Three impedance screening programs on a cohort of seven-year-old children. Can serial impedance screening reveal chronic middle ear disease?

In an unselected cohort of 387 seven-year-old children 90% got all 9 planed tympanometries in the first year at school. The results from these 348 children had been analysed in three different impedance screening models, as 3 consecutive screening series, each comprising two or three tests with an interval of 4 to 8 weeks. The specificity of the ASHA and Nashville models were too low, about 70%, there were too many false positive cases. Therefore a new impedance screening model, was constructed. It had about the same sensitivity (80%) and a fare better specificity (95%) than the ASHA and Nashville models. The predictive value of a positive test was 15% in the ASHA and Nashville models but three times better (48%) in the Hirtshals model. The Hirtshals model had a referral rate of 9% compared with the 32-36% in the ASHA and Nashville models. The frequencies of retesting were 31% in the ASHA model, 52% in the Nashville model and 45% in the Hirtshals model.