Normative Multifrequency Tympanometry in Infants and …...Palabras Clave: Admitancia del aire en el conducto auditivo, conductancia, infante, admitancia del oído medio, timpanometría

Delivered by Ingenta to American Academy of Audiology MembersIP 108723431 On Thu 04 Jun 2015 192507

J Am Acad Audiol 17470ndash480 (2006)

470

Department of Communication Sciences and Disorders Institute for Sensory Research Syracuse University Syracuse NYdaggerCurrently affiliated with the University of Maryland Department of Hearing and Speech Sciences

Lauren Calandruccio 621 Skytop Road Institute for Sensory Research Syracuse University Syracuse NY 13244 Phone315-442-7111 Fax 315-443-1113 E-mail lcalandrsyredu

Data contained in this paper were presented at the American Academy of Audiologyrsquos 17th Annual Convention WashingtonDC March 30ndashApril 2 2005

This work was funded by the NIDCD grant R29 DC20285

Normative Multifrequency Tympanometry inInfants and Toddlers

Abstract

Multifrequency tympanometry data were measured multiple times between theages of four weeks and two years from 33 infantstoddlers Tympanograms werealso measured from 33 adult participants Tympanograms recorded with fiveprobe-tone frequencies (226 400 630 800 and 1000 Hz) were classified usingthe Vanhuyse et al model classification system (Vanhuyse et al 1975)Admittance at +200 daPa (Y200) and middle ear admittance (YME) werecalculated The proportion of Vanhuyse et al patterns in infants and toddlerswas different than in adults especially for younger ages YME and Y200 bothincreased with age YME and Y200 data for all infanttoddler groups weresignificantly lower than adult values at all of the tested probe-tone frequenciesThese data can be used as a guide in the clinic to assess normal tympanometricvalues for infants and toddlers

Key Words Admittance of air in ear canal conductance infant middle earadmittance multifrequency tympanometry susceptance toddler Vanhuyse etal model

Abbreviations B = susceptance CNL = could not label ECV = ear canalvolume G = conductance JCIH = Joint Committee on Infant Hearing YME =middle ear admittance Y200 = admittance at +200 daPa

Sumario

Se colectoacute informacioacuten en 33 infanteslactantes de timpanometriacuteas demultifrecuencia realizadas en muacuteltiples ocasiones entre las edades de cuatrosemanas a dos antildeos Tambieacuten se midieron los timpanogramas de 33 adultosparticipantes Los timpanogramas registrados con sondas de prueba de cincofrecuencias diferentes (226 400 630 800 y 1000 Hz) se clasificaron usandoel sistema de clasificacioacuten del modelo de Vanhuyse y col (Vanhuyse y col 1975)Se calculoacute la admitancia a + 200 daPa (Y200) y la admitancia del oiacutedo medio(YME) La proporcioacuten de los patrones de Vanhuyse y col en infantes y lactantesfue diferente que en adultos especialmente en las edades maacutes joacutevenes Tantola YME como la Y200 se incrementaron con la edad La informacioacuten sobre la YMEy la Y200 en los grupos de infanteslactantes fue significativamente maacutes bajaque los valores en los adultos en todas las frecuencias del tono de prueba Estosdatos pueden ser utilizados como guiacutea en la cliacutenica para evaluar los valorestimpanomeacutetricos normales en nintildeos pequentildeos y lactantes

Palabras Clave Admitancia del aire en el conducto auditivo conductanciainfante admitancia del oiacutedo medio timpanometriacutea de multifrecuenciasusceptancia lactante modelo de Vanhuyse y col

Abreviaturas B = susceptancia CNL = sin poderse etiquetar ECV = volumendel canal auditivo G = conductancia JCIH = Comiteacute Conjunto sobre AudicioacutenInfantil YME = admitancia del oiacutedo medio Y200 = admitancia a + 200 daPa

Lauren CalandruccioTracy S FitzgeralddaggerBeth A Prieve

Delivered by Ingenta to American Academy of Audiology MembersIP 108723431 On Thu 04 Jun 2015 192507

TTyymmppaannoommeettrryy iinn IInnffaannttss aanndd TTooddddlleerrssCalandruccio et al

471

External and middle ear structureschange with development becomingadultlike at approximately nine years

of age The most rapid changes occur during thefirst several months of life for example theangle of the tympanic membrane increasesrelative to the ear canal and the middle earcavity continues to grow (review in Northernand Downs 2001) Developmental changes inmiddle ear admittance and reflectance havealso been reported (Holte et al 1991 Keefe etal 1993 Prieve et al 2005) and are not yetcomplete by two years of age (Keefe et al 1993Prieve et al 2005) These data suggest that age-specific norms for middle ear measurementsmay be necessary for infants and toddlersMoreover clinic norms specifically for infantsare necessary in light of the recommendationof the Joint Committee on Infant Hearing(JCIH) JCIH advocates that diagnosis ofhearing loss be made before three months ofage for those children who do not pass theiruniversal newborn hearing screening (JCIH2000) For proper diagnosis of hearing lossand subsequent intervention JCIH proposesthat middle ear admittance measuresincluding tympanometry be included in the testbattery

Normative data are available for neonates(Keith 1973 Kei et al 2003 Margolis et al2003) and for infants broadly grouped intoage categories starting at six months (Roushet al 1995 De Chicchis et al 2000) Currentlynormative data is not available for infantsbetween four weeks and six months of agealthough audiological diagnosis andintervention are most likely to occur duringthis period (JCIH 2000) Roush et al (1995)provided data for infants between 6 and 30months of age grouped into four age bracketseach spanning a six-month period De Chicchiset al (2000) grouped data into five age groupswith the youngest age group covering infants6 to 11 months of age Four older age groupseach covered an 11-month period Holte et al(1991) reported data from infants less than sixmonths but the intent of the paper was toshow developmental changes intympanograms Consequently normative datafor clinical use are not easily obtained fromHolte et al (1991)

Perhaps one reason why normativetympanometric data for infants aged youngerthan six months are not available is that mosttympanometry is performed using a 226 Hzprobe tone Tympanometry using this probe-

tone frequency is insensitive to middle eareffusion in infants less than six to sevenmonths of age (Paradise et al 1976 Shurin etal 1976 Hunter and Margolis 1992 Keefe atal 1993 Rhodes et al 1999 Purdy andWilliams 2000) The use of higher-frequencyprobe tones has been recommended whentesting young infants because higher-frequency tones appear to be more sensitiveto middle ear effusion (Williams et al 1995Purdy and Williams 2000 Kei et al 2003Margolis et al 2003) However using higher-frequency probe tones results in multipeakedtympanometric shapes making clinicalinterpretation more difficult Normative datafor tympanograms measured using higher-frequency probe tones (on newer instrumentsoperating in accordance with the AmericanNational Standards Institute [ANSI]standards [ANSI 1987]) exists only for 1000Hz probe tones in newborns (Kei et al 2003Margolis et al 2003)

The Vanhuyse et al model has been usedin the interpretation of tympanograms recordedwith higher-frequency probe tones (Vanhuyseet al 1975) The Vanhuyse et al model definesfour patterns of admittance tympanogramsbased on the number of combined minima andmaxima (or extrema) in both susceptance (B)and conductance (G) tympanograms Thismodel was developed to better understandmultipeaked tympanograms either due todifferences in probe-tone frequencies or middleear disease in adults but has been extendedto study middle ear transfer functioncharacteristics in infants (Margolis et al 2003Holte et al 1991 Prieve et al 2005)

McKinley et al (1997) and Sprague et al(1985) used the Vanhuyse et al model to classifytympanograms recorded in neonates McKinleyet al (1997) concluded that the Vanhuyse et almodel was not adequate for classifying andinterpreting the majority of their high-frequency neonatal tympanograms Spragueet al (1985) reported neonatesrsquo tympanogramsto be less complex than those observed inadults Holte et al (1991) also found thepercentage of tympanometric types wasdifferent among infants of different ages (birthto 45 months) and adults but felt that themodel could be used to classify a largepercentage of the tympanograms No dataexist describing tympanometric types for olderinfants and toddlers

The purpose of the current study wastwofold First we sought to describe

Delivered by Ingenta to American Academy of Audiology MembersIP 108723431 On Thu 04 Jun 2015 192507

tympanogram patterns using the Vanhuyseet al model for infants and toddlers aged fourweeks to two years of age Second we wantedto provide normative tympanometric datafor infants and toddlers measured with 226630 and 1000 Hz probe tones

MMEETTHHOODD

SSuubbjjeeccttss

The 33 infantstoddlers (12 female 21male) that were enrolled in the current studywere a part of a larger longitudinal studydesigned to explore the relationship betweenchanges in middle ear admittance andcorresponding changes in otoacousticemissions None of the participants had anyrisk factors for hearing loss (JCIH 2000)and were cared for in the well-baby nurseryof the hospital

Data were collected over a 24-monthobservation period An attempt to collect dataat specific ages (1 2 3 4 6 8 10 12 16 20and 24 months) was made However themaintenance of this schedule was difficult dueto reasons such as negative middle ear pressurenoncooperation missed appointmentsrelocation of families and equipment failureData were collected on individual participantsan average of 77 test ageschild (standarddeviation = 3 test ages) Each participanthad data collected on one randomly selectedear (18 left ears 15 right ears) Participantsrsquodata were separated by age into five groups4 to 10 weeks (mean = 69 weeks n = 39) 11to 19 weeks (mean = 156 weeks n = 41) 20to 26 weeks (mean = 232 weeks n = 15) 6to 12 months (mean = 381 weeks n = 50) and2 years (mean = 1028 weeks range 99ndash113weeks n = 26) TEOAEs or DPOAEs werepresent and robust or behavioral thresholdswere le15 dB HL for at least two of fourfrequencies (500 1000 2000 and 4000 Hz)on the days tympanograms were recorded

Data were also collected on one ear from33 adult subjects (18 female 15 male 24right ears 9 left ears) that were used as acontrol in examination of tympanogram type(mean age = 303 years standard deviation= 79 years) One tympanogram for eachfrequency per adult subject were used in theanalyses for this study None of the adults hada history of ear pathologies and all had

normal hearing (lt15 dB HL) bilaterally atoctave frequencies between 250 and 8000Hz Adult subjects and the parents of infantsubjects signed an informed consent formand were paid for their participation

IInnssttrruummeennttaattiioonn aanndd PPrroocceedduurreess

The Virtual Model 310 acousticadmittance instrument was used to collect alldata The Virtual 310 is a two-componentadmittance meter with the ability to analyzeconductance (G) and susceptance (B)separately The Virtual 310 was controlled bya Macintosh Powerbook 165 with softwareprovided by Virtual Corporation Prior to alldata collection calibration measurementswere made for every probe-tone frequencyin hard-walled calibration cavities (05 20and 50 cc) supplied by the manufacturer

Measurement of five sweep-pressuretympanograms were attempted using apositive to negative air pressure sweep at arate of 125 daPasec and probe level of 85 dBSPL The five probe-tone frequencies were226 400 630 800 and 1000 Hz Of thesefrequencies detailed analyses were completedon tympanograms measured with probe-tonefrequencies of 226 630 and 1000 Hz becausethese are typical frequencies used in theclinic The order of the probe-tone frequenciesused was quasirandomized that is as manytympanograms as possible were collectedwhile the participant remained cooperativePediatric participants either were held asleepandor awake by their parent or for sometoddlers sat on their own while testingoccurred If the participant was awakeexperimenters attempted to distract the childfrom the testing procedure with toys

AAnnaallyysseess

Tympanograms with tympanometricpeak pressure more negative than -150 daPawere not included in analyses to avoidincluding any ears with middle eardysfunction There were ten infants who hadnegative middle ear pressure during at leastone visit Data were not collected from theseears on those days B and G values from thetympanograms were measured at +200 daPaand at the air pressure with the greatestadmittance The admittance at +200 daPa

472

JJoouurrnnaall ooff tthhee AAmmeerriiccaann AAccaaddeemmyy ooff AAuuddiioollooggyyVolume 17 Number 7 2006

Delivered by Ingenta to American Academy of Audiology MembersIP 108723431 On Thu 04 Jun 2015 192507

TTyymmppaannoommeettrryy iinn IInnffaannttss aanndd TTooddddlleerrssCalandruccio et al

473

(Y200) and middle ear admittance (YME) werecalculated

Middle ear admittance was calculatedwith the following equation

YME =

The tip is the extrema (either peak or dipdepending on tympanometric shape) at theapplied air pressure and the tail is theadmittance at +200 daPa

Admittance at +200 was calculated usingthe following equation

Y200 =

Y200 was reported rather than ear canalvolume (ECV) because most clinicalequipment calculates ECV only at 226 Hz Ina purely compliant system the relationshipbetween ECV and Y200 is as follows (Margolisand Shanks 1990)

ECV = Y200

Nonparametric one-way ANOVAs wereperformed for YME and Y200 data at a

significance level of α = 005 Nonparametricstatistics were used for these data becausemore than one tympanogram from the samesubject was usually included in a given agerange (due to the longitudinal nature of thestudy) Consequently the assumption ofindependence across the data was violatedTo account for violations of independencethis type of ANOVA ranks the data prior tostatistical testing Post hoc testing wasperformed using Wilcoxon two-sample t-tests(Wilcoxon 1945) to compare the relationshipamong all five groups (αFW = 005)

RREESSUULLTTSS

Atotal of 204 tympanograms weresuccessfully recorded at 226 630

and 1000 Hz All tympanograms wereclassified as one of the four Vanhuyse etal patterns or deemed ldquocould not belabeledrdquo (CNL) Figure 1 shows an exampleof the four Vanhuyse et al tympanogramtypes recorded from the current studyrsquosdata The top left panel of Figure 1 showsan example of a 1B1G tympanogram

FFiigguurree 11 The four classic Vanhuyse et al patterns These examples are actual data from the current studyThe top left plot is a 1B1G type pattern the bottom left plot is 93B1G plot the top right plot is a 3B3G plotand the bottom right plot is a 5B3G plot All tympanograms were characterized as one of these four Vanhuyseet al types or were deemed ldquocould not be labeledrdquo (CNL)

Delivered by Ingenta to American Academy of Audiology MembersIP 108723431 On Thu 04 Jun 2015 192507

JJoouurrnnaall ooff tthhee AAmmeerriiccaann AAccaaddeemmyy ooff AAuuddiioollooggyyVolume 17 Number 7 2006

474

where both the B and G tympanograms havea single peak The top right panel illustratesa 3B1G tympanogram For this classificationthe G tympanogram continues to have asingle peak however the B tympanogramnow has two maxima and one minimum orthree extrema In the example of the 3B3Gtympanogram shown in the bottom left panelof Figure 1 both the B and G tympanogramshave three extrema In the bottom rightpanel the 5B3G tympanogram the Gtympanogram has three extrema while the Btympanogram has five extrema

Figure 2 illustrates the percentage oftympanogram types for the six different agegroups assessed in this study As expected theshapes of the tympanograms become morecomplex as probe frequency increases Theresults shown in the top panel of Figure 2were collected using a 226 Hz probe toneThe majority of 226 Hz tympanograms for allage groups had a 1B1G Vanhuyse et al-typepattern However the two youngest groups(4ndash10 weeks and 11ndash19 weeks) also had asmall proportion of 3B1G tympanograms(231 and 68 respectively)

When a 630 Hz probe tone was used(the middle panel of Figure 2) all age groupscontinued to have predominantly 1B1Gtympanograms This type comprised morethan 90 of tympanograms for the threeoldest infanttoddler and adult age groupsHowever the youngest group had a largerproportion of 3B1G and 3B3Gtympanograms Data for the 1000 Hz probetone are shown in the bottom panel of Figure2 Adults had predominantly 3B1G patternswhereas infants and toddlers had a widervariety of tympanogram types at thisfrequency

The distribution of tympanogram typesas a function of frequency for all age groupsis examined in Figure 3 The percentage oftympanogram type for five probe-tonefrequencies (226 400 630 800 and 1000 Hz)are shown The top panel displays data fromthe 4ndash10-week-old group The percentages oftympanogram types for the 4ndash10-week-oldgroup are markedly different from thepercentages for all other age groups Thetwo most notable differences are a lowerpercentage of 1B1G tympanograms for allprobe-tone frequencies and a greaterpercentage of 3B3G tympanogramsespecially at 400 and 600 Hz With respectto the three youngest age groups (the leftcolumn of plots) the overall proportion of3B3G and CNL tympanograms decreasedwith age from 88 to 2 and 158 to 66respectively when data were averaged acrossprobe-tone frequencies while the averagepercentage of 1B1G tympanograms increasedwith age from 522 to 781 However thepercentages of tympanogram types continuedto change between six months and adulthood(the right column of plots) The mostnoticeable of these changes is the increase in3B1G tympanograms at 1000 Hz

Descriptive statistics including medians

FFiigguurree 22 Percentage of tympanogram types as afunction of Vanhuyse et al model tympanogram typesare shown for six age groups The tympanogramsshown on the top panel were recorded using a 226 Hzprobe tone the middle panel with a 630 Hz probe toneand the bottom panel with a 1000 Hz probe tone

Delivered by Ingenta to American Academy of Audiology MembersIP 108723431 On Thu 04 Jun 2015 192507

TTyymmppaannoommeettrryy iinn IInnffaannttss aanndd TTooddddlleerrssCalandruccio et al

475

and 5th and 95th percentiles for YME and Y200for 226 630 and 1000 Hz probe tones areshown in Tables 1 and 2 respectively Thenonparametric statistics reflect comparisonsacross age groups not across probe-tonefrequencies therefore statistical comparisonscan be made only within frequencies(displayed in columns) Group data within acolumn that were not significantly differentfrom one another are enclosed in bracketsThus those groups not within the samebracket were significantly different from oneanother

For a 226 Hz probe tone there were fewsignificant differences in middle earadmittance among infant age groups Thedashed portion of the leftmost bracketindicates that YME was significantly higherfor the youngest group than for the 11ndash19-week-oldgroup but not significantly different fromthe other infanttoddler age groups Therewere more significant differences between

age groups for the two higher probefrequencies For a probe tone of 630 Hz theyoungest age group was significantly differentfrom all other age groups Infants aged 11ndash19weeks had significantly lower middle earadmittance than infants aged six months totwo years and infants aged 20ndash26 weekshad significantly lower middle ear admittancethan those aged two years Using a 1000 Hzprobe tone infants aged 4ndash19 weeks hadsignificantly lower middle ear admittancethan those aged six months to two years

There were striking differences inmiddle ear admittance betweeninfantstoddlers and adults All infant agegroups had significantly lower middle earadmittance than did adults for all threefrequencies These results are similar tothose seen in Keefe et al (1993) and Prieveet al (2005) and indicate that developmentalchanges in the middle ear are not yetcomplete even at 24 months of age

FFiigguurree 33 Distribution of tympanogram types as a function of frequency for all age groups Each of the six pan-els illustrate data from one of the six age groups reported in this study For example the bottom plot in theright column illustrates the percentage of tympanogram types as a function of frequency for children 20ndash26weeks of age The filled black area represents 1B1G type tympanograms At 226 Hz all of the tympanogramsfor the 20ndash26 week age group were 1B1G (filled black area reaches 100 at 226 Hz) The other tympanogramtypes are represented by different color patterns indicated in the key

Delivered by Ingenta to American Academy of Audiology MembersIP 108723431 On Thu 04 Jun 2015 192507

JJoouurrnnaall ooff tthhee AAmmeerriiccaann AAccaaddeemmyy ooff AAuuddiioollooggyyVolume 17 Number 7 2006

476

Age differences for admittance at +200daPa were also frequency dependent At 226Hz infants aged four weeks to 26 months hadsignificantly lower admittance at +200 daPathan children aged two years At 630 Hzinfants aged 4ndash19 weeks had significantlylower admittance at +200 daPa than didinfants aged six months to two years At 1000Hz infants aged 4ndash10 weeks had significantlylower admittance at +200 daPa than didthose aged 20 weeks to two years of age andthose aged 11ndash19 weeks had lower valuesthan those aged two years All infants had

lower admittance at +200 daPa than didadults at all frequencies tested Loweradmittance at +200 daPa would result insmaller calculated ear canal volumes

Figure 4 is an illustration of the mostclinically relevant data from Tables 1 and 2The dark line in each plot represents themedian data for each age group while thegray shaded area represents the 5th to 95thpercentile For all four measures the lowerlimit (5th percentile) did not vary asconsiderably with age as did the upper limit(95th percentile)

Table 1 Middle Ear Admittance (YME) for Tympanograms Using 226 630 and 1000 Hz ProbeTones for Five Young Age Groups and Adults

MIDDLE EAR 226 Hz Median 630 Hz Median 1000 Hz MedianADMITTANCE (YME) 5thndash95th percentile 5thndash95th percentile 5thndash95th percentile

4ndash10 Weeks 0400 0325 1067n=39 0140ndash0733 0075ndash1278 0201ndash2258

11ndash19 Weeks 0249 0552 1453n=39 0126ndash0567 0255ndash1247 0701ndash2272

20ndash26 Weeks 0276 0763 1409n=15 0145ndash0544 0440ndash1158 0616ndash2177

6ndash12 Months 0310 0930 1933n=50 0158ndash0597 0401ndash1919 0948ndash4197

2 Years 0310 1114 2466n=26 0209ndash1029 0686ndash4543 1118ndash4073

Adults 0626 2033 2761n=33 0360ndash1113 0814ndash4139 0867ndash4690

Note Statistical significance of the data were classified by brackets where group medians within the same bracket were not significantlydifferent from one another Statistical analyses were conducted across age groups not across probe-tone frequencies therefore comparisonsshould be made within columns only Note that the dashed line in the furthest bracket on the left indicates that the 11ndash19-week group wassignificantly different from the youngest group however the youngest group was not significantly different from all other pediatric groups

Table 2 Admittance at +200 daPa (Y200) for Tympanograms Using 226 630 and 1000 Hz ProbeTones for Five Young Age Groups and Adults

ADMITTANCE AT 226 Hz Median 630 Hz Median 1000 Hz Median+200 daPa (Y200) 5thndash95th percentile 5thndash95th percentile 5thndash95th percentile

4ndash10 Weeks 0358 1006 1263n=39 0228ndash0551 0731ndash1692 0997ndash1714

11ndash19 Weeks 0363 1064 1432n=41 0249ndash0602 0839ndash1429 1087ndash1857

20ndash26 Weeks 0288 0870 1525n=15 0193ndash0506 0755ndash1365 1310ndash1974

6ndash12 Months 0400 0971 1768n=50 0129ndash0687 0727ndash1576 1432ndash2513

2 Years 0435 1207 2040n=26 0184ndash0577 0789ndash1599 1469ndash3375

Adults 1051 2729 3896n=33 0700ndash1685 1740ndash3703 2567ndash5239

Note Group medians within the same bracket were not significantly different from one another

Delivered by Ingenta to American Academy of Audiology MembersIP 108723431 On Thu 04 Jun 2015 192507

TTyymmppaannoommeettrryy iinn IInnffaannttss aanndd TTooddddlleerrssCalandruccio et al

DDIISSCCUUSSSSIIOONN

Data from this study indicate that theproportions of Vanhuyse et al-type

tympanograms in infants and toddlers aredifferent than those in adults especiallyfor younger infants One of the mostnoticeable differences occurred fortympanograms recorded with a 1000 Hzprobe tone At this frequency adults hadpredominantly 3B1G tympanograms (80)whereas infants and toddlers tended tohave a more equal distribution between1B1G and 3B1G types An equal distributionof 1B1G and 3B1G tympanograms wasnoted for adults at approximately 800 HzThe percentages of the different Vanhuyseet al types as a function of frequency foradults are similar to data reported inMargolis and Goycoolea (1993)

Few researchers have attempted to use

the Vanhuyse et al model to classifytympanogram types in infants Data inthese studies were collected only fromneonates (Shurin et al 1976 Sprague et al1985) or using slightly different probe-tonefrequencies than those used in the presentstudy (Holte et al 1991) A gross comparisoncan be made between the current resultsand those of Holte et al (1991) Holte et al(1991) reported Vanhuyse et altympanometric types for 226 Hz probe tonesin infants aged 26ndash47 51ndash66 and 103ndash133days old If we combine the two youngergroups from their study (26ndash47 51ndash66 daysold) and average the data it can becompared to those from our youngest group(4ndash10 weeks old) Results from Holte et al(1991) for this age range indicated that545 had 1B1G 20 had 3B1G 195had 3B3G and 55 had 5B3Gtympanograms In the current study infants

477

FFiigguurree 44 Middle ear admittance (YME) for tympanograms using 226 (top left) 630 (top right) and 1000 (bot-tom left) Hz probe tones for five young age groups (data taken from Table 1) Admittance at +200 daPa (Y200)for tympanograms using a 1000 Hz probe tone (bottom right) for five young age groups (data taken from Table2) Note that the y-axes of the four subplots are scaled differently for each condition

Delivered by Ingenta to American Academy of Audiology MembersIP 108723431 On Thu 04 Jun 2015 192507

in the youngest group have a similarpercentage of 3B1G tympanograms (231)but a higher percentage of 1B1Gtympanograms (769) and notympanograms classified as 3B3G or 5B3GThe oldest group (103ndash133 days old) reportedin Holte et al (1991) can be compared to our11ndash19-week-old group Results for this agerange were similar between our study and theHolte et al (1991) study 909 and 100demonstrated 1B1G types respectively Holteet al (1991) measured tympanograms withslightly different probe-tone frequencies thanthose used in this study with the exceptionof 226 Hz For example the results in Holteet al (1991) at 450 710 and 900 Hz weresimilar to our results at 400 800 and 1000Hz respectively Because higher probe-tonefrequencies were different between studiesdirect comparisons cannot be made Howeverpercentages of tympanogram types for eachof the higher probe frequencies even whenthe probe frequencies differed byapproximately 100 Hz were similar betweenthe two sets of data Based on our results andthose from others the Vanhuyse et al modelcan be used to classify tympanograms ininfants and toddlers aged four weeks to twoyears

Data in the present study are similar toYME reported by others although somedifferences in data collection and reportingexist One difference is that various pumpspeeds have been used Current clinicalequipment uses a pump speed of 200 or600200 daPasec The pump speed in thecurrent study was 125 daPasec A review ofthe literature reported by Margolis andShanks (1990) indicates that in infants agedtwo to four-and-a-half months YME weresmaller using a pump sweep of 200 daPaseccompared to 50 daPasec In children agedthree to five years the opposite was truethat is YME increased with faster pumpspeeds Because our data span the ages of fourweeks through two years of age we areuncertain how different pump speeds wouldaffect our data Previously published datawere collected using various pump speedsandor from different age groups than thedata in the present study thereforecomparisons among studies are broadlyinterpreted

Data from the present study comparefavorably with the interim normative adulttympanometric data published by ASHA

(1990) based on the results reported byMargolis and Heller (1987) The median YMEat 226 Hz from the current study (0626mmhos) is similar to their mean YME (072mmhos) The 90 range for our YME data(0360ndash1113 mmhos) also fell within the 90range reported by Margolis and Heller (1987)(027ndash138 mmhos)

Median YME results for the 6ndash12-month-oldage group using a 226 Hz probe tone (031mmhos) are similar to those results reportedin both De Chicchis et al (2000) and Roushet al (1995) These authors reported meanadmittance values of 032 and 039 mmhosrespectively The range reported in the Roushstudy for this age range is similar to ourreported range 020ndash050 and 016ndash060mmhos respectively However the upperlimit for the range of our two-year-old group(020ndash1029 mmhos) is larger than the resultsreported for the 18ndash24 month group reportedin Roush et al (1995) (030ndash070 mmhos)

Data from our 11ndash19-week-old group issimilar to YME in infants (mean age = 14weeks) reported by Margolis and Popelka(1975) Margolis and Popelka (1975) reportedmean YME of 0312 and 0679 mmhos for 220and 630 Hz probe tones respectively whereasthe median YME for similar frequencies in thecurrent study were 0249 and 0552 mmhosrespectively Holte (1989) reported a meanYME value of 032 mmhos for two- to four-and-a-half-month olds when using a 226 Hz probetone Median data for our two youngestgroups (4 to 19 weeks) was similar (YME of0316 mmhos) Results from the current studyare similar to previously reported data if thedifferences between mean and mediancalculations are accounted for

It is not possible to compare our 1000 HzYME and Y200 data with other recent studiesthat have measured 1000 Hz tympanograms(Kei et al 2003 Margolis et al 2003) Theoldest infants tested by Kei et al (2003) weresix days old while the oldest tested byMargolis et al (2003) were four weeks oldThus our youngest age group was older thanboth studiesrsquo populations In addition bothKei et al (2003) and Margolis et al (2003)used the term ldquostatic admittancerdquo to indicatethe peak-to-tail difference in admittancemagnitude Subtracting the tail from the tipusing admittance magnitude does not takeinto account any phase differences that couldbe present between the pressures at the tailand the tip and does not accurately estimate

JJoouurrnnaall ooff tthhee AAmmeerriiccaann AAccaaddeemmyy ooff AAuuddiioollooggyyVolume 17 Number 7 2006

478

Delivered by Ingenta to American Academy of Audiology MembersIP 108723431 On Thu 04 Jun 2015 192507

the admittance at the tympanic membraneYME data reported in the current study isadmittance recorded at the tympanicmembrane Additionally Margolis et al (2003)recommended that YME be calculated forinfants using the tip value minus the tail at-400 daPa applied pressure Our data werebased on calculations at +200 daPa appliedpressure because data were collected usinga positive-to-negative sweep as recommendedby Holte et al (1991) Tympanogram traceswere generally less noisy at the beginning ofthe trace than at the end of the trace somore accurate measurements could be madeIn addition the most negative pressure testedin the current study was -280 daPa makingdirect comparisons between studiesimpossible

Our results indicated that YME and Y200generally increased with age howeversignificant differences among infanttoddlerage groups varied across frequencies Takenas a whole age-specific probe-tone-frequency-specific norms for YME and Y200 are necessaryfor infants younger than six months of ageComparing YME and Y200 measurements to anormative range such as the one presentedin Figure 4 may prove to be useful indiagnostic evaluations

Future research should focus ontympanometric types and admittance ininfants and toddlers with confirmed middleear disease The distribution of data fromears with pathology should be compared tothe distribution from a normal populationto determine whether these data can be usedto diagnose middle ear disorders Futureresearch should also seek to determine whichadmittance measures are the most beneficialfor instance whether both subcomponentsshould be monitored separately or whethersimply looking at admittance magnitudeprovides the most useful information

AAcckknnoowwlleeddggmmeenntt The authors would like to thankseveral masters and undergraduate students andtechnicians who helped in collecting the data andorganizing the database especially Holly HutaMichael Jackson Nadia Zebluim and Anita SternsWe also thank Drs Michael Anzalone and Paul CNelson for their technical assistance and helpful com-ments while writing this manuscript The authorswould like to thank Dr Robert Margolis and an anony-mous reviewer for valuable comments on an earlierversion of this manuscript

RREEFFEERREENNCCEESS

American National Standards Institute (1987)Specifications for Instruments to Measure AuralAcoustic Impedance and Admittance (Aural AcousticImmitance) S339 New York American NationalStandards Institute

American Speech-Language-Hearing Association(1990) Guidelines for screening for hearing impair-ments and middle ear disorders ASHA 3217ndash24

De Chicchis AR Todd NW Nozza RJ (2000)Developmental changes in aural acoustic admittancemeasurements J Am Acad Audiol 1197ndash102

Holte LA (1989) Longitudinal tympanometry andpneumatic otoscopy in healthy newborn infants PhDdiss Syracuse University

Holte L Margolis RH Cavanaugh R (1991)Developmental changes in multifrequency tym-panograms Audiology 301ndash24

Hunter LL Margolis RH (1992) Multifrequency tym-panometry current clinical application Am J Audiol133ndash43

Joint Committee on Infant Hearing (2000) JCIH year2000 position statement principles and guidelinesfor early hearing detection and intervention programsAm J Audiol 99ndash29

Keefe DH Bulen JC Arehart KH Burns EM (1993)Ear-canal impedance and reflection coefficient inhuman infants and adults J Acoust Soc Am942617ndash2638

Kei J Allison-Levick J Dockray J Harrys RKirkegard C Wong J Maurer M Hegarty J YoungJ Tudehope D (2003) High-frequency (1000 Hz) tym-panometry in normal neonates J Am Acad Audiol1420ndash28

Keith RW (1973) Impedance audiometry withneonates Arch Otolaryngol 97465ndash467

Margolis RH Bass-Ringdahl S Hanks WD Holte LZapala DA (2003) Tympanometry in newborninfantsndash-1 kHz norms J Am Acad Audiol 14383ndash392

Margolis RH Goycoolea HG (1993) Multifrequencytympanometry in normal adults Ear Hear14408ndash413

Margolis RH Heller JW (1987) Screening tympa-nometry criteria for medical referral Audiology26197ndash208

Margolis RH Popelka GR (1975) Static and dynamicacoustic impedance measurements in infant ears JSpeech Hear Res 18435ndash443

Margolis RH Shanks JE (1990) Tympanometry basicprinciples and clinical applications In RintelmannWF ed Hearing Assessment 2nd ed Austin TX Pro-Ed 179ndash246

McKinley AM Grose JH Roush J (1997)Multifrequency tympanometry and evoked otoacousticemissions in neonates during the first 24 hours oflife J Am Acad Audiol 8218ndash223

TTyymmppaannoommeettrryy iinn IInnffaannttss aanndd TTooddddlleerrssCalandruccio et al

479

Delivered by Ingenta to American Academy of Audiology MembersIP 108723431 On Thu 04 Jun 2015 192507

Northern JL Downs MP (2001) Hearing in Children5th ed Philadelphia PA Lippincott Williams andWilkins

Paradise JL Smith CG Bluestone CD (1976)Tympanometric detection of middle ear effusion ininfants and young children Pediatrics 58198ndash210

Prieve BA Chasmawala S Jackson M (2005)Development of middle-ear admittance in humansAssoc Res Otolaryngol Abs 692

Purdy SC Williams MJ (2000) High frequency tym-panometry a valid and reliable immittance testprotocol for young infants N Z Audiol Soc Bull10(3)9ndash24

Rhodes MC Margolis RH Hirsch JE Napp AP (1999)Hearing screening in the newborn intensive care nurs-ery comparison of methods Otolaryngol Head NeckSurg 120808

Roush J Bryant K Mundy M Zeisel S Roberts J(1995) Developmental changes in static admittanceand tympanometric width in infants and toddlers JAm Acad Audiol 6334ndash338

Shurin PA Pelton SI Klein JO (1976) Otitis mediain the newborn infant Ann Otol Rhinol Laryngol85(S25)216ndash222

Sprague B Wiley T Goldstein R (1985)Tympanometric and acoustic-reflex studies inneonates J Speech Hear Res 28265ndash272

Vanhuyse VJ Creten WL Van Camp KJ (1975) Onthe W-notching of tympanograms Scand Audiol44550

Wilcoxon F (1945) Individual comparisons by rank-ing methods Biometrics 180ndash83

Williams MJ Purdy SC Barber C (1995) High fre-quency probe tone tympanometry in infants withmiddle ear effusions Austr J Otolaryngol 2169ndash173

JJoouurrnnaall ooff tthhee AAmmeerriiccaann AAccaaddeemmyy ooff AAuuddiioollooggyyVolume 17 Number 7 2006

480