Objective: We present population-based, childhood prevalence rates of and neonatal risk factors for permanent hearing loss among extremely premature infants.
Methods: By using an inception-cohort, longitudinal study design for 1974-2003, we studied permanent hearing loss among 1279 survivors with gestational age of < or =28 weeks and birth weight of <1250 g (mortality rate: 42.7%; lost to follow-up monitoring: 4.7%) Newborn hearing screening, performed by experienced pediatric audiologists, used click-evoked auditory brainstem response testing after 1975. Survivors underwent repeated behavioral audiologic testing and multidisciplinary follow-up monitoring. Permanent hearing loss was defined as mild/moderate (26-70 dB hearing level), severe/profound (71 to >90 dB hearing level), delayed-onset (diagnosed after previously normal hearing), or progressive (increase in loss of > or =15 dB hearing level). Permanent hearing loss rates were established at 3 years of age, with newborn, infant, and >5-year final hearing outcomes being recorded. Risk factors were compared for children with and without hearing loss, odds ratios were calculated, and prediction performance was determined through area under the curve analysis.
Results: Forty (3.1%) of 1279 survivors 3 years of age had permanent hearing loss and 24 (1.9%) had severe/profound loss, with no changes over time. Bilateral delayed-onset loss occurred for 4 children (10%) and progressive loss for 11 children (28%). One child had auditory neuropathy, and 29 (73%) had multiple disabilities. Prolonged oxygen use, gastrointestinal surgery, patent ductus arteriosus ligation, and low socioeconomic index yielded good prediction of permanent hearing loss; oxygen use was the most significant predictor of severe/profound loss.
Conclusions: Permanent hearing loss remains an adverse outcome of extreme prematurity, complicated by significant delayed-onset and progressive loss. Prolonged supplemental oxygen use is a marker for predicting permanent hearing loss; this requires detailed analysis of the pathophysiologic features, to reduce the prevalence of permanent hearing loss.