News Release

Common hearing test for newborns may be wrong up to a third of the time

Peer-Reviewed Publication

Michigan Medicine - University of Michigan

U-M study may help hospitals choose equipment for screening, keep new parents from worrying

ANN ARBOR, MI - A test commonly used to screen newborns for hearing problems may give false positives in as many as 35 percent of cases, a new University of Michigan study finds. Besides causing new parents to worry needlessly about their child's hearing, the technique may also lead to more costs in the long run for follow-up tests. By contrast, the study finds, a slightly more complex test seems to give normal-hearing babies a passing grade nearly all the time.

The new finding, published in the December issue of the American Journal of Audiology, may help hospitals choose which test to use as they try to meet some states' requirements that all newborns be screened for hearing loss before going home. In fact, it has already led the U-M to pick the more accurate test for its new universal screening program. The data may also help audiologists fine-tune their use of the quicker test to give the fewest false positive results.

The U-M study directly compared the results of distortion product otoacoustic emission (DPOAE) tests, analyzed using four different "pass" criteria, with those from auditory brain stem response (ABR) tests. The four different criteria were used because there are currently no universally accepted standards for a "pass" in DPOAE testing. Data came from 596 babies born at the U-M Health System who had no hearing-loss risk factors. All were tested with both techniques within 24 hours of birth.

All the infants passed the ABR test in both ears. But in the same group of infants, DPOAE tests performed with the strictest passing requirements suggested that 35 percent of the babies tested would need to return for follow-up testing to rule out hearing loss. Even when the bar was lowered, 11 percent of infants still failed the DPOAE test.

"These are both good tests, and both have important applications, but if we're going to do universal newborn screening, we should do it right," says author Paul Kileny, Ph.D., director of Audiology and Electrophysiology and professor in the Department of Otolaryngology. "These data show that ABR is more accurate and results in far fewer false positives. Why have parents on pins and needles for weeks when you can clear their normal-hearing baby on day one?"

The carefully designed study was not funded by any manufacturer of hearing screening equipment. Support came from the private Carls Foundation, which funds pediatric research.

As many as 24,000 babies are born each year with some level of hearing loss, at a rate of one to three in every 1,000 births that makes it the most common disability present at the start of life. Thousands more babies and toddlers develop hearing loss in the first year or two of life.

Catching those hearing deficits early requires specialized testing, but it's crucial, Kileny says. Babies whose problem is overlooked for years have a much harder time acquiring language and speech skills. But those who get hearing aids or cochlear implants early because of an in-hospital test or an alert pediatrician can often progress normally.

As a result, a 1993 National Institutes of Health report and a 1995 joint commission endorsed universal newborn screening. More than 30 states have since called for mandatory screening. The new mandates have prompted many hospitals to start programs, though they often have little or no extra funding for the equipment and trained staff needed to screen infants before their short hospital stays are over. Even in states like Michigan where no mandate has been passed, hospitals like UMHS have voluntarily started programs for both high-risk and low-risk babies.

Kileny and his colleagues launched their study in part to evaluate which test UMHS should use. Already, they had shown in a 1999 study that ABR tests could spot hearing loss reliably and cost-effectively in infants whose family history, birth conditions or medical problems put them at risk for hearing deficits. In the new trial, they looked at babies with no special risks. All the babies were tested while lying quietly in their mothers' hospital rooms.

The two approaches they tried, DPOAE and ABR, together account for most of the hearing tests done on newborns in the U.S. Both are also used to test older children and adults. DPOAE was invented in the late 1970s, by which time ABR had been established as the "gold standard".

The two techniques differ in their basic underlying principle. DPOAE measures sounds made in the inner ear in response to two simultaneous tones of slightly different frequency fed into the ear via a microphone, while ABR plays clicks into the ear and measures the brain's response through electrodes on the head. Another kind of OAE, transient-evoked, is also available.

DPOAE can be done with sounds of many frequencies, and patients are often passed or failed depending on how many they respond to. But there is no set standard for a "passing" score. Kileny and his colleagues used three different frequencies, including the low frequency of 2000 Hertz, as well as 3200 and 4000 Hz, for the DPOAE test. They performed each test twice. The four passing criteria, in descending order of stringency, were: repeated response to 2000 Hz and one other frequency, repeated response to 3200 and 4000 Hz, non-repeated response to 2000 Hz and one other, and non-repeated response to 3200 and 4000 Hz.

With the strictest standard applied, 35 percent of ears did not pass. With the second most strict, it was 28.5 percent, and with the third strictest standard, it was 16.5 percent. The least stringent standard yielded only 11 percent of ears not passing. With ABR, all ears passed.

"This result gives more reason to develop a universally acceptable standard for DPOAE, and highlights the difference between ABR and DPOAE," says Kileny. "Hospitals should consider this when choosing their preferred testing method, and not just the up-front cost of the equipment. The potential cost of false positives may be much greater, because of the need to re-test relatively large numbers of babies."

He adds, "Good follow-up is a key component of a successful hearing screening program, but it may be somewhat jeopardized by having an unnecessarily high need for further testing."

###


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.