Objective: Universal newborn hearing screening has become standard practice in many countries. The primary goal of this study was to assess the impact of early identification of permanent childhood hearing loss on oral communication development.
Setting: Participants were recruited from three clinical programmes in two cities in the province of Ontario, Canada. The study sample was born during two consecutive periods of newborn hearing screening. The first period, prior to 2002, was targeted on high-risk infants only, and the second, from 2002, included both high- and standard-risk infants (universal newborn hearing screening - UNHS). All children were enrolled in rehabilitation programmes focused on oral language development.
Methods: In this multicentre observational study, 65 children under the age of five years with onset of hearing loss before six months of age, 26 identified through systematic newborn screening (14 through targeted screening and 12 through UNHS) and 39 without screening, were assessed with an extensive battery of child- and parent-administered speech and language measures. The degree of hearing loss ranged from mild to profound with 22 children in the mild, moderate and moderately severe categories and 43 in the severe and profound categories. Data are reported for the three-year study period.
Results: The screened group of children was identified at a median age of 6.6 (interquartile range, 3.0-8.2) months and children referred from sources other than newborn screening were diagnosed at a median age of 16.5 (interquartile range, 10.2-29.0) months. Assessment of oral communication development showed no significant difference between the screened and unscreened groups. The communication outcomes for children identified before 12 months of age did not differ from those of later identified children.
Conclusions: Systematic screening of newborn hearing results in earlier identification and intervention for children with permanent hearing loss. Superior language outcome following newborn screening was not demonstrable in the setting of this study.