen. Definite pathogens included Haemophilus influenzae in 17% of the children, followed by Moraxella catarrhalis (9%) and Streptococcus pneumoniae (6%); ampicillin nonsusceptibility was found in 40, 100 and 24%, respectively. Overall, 123 children (20.5%) were found to have definite pathogens with resistance to ampicillin/penicillin, trimethoprim-sulfamethoxazole, or clarithromycin/erythromycin. Patient characteristics included premature birth and/or long length of stay in the nursery (23%), first infection before the age of 6 months (26%), put to bed with a bottle (28%), household smoker (34%), in out-of-home child care (38%), history of eczema, bronchiolitis and/or asthma (39%), and use of pacifiers (40%). Household characteristics were smoking (34%), married/common law parents (85%), and 60% had completed college or university; in 26% both parents were born outside of Canada; 73% of children were Caucasian. Of the 75% who responded to the question regarding income, 42% had household income over $60,000 (CAN). Risk factors for the presence of a pathogen and for a resistant pathogen in multivariate analysis included younger age, lower maternal education, day care centre attendance, no previous adenoidectomy and bilateral, primarily winter infections as well as amoxicillin use in the previous 6 months.
Conclusion: Modifiable risk factors for otitis media including household smoking and pacifier use are present in many children undergoing tympanostomy tube placement; child care centre attendees are over-represented. Multiple antibiotic courses were commonly prescribed prior to surgery. H. influenzae and M. catarrhalis are important pathogens and therapy in clinical failures should be directed against them. The 7-valent protein conjugate polysaccharide vaccine (Prevnar) would have covered 73% of the serotypes of S. pneumoniae isolated in this study.