Evaluating a prediction model for infant hearing loss

Judith E.C. Lieu, Felicia Ratnaraj, Banan Ead

Research output: Contribution to journalArticlepeer-review

8 Scopus citations


Objectives/Hypothesis The objective of this study was to determine whether a prognostic model using risk factors for hearing loss could predict the chance that infants who failed a newborn hearing screen would subsequently be found to have hearing loss diagnosed by auditory brainstem response testing. Study Design Individual retrospective case-control study. Methods We studied 229 infants with hearing loss compared with 458 infants with normal hearing. All infants had undergone natural sleep or sedated auditory brainstem response, predominantly for not passing a newborn hearing screen. Risk factors, birth history, and other information were extracted via medical record review. Multiple logistic regression analyses identified independent predictors of hearing loss. Results Four risk factors were independently predictive of hearing loss diagnosed by sleep or sedated auditory brainstem response: prematurity, 5-minute APGAR score ≤ 6, intracranial complication, and craniofacial abnormality. A prognostic model developed from these risk factors was associated with a 15% rate of hearing loss in stage I, 52% rate of hearing loss in stage II, and 96% rate of hearing loss in stage III. Conclusions The presence of any one of four independently predictive risk factors in infants who did not pass newborn hearing screen was associated with a 50% rate of hearing loss; having three or more was associated with a 90% rate of hearing loss. Knowing that an infant is at high risk of hearing loss can motivate parents to follow up with diagnostic auditory brainstem response testing so that early identification can lead to early intervention.

Original languageEnglish
Pages (from-to)2873-2879
Number of pages7
Issue number11
StatePublished - Nov 1 2013


  • Hearing loss
  • auditory brainstem response
  • infant
  • newborn hearing screening
  • prediction
  • risk factors


Dive into the research topics of 'Evaluating a prediction model for infant hearing loss'. Together they form a unique fingerprint.

Cite this