A randomized clinical trial to assess the effects of tympanometry on the diagnosis and treatment of acute otitis media

Academic Article

Abstract

  • Objectives. Acute otitis media (OM) is the most common indication for the use of antibiotics among children. Otoscopy alone is an imprecise method for the diagnosis of OM, which may lead to inappropriate antibiotic use. We sought to determine whether tympanometry, as an adjunct to otoscopy, would significantly change physician prescribing behavior and whether physicians overprescribe antibiotics for OM, using independently graded tympanometry results as a standard. Methods. A randomized, clinical trial was conducted among children 6 to 35 months of age who presented to a pediatric emergency department with either fever or upper respiratory infection symptoms. Children were randomized into 2 groups, in which the attending physician evaluated tympanometry results (Tymp Aware) or the attending physician was blinded to the tympanometry findings (Tymp Unaware). Tympanometry curves were graded independently by using a modified version of the Jerger scale. Results. Of the 698 patients enrolled, tympanometry was performed successfully for 99.3%. Antibiotics were prescribed for OM for 27.9% of all patients. No statistically significant difference in antibiotic prescription rates for OM between the Tymp Aware group (28.8%) and the Tymp Unaware group (26.8%) was found. Of all patients for whom antibiotics were prescribed for OM, 14% had normal curves for both ears and 40% had some tympanographic movement bilaterally. Conclusions. Tympanometry did not seem to change diagnoses or prescribing behavior in the group of physicians studied. Antibiotics were commonly prescribed for presumed OM in the absence of effusions documented with tympanometry.
  • Authors

    Published In

  • Pediatrics  Journal
  • Digital Object Identifier (doi)

    Author List

  • Spiro DM; King WD; Arnold DH; Johnston C; Baldwin S
  • Start Page

  • 177
  • End Page

  • 181
  • Volume

  • 114
  • Issue

  • 1