Test and retest reliability of objective screening tests in neonatal hearing screening program in developing countries.

BACKGROUND

Our study aimed to assess the reliability of screening tests in healthy newborns across diverse hospital environments and neonatal intensive care unit (NICU) infants hospitalized for over 5 days, particularly within the framework of hearing screening in developing countries.

METHOD

The study comprised 100 neonates in each group: G1 (healthy infants in government general wards), G2 (healthy infants in private special wards), and G3 (infants in NICU for over 5 days). Intra-session (within 5 min) and inter-session (within a month) otoacoustic emissions (OAE) and automated auditory brainstem responses (AABR) recordings were conducted and the reliability of each test was evaluated across sessions and groups.

RESULTS

The weighted Kappa results showed poor within-session reliability of OAEs in G1 and G3, while G2 exhibited good reliability. AABR intra-session reliability was consistently good across all three groups. The inter-session reliability of OAEs remained poor in G1 and G3 but better in G2. Significantly, the inter-session reliability for AABR decreased in G1 and G3, with Chi-square analysis revealing a notable referral discrepancy between the initial and final assessments. Such a disparity was absent in G2, where reliability remained high.

DISCUSSION

The study highlights the compromised reliability of OAEs in noisy environments, while AABR maintains good reliability under similar conditions. Additionally, AABR shows poorer reliability when conducted within a specific early timeframe, emphasizing the importance of screening after this period. This issue does not affect OAE, but OAE is limited in detecting mild conductive components. The study highlights the poor reliability of AABRs in the NICU group, attributing this to the potential for improved thresholds over time within this group.