381 REVIEW OF DIAGNOSTIC CRITERIA FOR ACUTE OTITIS MEDIA

Background In a classic paper, Hayden (1981) showed that clinicians and researchers used a wide variety of criteria to diagnose acute otitis media (AOM). The American Academy of Pediatrics (AAP) published AOM guidelines (2004) to provide a framework for clinical decision making. These guidelines specified criteria for the diagnosis of AOM. Objectives To assess the consistency of AOM diagnostic criteria used in the literature from 1994-2004, and to assess the agreement with the AAP criteria. Methods We (DPM and SSG) conducted a literature search of clinical trials on AOM in children from 1994-2004. We independently reviewed sixty-two articles from 17 journals. The 51 articles included in the final analysis were clinical trials exclusively focusing on the treatment of uncomplicated AOM in children. We applied a strict interpretation of the three AAP criteria for the diagnosis of AOM: (1) required a specified time of acute onset, (2) required documentation of a middle ear effusion as verified by the presence of bulging, decreased mobility, an air/fluid level, or otorrhea, or an abnormal tympanogram or tympanocentesis, and (3) required the presence of tympanic membrane erythema or otalgia. We also recorded the frequency of other signs and symptoms of AOM assessed, but not necessarily required. Results 25% of studies specified a time for acute onset, 75% required a middle ear effusion, and 27% required erythema and/or otalgia. Of the three AAP criteria, 12% of the articles required all three, 25% required two, 41% required one of the criteria, and 22% required none. The three symptoms most commonly assessed (but not necessarily required) were otalgia (98%), fever (90%), and erythema (82%). The two signs/symptoms least frequently assessed were sleep disturbances (10%) and air-fluid level (6%). Conclusions Clinical investigators have made progress in the use of specific criteria for the diagnosis of AOM. In 1981, 60% of the articles reviewed by Hayden described their criteria for AOM. This compares with our data, in which 100% of the authors described their criteria. However, investigators are still inconsistent in their use of specific criteria for the diagnosis of AOM. Hayden9s 1981 paper described 18 different sets of criteria from 26 different articles; we found, using the AAP guidelines as a standard, that only 6/51 (12%) used all 3 criteria defined by the AAP, 13/51 (25%) required 2 of the criteria, 21/51 (41%) required one, and 11/51 (22%) required none of the criteria described by the AAP. We recommend that investigators increase the validity of their studies by standardizing the criteria they use for the diagnosis of AOM. Criteria should be in agreement with the AAP guidelines.