PubMed2002
BACKGROUND AND OBJECTIVE: To make a rational decision as to which screening test might be adequate as a universal newborn hearing screening, different methods have to be tested under "real-life" conditions. In addition, a good reference is required as "golden standard." PATIENTS/METHODS: In the study presented here, an ABR with a novel algorithm for threshold-estimation was optimized and compared to TEOAE (Echoscreen) and DPOAE (GSI 60) in a three-step protocol using a standard click-evoked ABR (Evoselect) as reference of sensitivity and specificity. RESULTS: 26 ears were found to be hearing-impaired. All of them were also detected by each of the screening method (sensitivity 100%). Specificity in the final "step 2" of the study was around 87.7/92.3% for Echoscreen, 82.4/84.4% for DPOAE and 82.4/89.1 for Evoflash (left/right, respectively). Differences were not at all significant (X2-test, p > > 0.05). In summary, results were somewhat inferior to those found by other authors, probably due to more difficult conditions in the "real-life" setting. CONCLUSIONS: A higher rate of false-alarm shows that a two-step screening is necessary. Therefore, increased resources for diagnostic procedures are required.