Can risk factors for bilateral hearing loss (HL) be used to predict HL among infants who do not pass the initial newborn hearing screening?
Background: Children with HL often have delayed academic achievement, lower average reading ability and lower-than-average cognitive test performance. Risk factors are used to help identify infants who pass the initial newborn hearing screening (NHS) in the United States but are at risk for developing delayed-onset HL, or those who may have milder forms of permanent HL.
Explore this issue:
November 2013Study design: Individual retrospective case-control study using medical records of 687 infants (229 HL, 458 control) aged zero to six months of age referred for natural sleep or sedated auditory brainstem response (ABR) in 2005 through 2009.
Setting: Audiology department, St. Louis Children’s Hospital, BJC HealthCare system.
Synopsis: Two children with normal hearing were identified as controls for each one child diagnosed with HL. The most common reason for diagnostic ABR testing was not passing NHS. Mean birth weight and estimated gestational age (EGA) were lower in infants with HL, and more infants with HL had been admitted to a NICU. The ear examination suggested middle ear effusion in 76 HL cases, and nine HL cases had atresia in at least one ear; 65 HL cases had prolonged latencies on ABR. Fifteen HL cases were diagnosed with conductive HL. Congenital cytomegalovirus infection, hyperbilirubinemia, eye problems and Down syndrome were not associated with HL. Among infants who received diagnostic ABR for other reasons, there was an increased rate of several risk factors compared with infants who had not passed NHS. Having none of the risk factors was associated with a low HL risk; having three or more was associated with an extremely high HL risk. Having even one factor raised the HL probability to 50 percent. Limitations included a lack of differentiation in hearing loss between temporary conductive HL, permanent HL and auditory neuropathy spectrum disorder.
Bottom line: The presence of any one of four independently predictive risk factors was associated with a 50 percent hearing loss rate; having three or more was associated with a 90 percent rate of hearing loss.
Citation: Lieu JE, Ratnaraj F, Ead B. Evaluating a prediction model for infant hearing loss. Laryngoscope. 2013;123:2873-2879.
—Reviewed by Amy Eckner