Secondary Logo

Journal Logo

Tot 10: Monkey (Wrench) in the Middle Evaluating Middle Ear Function in Young Children

Smith, Joanna T. MS; Wolfe, Jace PhD

doi: 10.1097/01.HJ.0000436553.87379.a9

Ms. Smith, left, is cofounder and executive director of Hearts for Hearing in Oklahoma City. Dr. Wolfe is director of audiology at Hearts for Hearing and an adjunct assistant professor at the University of Oklahoma Health Sciences Center and Salus University.



Over the past few years, middle ear measurements have suffered from middle child syndrome, with our profession failing to fully appreciate their role in our bag of tools for assessing tiny tots.



Well, just like a middle child typically overachieves to overcome a perceived lack of attention, the middle ear plays a huge part of every audiology measure we use in the clinic. In addition, middle ear measurement is arguably more complex than its sibling measurements.

Before we start on the ten pearls audiologists should keep in mind when evaluating middle ear function in infants and young children, we would like to acknowledge Robert Margolis, Doug Keefe, James Jerger, Lisa Hunter, Navid Shahnaz, Stephen Painton, Jay Hall, and Joseph Kei for helping us reconcile the complex physics of acoustic immittance in infants.

Back to Top | Article Outline


You are probably well aware that the middle ear measurement protocols we use for adults are not appropriate for infants. This is because of significant differences in the anatomy and physiology of the external and middle ears during the first few months of life.

First, the outer third of the adult external ear canal is made of cartilage, while the inner two-thirds is comprised of bone. In contrast, the entire infant external ear canal consists of soft cartilaginous tissue.

In addition, Piza and colleagues reported that every infant temporal bone they studied possessed mesenchyme in the middle ear space (Laryngoscope 1996;106[7]:856-864"). Other researchers have suggested an increased likelihood of middle ear effusion in infants (Laryngoscope 1993;103[1 pt 2 suppl 58]:1-31). The presence of mesenchyme and effusion would definitely mass-load the middle ear system and reduce its resonant frequency, and it also may increase the resistance of the system.

Furthermore, the tympanic membrane of infants is likely thicker than the adult tympanic membrane, which also may contribute to a greater mass effect.

Back to Top | Article Outline




The aforementioned physical differences in adult and infant ear anatomy affect tympanometry in a handful of ways. We should start by noting that researchers are still not entirely certain of the exact mechanisms responsible for the differences observed between adults’ and infants’ tympanograms.

When highly positive or negative pressure is introduced into the ear canal, the difference in pressure between the external and middle ear spaces is substantial, and the eardrum is “pushed into” or “pulled out” of the middle ear space, respectively. This essentially results in the eardrum acting as a “wall” to the passage of sound so that the probe tone remains within the hard, bony walls of the external ear canal.

In contrast, an infant's cartilaginous ear canal wall is compliant and will likely expand in response to the positive pressure changes introduced into the ear canal at the beginning of the tympanometric measurement.

Second, the presence of increased mass in a system will decrease the resonant frequency of that system. The typical adult middle ear system has an input resonant frequency between 800 and 1,200 Hz at the eardrum, while the infant middle ear has a primary resonance around 450 Hz, with a second resonance near 700 Hz. This difference in middle ear resonance is also responsible for differences in an infant's tympanometric results.

Back to Top | Article Outline


The overwhelming consensus among researchers and clinicians calls for the use of a 1,000-Hz probe tone for tympanometry in infants, as various reports have described erroneous or complex findings with 226-Hz tympanometry in these patients.

To start, Keefe et al showed that transmission of sound into the middle ear of infants was relatively inefficient at 220 and 660 Hz, which the researchers attributed to the proximity of these frequencies to the frequency of vibrations of the infant's cartilaginous ear canal wall (J Acous Soc Am 1993;94[5]:2617-2638 Sounds in the 1,000- to 4,000-Hz range were most efficiently transmitted into the infant middle ear.

Additionally, many researchers who have studied 226-Hz tympanometry in infants have reported notched tympanograms, which are complicated to interpret (J Am Acad Audiol 2003;14[1]:20-28; Arch Otolaryngol 1973;97[6]:465-467).

As a good rule of thumb, the clinician should consider using a 1,000-Hz probe tone through the first two months, both a 226- and 1,000-Hz probe tone from two months to six months, and a 226-Hz tympanogram from 7 months of age and beyond.

The same pitfalls that exist for 226-Hz tympanometry with neonates also exist with the use of a 226-Hz probe tone for acoustic immittance assessment. We use a 1,000-Hz probe tone for acoustic reflex assessment through 6 months of age and a 226-Hz probe tone thereafter.

Back to Top | Article Outline


Interestingly enough, a 1,000-Hz probe-tone admittance tympanogram from an infant with normal middle ear function typically looks like a “normal” 226-Hz probe tympanogram from an adult (see the figure on page 24).

The most salient characteristic is a peak around 0 daPa that closely resembles a type A tympanogram in the Jerger classification system for adult tympanograms and is consistent with normal eardrum movement in infants.

If you were awake in graduate school and can still recall your professor's lecture on multiple component, multiple frequency tympanometry, then you may be asking why the infant's admittance tympanogram does not typically notch to a 1,000-Hz probe tone. We are sorry to say that as a field, we simply do not understand enough about the developing middle ear to have a clear-cut answer.

Back to Top | Article Outline


Well, it would be great to use admittance tympanometry to estimate the size of the infant ear canal, but the measurement is complicated when using 1,000-Hz tympanometry.

Remember that one of the virtues of using a 226-Hz probe tone is that 1 mmho of compliance is equal to 1 cc of cavity size when the driving frequency is 226 Hz. This one-to-one relationship does not exist at other probe tone frequencies.

Since the external and middle ear of an infant begin to become more adultlike around 6 months of age, the clinician may obtain a valid estimate of ear canal volume using 226-Hz admittance tympanometry once the infant reaches 6 to 7 months.

Back to Top | Article Outline


The acoustic reflex test may very well be one of the most underappreciated measures in all of pediatric audiology. It is our most sensitive measure of conductive pathology.

Research has shown that we can only record the acoustic reflex 20 percent of the time when the average air–bone gap is 10 dB (Arch Otolaryngol 1974;99[3]:165-171 Additionally, Kei has suggested that the presence of acoustic reflexes essentially rules out middle ear effusion (Assessing Middle Ear Function in Infants [San Diego: Plural Publishing, 2012]

A collective finding of present acoustic reflexes and absent otoacoustic emissions suggests that cochlear dysfunction, rather than middle ear pathology, is the root cause for abnormal auditory dysfunction.

Furthermore, Berlin and colleagues showed that acoustic reflexes were typically absent—and, if present, always significantly elevated—in all patients with auditory neuropathy spectrum disorder (J Am Acad Audiol 2005;16[8]:546-553

Back to Top | Article Outline


The following points are pearls for the clinician to keep in mind when testing acoustic reflexes in neonates:

  • Keep the probe tone frequency high. The most accurate acoustic reflex results are obtained with a 1,000-Hz probe tone.
  • Broadband noise and pure tones may both be used to elicit a reflex response in infants. The range of normal response thresholds for a broadband noise (BBN) stimulus in infants is 50 to 80 dB HL.

A 1,000-Hz eliciting tone should not be used because of the potential for interaction and subsequent artifacts between the probe and eliciting tones. Acoustic reflexes are typically present between 60 and 90 dB HL for eliciting tones of 500 to 2,000 Hz. As such, a pure tone-elicited reflex threshold above 95 dB HL exceeds normal limits for infants.

  • Beware excessive levels! A good rule of thumb is to avoid a presentation level above 105 dB HL. For this reason, broadband noise may be preferable, as acoustic reflex thresholds to BBN are generally lower compared with tones, and the energy is spread across the basilar membrane, reducing the chances of noise-induced hearing loss.
  • Responses may be upwardly mobile. Acoustic reflex response traces in infants are occasionally different from those observed in adults. Instead of seeing the characteristic downward decrease in admittance that is time locked to the stimulus, we may see an upward deflection.
Back to Top | Article Outline




OK, so we realize that nobody gets fired up about norms and calibration, but we've got to hit the high points. Let's get calibration out of the way first.

For now, clinicians should adhere to the American National Standards Institute (ANSI) guideline for calibrating acoustic immittance units (ANSI S3.39-1987). In the future, however, changes may be in order.

First, the typical level of the acoustic immittance probe tone is 85 dB SPL. For 226-Hz probe-tone tympanometry in adults, this level is unlikely to trigger the acoustic reflex. However, a 1,000-Hz probe tone presented at 85 dB SPL will likely trigger the acoustic reflex in most infants with normal auditory function.

As such, future calibration protocols and the manufacturers of immittance equipment may consider the use of a lower level probe tone (e.g., 75 dB SPL) for infant assessment.

Furthermore, the current standard calls for the use of a 226-Hz tone during calibration of acoustic immittance equipment. It is possible that alternative calibration techniques may be needed for high-frequency tympanometry.

Here are the research-based normative values we use at Hearts for Hearing for acoustic immittance with infants and children (1,000-Hz probe tone) from birth to 6 months: a single-peaked tympanogram with maximum amplitude near 0 daPa and acoustic reflex to broadband noise not exceeding 80 dB HL, with tonal reflexes no greater than 95 and 85 dB HL for eliciting tones of 500 and 2000 Hz, respectively.

Back to Top | Article Outline


Without a doubt, wideband acoustic reflectance measures may someday have a routine place in the battery of tools used to evaluate middle ear function in infants and young children.

Recent research has suggested that wideband reflectance may be used to identify middle ear effusion, distinguish between various disorders, and estimate the extent of the air–bone gap (Int J Audiol 2012;51[12]:880-891

Since wideband reflectance does not require the introduction of air pressure into the ear canal, it can be achieved with the same instrumentation used for otoacoustic emissions.

Back to Top | Article Outline


In closing, parents must be certain to shower the middle child with love. Likewise, pediatric audiologists must show the love and incorporate gold-standard middle ear assessment into their diagnostic test batteries.

© 2013 by Lippincott Williams & Wilkins, Inc.