A Two-Step Otoacoustic Emission Protocol With Brainstem Evoked Response Audiometry For Universal Newborn Hearing Screening: A Prospective Observational Study Of 500 Neonates

Abstract

Background: Congenital hearing loss is an important cause of preventable impairment in speech, language, cognitive and social development. Screening restricted to high-risk neonates may fail to identify a substantial proportion of affected infants. Physiological screening using otoacoustic emissions (OAE) is rapid and suitable for universal screening, but early testing can generate transient referrals; a staged protocol may reduce unnecessary diagnostic testing.
Objective: To determine the incidence of congenital hearing loss among newborns using a two-step distortion-product otoacoustic emission (DPOAE) protocol followed by brainstem evoked response audiometry (BERA) in persistent referrals; to identify high-risk neonates, evaluate risk factors associated with hearing loss, and determine the false-positive referral rate.
Methods: This prospective observational study included 500 term and preterm newborns delivered at our hospitali, between January and October 2016. After parental informed consent, demographic, antenatal, perinatal, postnatal, family-history and otological data were recorded. Bilateral DPOAE screening was performed at 24–72 hours of age. Newborns with a REFER result in either ear were rescreened by DPOAE at 10–15 days. Persistent REFER results underwent diagnostic BERA at 1 month. Descriptive statistics were calculated using SPSS version 17.0; categorical variables were compared using the chi-square test or Fisher exact test, with p<0.05 considered significant.
Results: At the first DPOAE screen, 453/500 newborns (90.6%) passed bilaterally and 47 (9.4%) had a REFER result: 20 bilaterally, 14 in the left ear only, and 13 in the right ear only. On repeat DPOAE, 43 of 47 referrals passed and 4 remained bilateral REFER. All four persistent referrals had bilateral hearing loss on BERA at 1 month, giving an observed incidence of 0.8% (8 per 1000). The first-stage transient referral/false-positive proportion was 43/500 (8.6%), whereas no additional false-positive referral was observed after the second DPOAE among the four infants sent for BERA. High-risk features were present in 116/500 infants (23.2%); all four BERA-confirmed cases occurred in this group (3.4% vs 0% among infants without identified risk factors; p<0.001). Significant associations were reported for low birth weight (p<0.001), NICU stay with ventilation for >5 days (p<0.001), birth asphyxia (p=0.005), hyperbilirubinemia (p=0.011), maternal infection (p=0.010), and increasing burden of multiple risk factors (p<0.001).
Conclusion: A two-step DPOAE strategy followed by BERA for persistent referrals was feasible in this hospital cohort and markedly reduced the referral burden between the first and second screens. Universal screening identified an observed congenital hearing-loss incidence of 0.8%, with all confirmed cases occurring among high-risk neonates. Repeating OAE before diagnostic BERA may reduce transient referrals while retaining a structured pathway for early diagnosis.