Are the tonsils and adenoids a reservoir of infection in otitis media with effusion (glue ear)?

In a prospective investigation of the treatment of glue ear in children, the possible role of persistent infection in the tonsils and adenoids was assessed by comparing bacterial cultures of swabs and resected tissue from the tonsils and adenoids of patients with cultures of swabs from similar sites in control subjects without ENT abnormality. For almost all potential pathogens, including Streptococcus pneumoniae and Haemophilus influenzae, no statistically significant differences were demonstrated between patients and controls. The same was true of Streptococcus pyogenes in tonsil specimens, while in adenoid specimens rather more isolates were obtained in patients than controls. This difference was just statistically significant at the 5% level, but only when all isolations, including very scanty growths, were compared. On present evidence, persistent infection should not be invoked to explain the success of adenoidectomy in otitis media with effusion or to justify adenotonsillectomy.