Comparison of old, professional and consumer model acoustic reflectometers in the detection of middle-ear fluid in children with recurrent acute otitis media or glue ear.

Objective: Acoustic reflectometry is an alternative method of determining the probability of middle-ear fluid. Its sensitivity and specificity are comparable to those of pneumatic otoscopy and tympanometry. The relative superiority between the first-generation model and second-generation spectral gradient acoustic reflectometers (introduced for both professional and consumer use) has not been studied. This study aimed to assess the diagnostic performance of three different acoustic reflectometry devices, and to evaluate the intercorrelation of the results.

Methods: Fifty children (100 ears) aged <7 years undergoing ambulatory otomicroscopic myringotomy in anesthesia were examined with three different acoustic reflectometers, and the results were compared to the myringotomy findings (gold standard method). Specificity, sensitivity and positive and negative predictive values of each instrument in detecting middle-ear fluid were determined. The intercorrelations between the three devices were assessed descriptively (percentage agreement) and statistically (kappa correlation).

Results: Second-generation professional and consumer spectral gradient acoustic reflectometers performed equally well in detecting middle-ear fluid (specificities 84 and 85%, sensitivities 83 and 77%, respectively) and were superior to the old model. Accordingly, the results obtained with professional and consumer models correlated well with each other (kappa 0.60) but not with those of the first-generation device. Acoustic reflectometry is not reliable when performed in anesthesia with the patient lying down.

Conclusions: Newer spectral gradient acoustic reflectometers perform better than the older first-generation model. In this study, the consumer model gave the same information as professional model, but its diagnostic validity in the hands of layman consumers needs to be confirmed.