Tympanometry is a measure of the acoustic admittance or ease with which acoustic energy flows into the middle ear transmission system as air pressure is varied in the ear canal. This measure is accomplished by sealing a small probe device into the ear canal. A speaker delivers a probe signal, typically 226 Hz, into the ear canal, and a microphone measures the amplitude and phase of the probe signal admitted into the middle ear system. The acoustic admittance is determined by the combined stiffness (or conversely, compliance), mass, and resistance of the eardrum and all middle ear structures. In the presence of middle ear pathology, these admittance characteristics are altered, and therefore the amplitude and phase of the probe signal measured in the ear canal are also altered. In pathology such as middle ear effusion, the eardrum is stiffened by fluid in the middle ear cavity, and only minimal acoustic energy from the probe signal is admitted into the middle ear; acoustic admittance in the plane of the eardrum is described as low. In contrast, pathology such as ossicular discontinuity makes the ear less stiff, so that most of the acoustic energy from the probe signal is admitted into the middle ear system, and acoustic admittance is high.

In addition to the loudspeaker and microphone, the probe system is connected to a pneumatic pump that adjusts ear canal pressure over a range from −600 to +400 daPa. The dekapascal (daPa) is the unit of pressure that has replaced mm H₂O (ANSI S3.39-1987). The two units, however, are nearly interchangeable (1 daPa = 1.02 mm H₂O).

Tympanometry became a routine clinical procedure following the landmark paper of Jerger (1970). Jerger identified three basic tympanogram shapes. A tympanogram is a graphic display of acoustic admittance measured as a function of changing ear canal pressure. A normal type A tympanogram is shown in Figure 1. Introduction of extreme pressures into the sealed ear canal stiffens the eardrum, and theoretically, all of the acoustic energy from the probe signal is reflected at the surface of the eardrum, and admittance reaches a minimum. Acoustic admittance gradually increases to a maximum, and the probe signal becomes most audible, when the pressure in the ear canal equals the pressure in the middle ear cavity. When the eustachian tube is functioning normally, atmospheric pressure of 0 daPa is maintained in the middle ear cavity, and tympanogram peak pressure (TPP) also is 0 daPa. The ear canal pressure producing peak admittance, therefore, provides an estimate of middle ear pressure. When the eardrum is retracted and negative middle ear pressure exists, the peak of the tympanograms shifts to a corresponding negative value. This tympanogram pattern is designated type C in Figure 1. The third tympanogram, designated type B, has no discernible peak and is flat. This pattern is recorded from ears with middle ear effusion (MEE), perforated eardrums, or patent pressure-equalization tubes (PET).

Figure 1..  

Three patterns of tympanograms recorded using a 226 Hz probe signal. Type A is normal, type B is flat, and type C has a negative tympanogram peak pressure.

Tympanogram shape also has been quantified in an attempt to aid in the diagnosis of middle ear disease and to establish objective criteria for medical referral. Four commonly used calculations are depicted in Figure 2. The first, acoustic equivalent volume (V_ea), is an estimate of the ear canal volume between the probe device and the eardrum. This estimate typically is made using a 226-Hz probe signal and an ear canal pressure of 200 daPa. When the probe device is sealed in the ear canal, the measured acoustic admittance reflects the combined effects of the ear canal and the middle ear. Under extreme ear canal pressures, however, the eardrum theoretically becomes so stiff that acoustic admittance into the middle ear decreases to 0 mmhos. The admittance measured at extreme pressures then is attributed solely to the ear canal volume. When a 226-Hz probe signal is used, the acoustic admittance measured at 200 daPa is equal to the volume of the ear canal. In Figure 2, V_ea equals 0.6 cm³. In children less than 7 years old, V_ea ranges from 0.3 to 0.9 cm³ (Margolis and Heller, 1987; Shanks et al., 1992). In adults, V_ea averages 1.3 cm³ in women and 1.5 cm³ in men (Wiley et al., 1996). As a subsequent example will demonstrate, V_ea is useful in differentiating between intact and perforated eardrums when a flat type B tympanogram is recorded.

Figure 2..  

Four calculations made on 226 Hz tympanograms: acoustic equivalent volume (V_ea, in cm³), peak compensated static acoustic admittance (Y_tm, in acoustic mmhos), tympanogram width (TW, in daPa), and tympanogram peak pressure (TPP, in daPa).

Peak compensated static acoustic admittance (Y_tm) is the amplitude of the tympanogram between the peak and 200 daPa. This measure describes the acoustic admittance of the middle ear transmission system compensated for or minus the effects of the ear canal volume. In Figure 2, Y_tm is 1.1 acoustic mmhos, calculated as peak admittance (1.7 mmhos) minus ear canal admittance (0.6 mmhos). Many instruments “baseline correct” at 200 daPa, so that Y_tm can be read directly from the y-axis. If the tympanogram in Figure 2 were baseline corrected, zero admittance would be shifted upward to correspond with the V_ea at 200 daPa. If a middle ear problem produces abnormally high stiffness, the amplitude of the tympanogram, or Y_tm, will decrease. Conversely, if a middle ear problem decreases the stiffness of the eardrum or middle ear, Y_tm will increase. Y_tm at 226 Hz normally increases slightly from infancy to adulthood, with a mean of 0.5 acoustic mmhos at 4 months to 0.7 acoustic mmhos in adulthood (Margolis and Heller, 1987; Holte, Margolis, and Cavanaugh, 1991; Roush et al., 1995; Wiley et al., 1996).

Tympanogram width (TW), defined as the width in daPa at one-half Y_tm, is a measure of the broadness of a tympanogram peak. In Figure 2, TW is 85 daPa. TW is not highly correlated with Y_tm, and therefore it provides supplemental information regarding middle ear function (Koebsell and Margolis, 1986). TW has been most useful in identifying children with middle ear effusion (MEE). In some cases of MEE, Y_tm is normal but TW is abnormally broad. Nozza et al. (1992, 1994) reported that a TW greater than 275 daPa was associated with a high sensitivity (81%) and specificity (82%) in identifying MEE.

The fourth measure, TPP, provides an estimate of middle ear pressure or indirect measure of eustachian tube function. Figure 2 shows a normal TPP of 10 daPa. Not all individuals with negative middle ear pressure develop MEE. Results from school screening programs showed that medical referral on the basis of TPP alone resulted in unacceptably high overreferral rates, and therefore TPP is no longer used in referral criteria. A negative TPP in conjunction with a reduced Y_tm is a much stronger indication of MEE and cause for medical referral (Feldman, 1976).

Table 1 shows means and 90% normal ranges for V_ea, Y_tm, TW, and TPP from several large-scale studies in subjects ages 8 weeks to 90 years. These calculations are significantly affected by the procedures (e.g., rate and direction of pressure changes and the pressure used to estimate V_ea) used to record the tympanogram (Shanks and Wilson, 1986). The data presented in Table 1 were calculated from tympanograms recorded using the most commonly used parameters, descending pressure changes at rates of 200–600 daPa/s and correction for ear canal volume at 200 daPa.

The remaining figures depict tympanometry findings for a variety of middle ear pathologies. The probe signal frequency most commonly used to measure the admittance properties of the middle ear is 226 Hz. Although this low-frequency probe signal was selected partly at random during instrument development (Terkildsen and Scott Nielson, 1960), it remains the most commonly used probe signal. Acoustic admittance measurements at low frequencies are dominated by the stiffness characteristics of the eardrum and middle ear transmission system, whereas measurements made at high frequencies are dominated by mass characteristics. Although high-frequency probe signals of 660–1000 Hz are valuable in assessing the mass characteristics of the middle ear, the tympanogram patterns that result at high frequencies are more complex and have not enjoyed widespread use. Only low-frequency tympanograms are presented in subsequent examples, but cases where high-frequency probe signals might be advantageous are pointed out. Additional references on high-frequency tympanometry are provided.

Figure 3 shows a series of tympanograms recorded from a child recovering from a 3-month episode of otitis media with MEE. When first evaluated, the admittance tympanogram was flat (type B), with a normal V_ea of 0.45 mmhos. Sequential pure-tone audiograms showed air–bone gaps across all frequencies, ranging from 10 dB to 55 dB, that were greatest at 250 and 4000 Hz and smallest at 2000 Hz. Over time, the tympanogram changed to a type C pattern. In early recovery, the tympanogram, shown by the heavy line, had a shallow (Y_tm = 0.25 mmhos), broad peak (TW = 200 daPa) with negative peak pressure (TPP = –100 daPa). Air–bone gaps decreased to 10–25 dB and were largest at 4000 Hz. This tympanogram pattern has been demonstrated in human temporal bones injected with middle ear fluid up to the level of the umbo, producing a mass loading effect on the eardrum (Renvall, Liden, and Bjorkman, 1975). The mass effect is greatest at high frequencies, as reflected by large air–bone gaps at 4000 Hz, and is accentuated when tympanometry is performed using high-frequency probe signals such as 600–800 Hz. Further resolution of the otitis media produced a type C tympanogram, increasing to normal Y_tm (0.35 mmhos) with a TPP of –200 daPa. Small air–bone gaps of 15–20 dB at this time were confined to the 250–1000 Hz range and were virtually closed at 4000 Hz, indicating increased stiffness of the eardrum from the negative middle ear pressure without the mass loading effects of the middle ear fluid. The tympanogram gradually returned to a type A. This case study demonstrates a variety of tympanogram patterns associated with otitis media. Rather than being a drawback, the various tympanogram shapes help clinicians track the resolution of MEE.

Figure 3..  

Type B and two type C acoustic admittance tympanograms recorded using a 226 Hz probe signal from a child during recovery from otitis media with effusion.

The American Speech-Language-Hearing Association (1997) has developed guidelines for screening infants and children for chronic middle ear disorders with the potential for causing significant hearing loss or long-lasting speech, language, and learning deficits. Medical referral is advised for infants when Y_tm is less than 0.2 mmhos or TW is greater than 235 daPa, and for 1-to 5-year-olds when Y_tm is less than 0.3 mmhos or TW is greater than 200 daPa if these abnormal findings persist at a 6–8-week rescreening. Immediate medical referral is recommended for otalgia, otorrhea, or eardrum perforation noted otoscopically or from a flat tympanogram with V_ea greater than 1.0 cm³. Screening guidelines are not available for infants less than 7 months old. In this age group, tympanogram shapes at 226 Hz are irregular and difficult to interpret (Holte, Margolis, and Cavanaugh, 1991).

Figure 4 displays three type B tympanograms. The bottom tympanogram was recorded from an ear with an intact eardrum and MEE; V_ea of 0.45 cm³ is normal for a child's ear canal. The other two tympanograms also are flat but V_ea is 3.25 cm³ in one case and greater than 5.0 cm³ in the other. The middle tympanogram was recorded from an ear with a patent PET, and the top tympanogram was recorded from an ear with a traumatic perforation from a Q-tip. V_ea often is larger with a traumatic eardrum perforation than with a perforation associated with chronic middle ear disease and poorly developed mastoid air-cell system (Andreasson, 1977).

Figure 4..  

Type B acoustic admittance tympanograms recorded using a 226 Hz probe signal from an ear with middle ear effusion (bottom tympanogram), an ear with a patent pressure equalization tube (middle tympanogram), and an ear with a traumatic eardrum perforation (top tympanogram).

In a child less than 7 years old, a volume greater than 1.0 cm³ is indicative of a perforated eardrum, whereas in adults the volume must exceed 2.5 cm³ (Shanks et al., 1992). Volumes exceeding these ranges clearly indicate a perforated eardrum, but flat tympanograms with smaller volumes do not necessarily rule out eardrum perforation. A flat tympanogram with a normal V_ea can also be recorded from an ear with eardrum perforation and cholesteatoma filling the middle ear space and closing off the mastoid air-cell system. Case history and otoscopic examination are very important in these cases. No consistent pattern of hearing loss is associated with eardrum perforation; air–bone gaps can be absent or as large as 50–70 dB if necrosis of the incus also occurs.

Figure 5 demonstrates that otosclerosis also is associated with a variety of tympanogram shapes. Tympanograms vary from a normal type A pattern (shown in Fig. 1) to a low-admittance, stiff pattern (type A_s) shown by the lower tympanogram in Figure 5. A third pattern frequently recorded in otosclerosis is a normal type A pattern, but with a narrow tympanogram width (Shanks, 1984; Shahnaz and Polka, 1997). Pure-tone audiometry in otosclerosis shows a stiffness tilt, with the largest air–bone gaps at low frequencies and the smallest air–bone gap near 2000 Hz. Otosclerosis is virtually the only middle ear pathology where significant air–bone gaps are measured in conjunction with a normal type A tympanogram.

Figure 5..  

Type A acoustic admittance tympanograms recorded using a 226 Hz probe signal in two ears with surgically confirmed otosclerosis.

Figure 6 shows two cases of type A tympanograms with abnormally high Y_tm (2.5 mmhos). Normal tympanograms with deep peaks sometimes are designated type A_d. The bottom tympanogram was recorded from an ear with traumatic ossicular discontinuity; the audiogram showed a maximum conductive hearing loss with 30–70 dB air–bone gaps. The top tympanogram was recorded from an ear with a monomeric eardrum resulting from a healed perforation; the audiogram showed slight air–bone gaps at only 3000 and 4000 Hz. This A_d pattern also is typical of ears with tympanosclerotic plaques on the eardrum and status post stapedectomy. These cases of high-admittance pathology are another indication for high-frequency tympanometry. High-frequency tympanograms in ears with ossicular discontinuity typically exhibit broader, more undulating peaks than tympanograms recorded from ears with eardrum pathology. In cases of high Y_tm, otoscopic examination of the eardrum is crucial; high-admittance pathology of the eardrum can dominate or mask low-admittance pathology such as otosclerosis. Pure-tone audiometry also is invaluable in these cases. Eardrum pathology alone does not produce a significant conductive hearing loss, whereas ossicular discontinuity results in a maximum conductive hearing loss.

Figure 6..  

Type A_d acoustic admittance tympanograms recorded using a 226 Hz probe signal in an ear with traumatic ossicular discontinuity (bottom tympanogram) and in an ear with a monomeric tympanic membrane (top tympanogram).

The preceding cases demonstrate that tympanometry is most beneficial when used as one of a battery of tests that also include case history, otoscopic examination, and pure-tone audiometry. The cases also demonstrate that each unique middle ear problem does not produce one and only one tympanogram pattern. On the contrary, a single pathology can produce several different tympanometry patterns, and conversely, a single tympanogram pattern can result from several different middle ear problems. When used with a battery of tests, however, the contribution from tympanometry can be unique and informative.

See also middle ear assessment in the child.