- Email: [email protected]
Tympanometry of a diverse group of preschool aged children
International Journal of Pediatric Otorhinolaryngology (2006) 70, 1523—1527
www.elsevier.com/locate/ijporl
Tympanometry of a diverse group of preschool aged children Abby C. Meyer *, Kathryn J. Webb, Cynthia S. Davey, Kathleen A. Daly University of Minnesota, Department of Otolaryngology, MMC 396, 420 Delaware St SE, Minneapolis, MN 55455, USA Received 20 September 2005; received in revised form 15 February 2006; accepted 20 February 2006
KEYWORDS Otitis media; Tympanometry; Hmong
Summary Objective: Ethnicity has been previously described as a risk factor for middle ear disease. Little data exist on the presence of middle ear disease based on tympanometry screening comparing Asian children and children of other races. Methods: Two hundred and seventy children aged 3—5 were screened with tympanometry at six Head Start sites in St. Paul, Minnesota during the months of September and October of 2004. Gender, age, and race/ethnicity was recorded and entered into a database, along with values for canal volume, static admittance, peak pressure, and tympanometric width. Results: Criteria for abnormal tympanometry were based on American SpeechLanguage Hearing Association (ASHA) recommendations for a failed tympanogram for 1—5 year olds (admittance <0.3 mmho or width >200 daPa). There were no statistically significant differences in failure rates between males and females. There were, however, more failures for Asian (predominantly Hmong) children compared to children of other races/ethnicities after adjusting for age and gender differences (OR = 6.39, CI 3.65—11.2, p < 0.001) and for children <4-years-old compared to children 4—5-years-old after adjusting for race and gender differences (OR = 1.99, CI 1.03—3.84, p < 0.05). Conclusions: Asian children were more than six times as likely to fail tympanometry as children of other races/ethnicities. The explanation for this difference is likely to be multifactorial, and further research is needed to characterize this difference. # 2006 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
* Corresponding author. Tel.: +1 763 571 7556 E-mail address: [email protected] (A.C. Meyer).
Tympanometry is a valuable tool for the assessment of evidence of middle ear effusion. Watters et al. and Nozza et al. compared pre-operative tympanometry to surgical findings and found tympanometry
0165-5876/$ — see front matter # 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2006.02.008
1524 to be highly sensitive in predicting middle ear effusion [1,2]. Furthermore, tympanometry may help identify middle ear disease that may lead to chronic ear disease, hearing loss, and subsequent speech and language difficulties if not detected. However, recent research by Paradise et al. casts doubts on the relationship between persistent middle ear effusion and developmental outcomes at age six [3]. In recent years, the city of St. Paul, Minnesota has seen a large influx of refugees and immigrants from many Southeast Asian nations. According to the 2000 census, 41,800 Hmong (refugees and immigrants from Laos) reside in Minnesota, making it the most populous Asian ethnic group in the state [4]. The majority (58%) lives in St. Paul. As a result, many children of Hmong descent are enrolling in enrichment programs, including Head Start. In the past few years, staff from the Sight & Hearing Association (SHA), the organization that conducts both vision and hearing screening for Head Start programs in St. Paul, noticed that Hmong children had a high referral rate for abnormal tympanometry. The SHA approached researchers in the Department of Otolaryngology at the University of Minnesota with the hopes that this observation could be formally investigated. In reviewing the literature, several studies support race or ethnicity as a predisposing factor for middle ear pathology [5—14]. However, only one study has compared rates of middle ear disease in Asian and White children. Authors reported that 9.5% of White children and 1.3% of Chinese children had middle ear disease as determined with tympanometry and otoscopy [15].
2. Methods 2.1. Data collection Preschool aged children were screened with tympanometry by SHA personnel at six Ramsey Action Program Head Start sites in September and October 2004. Tympanometry was performed using a MAICO MI 26 Tympanometer/Audiometer that was routinely calibrated in August, prior to the start of screening in September. The same equipment was used at all screening sites and daily probe calibration checks were performed. A 226 Hz probe tone was presented with an intensity of 85 dB SPL measured in a 2 ml coupler. A positive to negative pressure sweep between +200 and 400 daPa was administered at a variable rate of 550 daPa/s maximum with 200 daPa/s near peak. Typical test time per ear was less than 3 s. Otologic examinations were not performed as part of the screening.
A.C. Meyer et al. Tympanogram tracings were obtained and printed by SHA personnel, who labeled them with the child’s age, gender, and race/ethnicity. At least two attempts were made to obtain a valid tympanogram. Race/ethnicity was determined by the screener from the child’s Head Start registration. The tympanogram tracings were provided to the University of Minnesota Department of Otolaryngology for interpretation and analysis. The institutional review board approved secondary analysis of anonymous tympanometry data.
2.2. Data analysis Ear canal volume (at 200 daPa), static admittance (compensated at +200 daPa), and peak pressure values were obtained from the tympanogram printout and entered into a database, along with identification number, age, gender, ethnicity, and screening site. One researcher (AM) measured tympanometric width to the nearest 1/16 in. at one-half the peak admittance and converted the value into daPa units with the following conversion: 1/16 in. is equal to 25 daPa. If the tympanogram was ‘‘noisy’’ as evidenced by upward departure from the defined peak, giving an erroneous admittance result, peak compensated admittance was also measured by hand. A 5% sample of tracings was re-measured by another researcher to check for accuracy against the data entered in the database. Criteria for abnormal tympanometry were based on American Speech-Language Hearing Association (ASHA) recommendations for a failed tympanogram for 1—5 year olds based on normative studies. These criteria were admittance <0.3 mmho or width >200 daPa [16]. Comparisons by age, race and gender were done with chi-square analyses, and logistic regression was used to construct a predictive model for tympanometry failure. To allow comparisons with previous screening studies using a flat tympanogram (defined as static admittance <0.2, or 0.1) as evidence of middle ear effusion, criteria of static admittance <0.2 mmho or width >200 daPa were also used to determine tympanometry fail rates by age, gender or race.
3. Results Among the children tested, 49% were female and 51% male; 20% were 3 years old, 54% were 4 years old, and 26% were 5 years old. Twenty-nine percent were African/African American, 39% were Asian (93% were Hmong), and 32% were Caucasian/other, which included Hispanic, Middle Eastern, and others.
Tympanometry of a diverse group of preschool aged children
1525
Table 1 Tympanometric summary statistics by ears Variable
N
Mean
S.D.
Admittance Pressure Width Canal volume
267 258 266 267
0.37 60.34 134.50 0.79
0.26 65.32 104.40 0.16
Five hundred and forty ears of 270 children were tested. Of these, 11 ears had abnormal admittance (<0.3) and abnormal ear canal volume. They were excluded from analysis. Seven of the 11 had canal volumes >1.0, suggesting a tympanostomy tube or tympanic membrane perforation. The other four had canal volumes <0.5, suggesting a blocked probe resulting in an invalid test. Of the 529 remaining ears, 39 had an unreadable tracing, so peak admittance and width could not be determined. Three children had missing data for both ears, leaving 267 children with usable tympanometric data in one or both ears. In order to base the data summaries on independent (non-correlated) observations, only data from one ear per child was selected for analysis of tympanogram measures based on the following rule: data was selected for right ear unless data for only the left ear was available. Tympanometric summary statistics by ears (one ear for each child as described above) are shown in Table 1. Of the 267 children with data for at least one ear, 130 (49%) failed tympanometry in at least one ear with admittance <0.3 mmho or width >200 daPa. Fail rates by age revealed a higher fail rate among 3 year olds (58%) than 4—5 year olds (46%); this difference was not statistically significant ( p = 0.12, Table 2). Asian children had a significantly higher fail rate than the other groups ( p < 0.001) but there Table 2 Tympanometric fail rate by age, race/ethnicity, gender Variable
N (%) failed
Age (years) 3 4 5
32/55 (58) 69/143 (48) 29/68 (43)
Race Asian African/African American Other Gender Female Male a b
3 years vs. 4—5 years. Asian vs. others.
78/105 (74) 34/90 (38)
p-value (chi-square) 0.49 0.12 a
0 385 25 0.5
<0.001 <0.001 b
0.22
Max
95% tile
2.2 100 600 1.6
0.8 8 225 1.0
5% tile 0.1 187 50 0.6
were no significant gender differences in tympanometry fail rates ( p = 0.22, Table 2). As previously mentioned, analyses using criteria of <0.2 mmho static admittance or width >200 daPa were also performed to allow us to compare our results to previous studies using these criteria for tympanometry failure. Fail rates using criteria of <0.2 mmho static admittance or width >200 daPa were 33% for 3 year olds and 19% for 4—5 year olds, 23% for females and 22% for males. Hmong/Asian children had the highest fail rate (37%), followed by African/ African American (14%) and White/other (10%). Logistic regression analysis including age, gender, and race/ethnicity revealed that the odds of tympanometric failure (defined as <0.3 mmho admittance or width >200 daPa) were significantly greater for 3 year olds and for Asian children after adjusting for other factors in the model (OR = 1.99, 95% CI 1.03—3.84, p = 0.042 for 3 versus 4—5 year olds, OR = 6.39, 95% CI 3.65—11.2, p < 0.001 for Asian versus non-Asian) (Table 3).
4. Discussion The purpose of this study was to determine if Asian (predominantly Hmong) children had higher rates of tympanometry failure than children of other races/ ethnicities. Our analysis demonstrated that these children had significantly more tympanometric failures, a finding that persisted after adjusting for age and gender. To our knowledge, this is the first paper comparing tympanometry results in a group of preTable 3 failure
18/72 (25) 69/131 (53) 61/135 (45)
Min
Logistic regression model for tympanometry
Factor
Odds ratio (95% CI)
Young 3 years vs. 4—5 years
1.99 (1.03, 3.84)
0.042
Gender Males vs. females
0.71 (0.42, 1.22)
0.216
6.39 (3.65, 11.2)
<0.001
Asian Asian vs. non-Asian
p-value
1526 dominantly Hmong children and children of other races. There are many possible explanations for these findings. First, genetic factors may result in anatomical differences that prevent proper drainage of the middle ear. Immunological differences may make this population more susceptible to particular organisms. Environmental factors may also play a role. For example, many immigrant families come to the United States and live with family members already established in a community. These homes often have extended families with many children. Exposure to several children on a regular basis would increase the risk of middle ear disease, as reported by previous studies [8,17—19]. Another possibility is the presence of poor general health as many Asian immigrants spend time in refugee camps before coming to the United States, where access to health care may be limited. For those who have been in the United States for a longer period of time, there may be decreased access to, or utilization of, health care due to financial difficulties or cultural concerns about Western medicine. Many Hmong people use traditional Hmong health care in which Shamans, not physicians, play an important role. The Hmong believe that Shamans communicate between the physical and spiritual world through performing rituals and sacrificing animals to keep the spirits happy and cure illness. Some Hmong people believe that surgery interferes with reincarnation or creates an access for evil sprits to enter the body and soul. Others believe that drawing blood disrupts the balance within the body that blood maintains [20]. The finding that younger children were more likely to fail tympanometry than older children is consistent with previous studies that have shown the incidence of middle ear disease decreases with age [18,21]. Otitis media (OM) is a seasonal disease. It is more prevalent in the winter than in other seasons, coincident with the peak prevalence of viral respiratory disease [22]. Several cross-sectional screening studies have reported on the prevalence of middle ear effusion using tympanometry, but only a few report data from September and October, the months in which our study was conducted. A prospective study done in the United Kingdom tested children at 3, 4, and 5 years of age [23]. Rates of OME (at least one tympanogram with admittance <0.2 mmho) ranged from 8.4% to 10.3% in September, and 13.1—17.6% in October. A study of 2—6 year olds in day care in Pittsburgh reported a similar rate of OME (7—14%) detected by tympanometry and pneumatic otoscopy in September and October [24]. In a large birth cohort studied in New Zealand, 10.5% of 5-yearold children had a flat tympanogram, examined
A.C. Meyer et al. across all seasons [25]. These studies report rates similar to those for African/African American and White/other, but markedly lower than Hmong rates in the current study. Our study has limitations. Children did not have otoscopy in conjunction with tympanometry, which would provide additional information about the middle ear and tympanic membrane. Tympanometry was done with a screening rather than a diagnostic instrument. We did not have access to social or medical history for the children to identify environmental factors or comorbid conditions that could have contributed to the development of middle ear disease. Finally, it would have been interesting to know whether the children were born in the United States or their native countries to determine if recent immigration was related to middle ear disease.
5. Conclusion Asian children were more than six times as likely to fail tympanometry as children of other races after adjusting for age and gender. This finding may be due to many factors, including genetic factors (e.g. anatomic or immunologic differences), or environmental factors (e.g. crowded living conditions) or decreased access to or utilization of health care. Further research with Hmong populations is needed to better understand the factors responsible for the increased risk of tympanometry failure in this population. With more knowledge, better care can be provided to the increasing number of Hmong pediatric patients in Minnesota.
Acknowledgements We would like to especially thank the Sight and Hearing Association for their commitment to this study and to Barbara Painter who did the screening of all the Head Start children. We also thank the Ramsey Action Program for allowing us to use the screening data for this study. We appreciate the advice and assistance of Robert Margolis Ph.D., Department of Otolaryngology, University of Minnesota, and Lisa Hunter Ph.D., Department of Communication Disorders, University of Utah.
References [1] G.W. Watters, J.E. Jones, A.P. Freeland, The predictive value of tympanometry in the diagnosis of middle ear effusion, Clin. Otolaryngol. 22 (1997) 343—345.
Tympanometry of a diverse group of preschool aged children [2] R.J. Nozza, C.D. Bluestone, D. Kardatzke, R. Bachman, Identification of middle ear effusion by aural acoustic admittance and otoscopy, Ear Hear. 15 (1994) 310—323. [3] J.L. Paradise, T.F. Campbell, C.A. Dollaghan, et al., Developmental outcomes after early or delayed insertion of tympanostomy tubes, N. Engl. J. Med. 353 (2005) 576—586. [4] US Census Bureau Website http://factfinder.census.gov/ servlet. QP-P7 race alone or in combination for American Indian, Alaska native, and for selected categories of Asian, and of native Hawaiian and other Pacific Islanders: 2000. Data set: census 2000 summary file 1 (SF-1) 100% data. Geographic area, Minnesota. [5] D. Bhoola, R. Hugo, Prevalence: outer and middle ear disorders in black and Indian preschool children from Durban, S. Afr. J. Commun. Disord. 42 (1995) 19—27. [6] J.A. Moore, Comparison of risk of conductive hearing loss among three ethnic groups of Arctic audiology patients, J. Speech Lang. Hear. Res. 42 (1999) 1311—1322. [7] H.G. Goycoolea, M.V. Goycoolea, C.R. Farfan, Racial and familial factors in otitis media. A point prevalence study on Easter Island, Arch. Otolaryngol. Head Neck Surg. 114 (1988) 147—149. [8] L. Vernacchio, S.M. Lesko, R.M. Vezina, et al., Racial/ethnic disparities in the diagnosis of otitis media in infancy, Int. J. Pediatr. Otorhinolaryngol. 68 (2004) 795—804. [9] B.P. Lanphear, R.S. Byrd, P. Auinger, C.B. Hall, Increasing prevalence of recurrent otitis media among children in the United States, Pediatrics 99 (1997) E1. [10] A.T. Curns, R.C. Holman, D.K. Shay, et al., Outpatient and hospital visits associated with otitis media among American Indian and Alaska native children younger than 5 years, Pediatrics 109 (2002) E41. [11] D.O. Robinson, D.V. Allen, Racial differences in tympanometric results, J. Speech Hear. Disord. 49 (1984) 140— 144. [12] D.O. Robinson, D.V. Allen, L.P. Root, Infant tympanometry: differential results by race, J. Speech Hear. Disord. 53 (1988) 341—346. [13] T.E. Griffith, Epidemiology of otitis media–—an interracial study, Laryngoscope 89 (1979) 22—30.
1527
[14] K.C. Pugh, H.W.K. Burke, H.M. Brown, Tympanometry measures in native and non-native Hawaiian children, Int. J. Pediatr. Otorhinolaryngol. 68 (2004) 753—758. [15] H.C. Rushton, M.C. Tong, V. Yue, P.J. Wormald, C.A. van Hasselt, Prevalence of otitis media with effusion in multicultural schools in Hong Kong, J. Laryngol. Otol. 111 (1997) 804—806. [16] American Speech-Language Hearing Association. Guidelines for Audiologic Screening. ASHA Desk Reference, vol. 4, Panel on Audiologic Assessment, Rockville, MD, 1997, pp. 74g—74q. [17] S. Bruneau, H. Ayukawa, J.F. Prouix, J.D. Baxter, K. Kost, Longitudinal observations (1987—1997) on the prevalence of middle ear disease and associated risk factors among Inuit children of Inukjuak, Nunavik, Quebec, Canada, Int. J. Circumpolar Health 60 (2001) 632—639. [18] S.A. Zeisel, J.E. Roberts, M. Burchinal, E. Neebe, F.W. Henderson, A longitudinal study of risk factors for otitis media in African American children, Matern. Child Health J. 6 (2002) 189—193. [19] A.J. Leach, Otitis media in Australian aboriginal children: an overview, Int. J. Pediatr. Otorhinolaryngol. 49 (Suppl. 1) (1999) S173—S178. [20] T.P. Lee, M.E. Pfeifer, Hmong 101. Available at: http:// www.hmongnet.org. [21] S.A. Skull, P.S. Morris, A. Yonovitz, et al., Middle ear effusion: rate and risk factors in Australian children attending day care, Epidemiol. Infect. 123 (1999) 57—64. [22] K. Daly, Definition and epidemiology of otitis media with effusion, in: J.E. Roberts, I. Wallace, F. Henderson (Eds.), Otitis Media, Language, and Learning in Young Children, Paul H Brooks Publishing, Baltimore, MD, 1997, pp. 3—41. [23] E.J. Midgley, C. Dewey, K. Pryce, A.R. Maw, ALSPAC Study Team, The frequency of otitis media with effusion in British preschool children: a guide for treatment, Clin. Otolaryngol. Allied Sci. 25 (2000) 485—491. [24] M.L. Casselbrant, L.M. Brostoff, E.I. Cantekin, et al., Otitis media with effusion in preschool children, Laryngoscope 95 (1985) 428—436. [25] D. Chalmers, I. Stewart, P. Silva, A. Mulvena, Otitis media with effusion in children—the Dunedin Study, Clin. Dev. Med. 108 (1989) 1—167.
www.elsevier.com/locate/ijporl
Tympanometry of a diverse group of preschool aged children Abby C. Meyer *, Kathryn J. Webb, Cynthia S. Davey, Kathleen A. Daly University of Minnesota, Department of Otolaryngology, MMC 396, 420 Delaware St SE, Minneapolis, MN 55455, USA Received 20 September 2005; received in revised form 15 February 2006; accepted 20 February 2006
KEYWORDS Otitis media; Tympanometry; Hmong
Summary Objective: Ethnicity has been previously described as a risk factor for middle ear disease. Little data exist on the presence of middle ear disease based on tympanometry screening comparing Asian children and children of other races. Methods: Two hundred and seventy children aged 3—5 were screened with tympanometry at six Head Start sites in St. Paul, Minnesota during the months of September and October of 2004. Gender, age, and race/ethnicity was recorded and entered into a database, along with values for canal volume, static admittance, peak pressure, and tympanometric width. Results: Criteria for abnormal tympanometry were based on American SpeechLanguage Hearing Association (ASHA) recommendations for a failed tympanogram for 1—5 year olds (admittance <0.3 mmho or width >200 daPa). There were no statistically significant differences in failure rates between males and females. There were, however, more failures for Asian (predominantly Hmong) children compared to children of other races/ethnicities after adjusting for age and gender differences (OR = 6.39, CI 3.65—11.2, p < 0.001) and for children <4-years-old compared to children 4—5-years-old after adjusting for race and gender differences (OR = 1.99, CI 1.03—3.84, p < 0.05). Conclusions: Asian children were more than six times as likely to fail tympanometry as children of other races/ethnicities. The explanation for this difference is likely to be multifactorial, and further research is needed to characterize this difference. # 2006 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
* Corresponding author. Tel.: +1 763 571 7556 E-mail address: [email protected] (A.C. Meyer).
Tympanometry is a valuable tool for the assessment of evidence of middle ear effusion. Watters et al. and Nozza et al. compared pre-operative tympanometry to surgical findings and found tympanometry
0165-5876/$ — see front matter # 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijporl.2006.02.008
1524 to be highly sensitive in predicting middle ear effusion [1,2]. Furthermore, tympanometry may help identify middle ear disease that may lead to chronic ear disease, hearing loss, and subsequent speech and language difficulties if not detected. However, recent research by Paradise et al. casts doubts on the relationship between persistent middle ear effusion and developmental outcomes at age six [3]. In recent years, the city of St. Paul, Minnesota has seen a large influx of refugees and immigrants from many Southeast Asian nations. According to the 2000 census, 41,800 Hmong (refugees and immigrants from Laos) reside in Minnesota, making it the most populous Asian ethnic group in the state [4]. The majority (58%) lives in St. Paul. As a result, many children of Hmong descent are enrolling in enrichment programs, including Head Start. In the past few years, staff from the Sight & Hearing Association (SHA), the organization that conducts both vision and hearing screening for Head Start programs in St. Paul, noticed that Hmong children had a high referral rate for abnormal tympanometry. The SHA approached researchers in the Department of Otolaryngology at the University of Minnesota with the hopes that this observation could be formally investigated. In reviewing the literature, several studies support race or ethnicity as a predisposing factor for middle ear pathology [5—14]. However, only one study has compared rates of middle ear disease in Asian and White children. Authors reported that 9.5% of White children and 1.3% of Chinese children had middle ear disease as determined with tympanometry and otoscopy [15].
2. Methods 2.1. Data collection Preschool aged children were screened with tympanometry by SHA personnel at six Ramsey Action Program Head Start sites in September and October 2004. Tympanometry was performed using a MAICO MI 26 Tympanometer/Audiometer that was routinely calibrated in August, prior to the start of screening in September. The same equipment was used at all screening sites and daily probe calibration checks were performed. A 226 Hz probe tone was presented with an intensity of 85 dB SPL measured in a 2 ml coupler. A positive to negative pressure sweep between +200 and 400 daPa was administered at a variable rate of 550 daPa/s maximum with 200 daPa/s near peak. Typical test time per ear was less than 3 s. Otologic examinations were not performed as part of the screening.
A.C. Meyer et al. Tympanogram tracings were obtained and printed by SHA personnel, who labeled them with the child’s age, gender, and race/ethnicity. At least two attempts were made to obtain a valid tympanogram. Race/ethnicity was determined by the screener from the child’s Head Start registration. The tympanogram tracings were provided to the University of Minnesota Department of Otolaryngology for interpretation and analysis. The institutional review board approved secondary analysis of anonymous tympanometry data.
2.2. Data analysis Ear canal volume (at 200 daPa), static admittance (compensated at +200 daPa), and peak pressure values were obtained from the tympanogram printout and entered into a database, along with identification number, age, gender, ethnicity, and screening site. One researcher (AM) measured tympanometric width to the nearest 1/16 in. at one-half the peak admittance and converted the value into daPa units with the following conversion: 1/16 in. is equal to 25 daPa. If the tympanogram was ‘‘noisy’’ as evidenced by upward departure from the defined peak, giving an erroneous admittance result, peak compensated admittance was also measured by hand. A 5% sample of tracings was re-measured by another researcher to check for accuracy against the data entered in the database. Criteria for abnormal tympanometry were based on American Speech-Language Hearing Association (ASHA) recommendations for a failed tympanogram for 1—5 year olds based on normative studies. These criteria were admittance <0.3 mmho or width >200 daPa [16]. Comparisons by age, race and gender were done with chi-square analyses, and logistic regression was used to construct a predictive model for tympanometry failure. To allow comparisons with previous screening studies using a flat tympanogram (defined as static admittance <0.2, or 0.1) as evidence of middle ear effusion, criteria of static admittance <0.2 mmho or width >200 daPa were also used to determine tympanometry fail rates by age, gender or race.
3. Results Among the children tested, 49% were female and 51% male; 20% were 3 years old, 54% were 4 years old, and 26% were 5 years old. Twenty-nine percent were African/African American, 39% were Asian (93% were Hmong), and 32% were Caucasian/other, which included Hispanic, Middle Eastern, and others.
Tympanometry of a diverse group of preschool aged children
1525
Table 1 Tympanometric summary statistics by ears Variable
N
Mean
S.D.
Admittance Pressure Width Canal volume
267 258 266 267
0.37 60.34 134.50 0.79
0.26 65.32 104.40 0.16
Five hundred and forty ears of 270 children were tested. Of these, 11 ears had abnormal admittance (<0.3) and abnormal ear canal volume. They were excluded from analysis. Seven of the 11 had canal volumes >1.0, suggesting a tympanostomy tube or tympanic membrane perforation. The other four had canal volumes <0.5, suggesting a blocked probe resulting in an invalid test. Of the 529 remaining ears, 39 had an unreadable tracing, so peak admittance and width could not be determined. Three children had missing data for both ears, leaving 267 children with usable tympanometric data in one or both ears. In order to base the data summaries on independent (non-correlated) observations, only data from one ear per child was selected for analysis of tympanogram measures based on the following rule: data was selected for right ear unless data for only the left ear was available. Tympanometric summary statistics by ears (one ear for each child as described above) are shown in Table 1. Of the 267 children with data for at least one ear, 130 (49%) failed tympanometry in at least one ear with admittance <0.3 mmho or width >200 daPa. Fail rates by age revealed a higher fail rate among 3 year olds (58%) than 4—5 year olds (46%); this difference was not statistically significant ( p = 0.12, Table 2). Asian children had a significantly higher fail rate than the other groups ( p < 0.001) but there Table 2 Tympanometric fail rate by age, race/ethnicity, gender Variable
N (%) failed
Age (years) 3 4 5
32/55 (58) 69/143 (48) 29/68 (43)
Race Asian African/African American Other Gender Female Male a b
3 years vs. 4—5 years. Asian vs. others.
78/105 (74) 34/90 (38)
p-value (chi-square) 0.49 0.12 a
0 385 25 0.5
<0.001 <0.001 b
0.22
Max
95% tile
2.2 100 600 1.6
0.8 8 225 1.0
5% tile 0.1 187 50 0.6
were no significant gender differences in tympanometry fail rates ( p = 0.22, Table 2). As previously mentioned, analyses using criteria of <0.2 mmho static admittance or width >200 daPa were also performed to allow us to compare our results to previous studies using these criteria for tympanometry failure. Fail rates using criteria of <0.2 mmho static admittance or width >200 daPa were 33% for 3 year olds and 19% for 4—5 year olds, 23% for females and 22% for males. Hmong/Asian children had the highest fail rate (37%), followed by African/ African American (14%) and White/other (10%). Logistic regression analysis including age, gender, and race/ethnicity revealed that the odds of tympanometric failure (defined as <0.3 mmho admittance or width >200 daPa) were significantly greater for 3 year olds and for Asian children after adjusting for other factors in the model (OR = 1.99, 95% CI 1.03—3.84, p = 0.042 for 3 versus 4—5 year olds, OR = 6.39, 95% CI 3.65—11.2, p < 0.001 for Asian versus non-Asian) (Table 3).
4. Discussion The purpose of this study was to determine if Asian (predominantly Hmong) children had higher rates of tympanometry failure than children of other races/ ethnicities. Our analysis demonstrated that these children had significantly more tympanometric failures, a finding that persisted after adjusting for age and gender. To our knowledge, this is the first paper comparing tympanometry results in a group of preTable 3 failure
18/72 (25) 69/131 (53) 61/135 (45)
Min
Logistic regression model for tympanometry
Factor
Odds ratio (95% CI)
Young 3 years vs. 4—5 years
1.99 (1.03, 3.84)
0.042
Gender Males vs. females
0.71 (0.42, 1.22)
0.216
6.39 (3.65, 11.2)
<0.001
Asian Asian vs. non-Asian
p-value
1526 dominantly Hmong children and children of other races. There are many possible explanations for these findings. First, genetic factors may result in anatomical differences that prevent proper drainage of the middle ear. Immunological differences may make this population more susceptible to particular organisms. Environmental factors may also play a role. For example, many immigrant families come to the United States and live with family members already established in a community. These homes often have extended families with many children. Exposure to several children on a regular basis would increase the risk of middle ear disease, as reported by previous studies [8,17—19]. Another possibility is the presence of poor general health as many Asian immigrants spend time in refugee camps before coming to the United States, where access to health care may be limited. For those who have been in the United States for a longer period of time, there may be decreased access to, or utilization of, health care due to financial difficulties or cultural concerns about Western medicine. Many Hmong people use traditional Hmong health care in which Shamans, not physicians, play an important role. The Hmong believe that Shamans communicate between the physical and spiritual world through performing rituals and sacrificing animals to keep the spirits happy and cure illness. Some Hmong people believe that surgery interferes with reincarnation or creates an access for evil sprits to enter the body and soul. Others believe that drawing blood disrupts the balance within the body that blood maintains [20]. The finding that younger children were more likely to fail tympanometry than older children is consistent with previous studies that have shown the incidence of middle ear disease decreases with age [18,21]. Otitis media (OM) is a seasonal disease. It is more prevalent in the winter than in other seasons, coincident with the peak prevalence of viral respiratory disease [22]. Several cross-sectional screening studies have reported on the prevalence of middle ear effusion using tympanometry, but only a few report data from September and October, the months in which our study was conducted. A prospective study done in the United Kingdom tested children at 3, 4, and 5 years of age [23]. Rates of OME (at least one tympanogram with admittance <0.2 mmho) ranged from 8.4% to 10.3% in September, and 13.1—17.6% in October. A study of 2—6 year olds in day care in Pittsburgh reported a similar rate of OME (7—14%) detected by tympanometry and pneumatic otoscopy in September and October [24]. In a large birth cohort studied in New Zealand, 10.5% of 5-yearold children had a flat tympanogram, examined
A.C. Meyer et al. across all seasons [25]. These studies report rates similar to those for African/African American and White/other, but markedly lower than Hmong rates in the current study. Our study has limitations. Children did not have otoscopy in conjunction with tympanometry, which would provide additional information about the middle ear and tympanic membrane. Tympanometry was done with a screening rather than a diagnostic instrument. We did not have access to social or medical history for the children to identify environmental factors or comorbid conditions that could have contributed to the development of middle ear disease. Finally, it would have been interesting to know whether the children were born in the United States or their native countries to determine if recent immigration was related to middle ear disease.
5. Conclusion Asian children were more than six times as likely to fail tympanometry as children of other races after adjusting for age and gender. This finding may be due to many factors, including genetic factors (e.g. anatomic or immunologic differences), or environmental factors (e.g. crowded living conditions) or decreased access to or utilization of health care. Further research with Hmong populations is needed to better understand the factors responsible for the increased risk of tympanometry failure in this population. With more knowledge, better care can be provided to the increasing number of Hmong pediatric patients in Minnesota.
Acknowledgements We would like to especially thank the Sight and Hearing Association for their commitment to this study and to Barbara Painter who did the screening of all the Head Start children. We also thank the Ramsey Action Program for allowing us to use the screening data for this study. We appreciate the advice and assistance of Robert Margolis Ph.D., Department of Otolaryngology, University of Minnesota, and Lisa Hunter Ph.D., Department of Communication Disorders, University of Utah.
References [1] G.W. Watters, J.E. Jones, A.P. Freeland, The predictive value of tympanometry in the diagnosis of middle ear effusion, Clin. Otolaryngol. 22 (1997) 343—345.
Tympanometry of a diverse group of preschool aged children [2] R.J. Nozza, C.D. Bluestone, D. Kardatzke, R. Bachman, Identification of middle ear effusion by aural acoustic admittance and otoscopy, Ear Hear. 15 (1994) 310—323. [3] J.L. Paradise, T.F. Campbell, C.A. Dollaghan, et al., Developmental outcomes after early or delayed insertion of tympanostomy tubes, N. Engl. J. Med. 353 (2005) 576—586. [4] US Census Bureau Website http://factfinder.census.gov/ servlet. QP-P7 race alone or in combination for American Indian, Alaska native, and for selected categories of Asian, and of native Hawaiian and other Pacific Islanders: 2000. Data set: census 2000 summary file 1 (SF-1) 100% data. Geographic area, Minnesota. [5] D. Bhoola, R. Hugo, Prevalence: outer and middle ear disorders in black and Indian preschool children from Durban, S. Afr. J. Commun. Disord. 42 (1995) 19—27. [6] J.A. Moore, Comparison of risk of conductive hearing loss among three ethnic groups of Arctic audiology patients, J. Speech Lang. Hear. Res. 42 (1999) 1311—1322. [7] H.G. Goycoolea, M.V. Goycoolea, C.R. Farfan, Racial and familial factors in otitis media. A point prevalence study on Easter Island, Arch. Otolaryngol. Head Neck Surg. 114 (1988) 147—149. [8] L. Vernacchio, S.M. Lesko, R.M. Vezina, et al., Racial/ethnic disparities in the diagnosis of otitis media in infancy, Int. J. Pediatr. Otorhinolaryngol. 68 (2004) 795—804. [9] B.P. Lanphear, R.S. Byrd, P. Auinger, C.B. Hall, Increasing prevalence of recurrent otitis media among children in the United States, Pediatrics 99 (1997) E1. [10] A.T. Curns, R.C. Holman, D.K. Shay, et al., Outpatient and hospital visits associated with otitis media among American Indian and Alaska native children younger than 5 years, Pediatrics 109 (2002) E41. [11] D.O. Robinson, D.V. Allen, Racial differences in tympanometric results, J. Speech Hear. Disord. 49 (1984) 140— 144. [12] D.O. Robinson, D.V. Allen, L.P. Root, Infant tympanometry: differential results by race, J. Speech Hear. Disord. 53 (1988) 341—346. [13] T.E. Griffith, Epidemiology of otitis media–—an interracial study, Laryngoscope 89 (1979) 22—30.
1527
[14] K.C. Pugh, H.W.K. Burke, H.M. Brown, Tympanometry measures in native and non-native Hawaiian children, Int. J. Pediatr. Otorhinolaryngol. 68 (2004) 753—758. [15] H.C. Rushton, M.C. Tong, V. Yue, P.J. Wormald, C.A. van Hasselt, Prevalence of otitis media with effusion in multicultural schools in Hong Kong, J. Laryngol. Otol. 111 (1997) 804—806. [16] American Speech-Language Hearing Association. Guidelines for Audiologic Screening. ASHA Desk Reference, vol. 4, Panel on Audiologic Assessment, Rockville, MD, 1997, pp. 74g—74q. [17] S. Bruneau, H. Ayukawa, J.F. Prouix, J.D. Baxter, K. Kost, Longitudinal observations (1987—1997) on the prevalence of middle ear disease and associated risk factors among Inuit children of Inukjuak, Nunavik, Quebec, Canada, Int. J. Circumpolar Health 60 (2001) 632—639. [18] S.A. Zeisel, J.E. Roberts, M. Burchinal, E. Neebe, F.W. Henderson, A longitudinal study of risk factors for otitis media in African American children, Matern. Child Health J. 6 (2002) 189—193. [19] A.J. Leach, Otitis media in Australian aboriginal children: an overview, Int. J. Pediatr. Otorhinolaryngol. 49 (Suppl. 1) (1999) S173—S178. [20] T.P. Lee, M.E. Pfeifer, Hmong 101. Available at: http:// www.hmongnet.org. [21] S.A. Skull, P.S. Morris, A. Yonovitz, et al., Middle ear effusion: rate and risk factors in Australian children attending day care, Epidemiol. Infect. 123 (1999) 57—64. [22] K. Daly, Definition and epidemiology of otitis media with effusion, in: J.E. Roberts, I. Wallace, F. Henderson (Eds.), Otitis Media, Language, and Learning in Young Children, Paul H Brooks Publishing, Baltimore, MD, 1997, pp. 3—41. [23] E.J. Midgley, C. Dewey, K. Pryce, A.R. Maw, ALSPAC Study Team, The frequency of otitis media with effusion in British preschool children: a guide for treatment, Clin. Otolaryngol. Allied Sci. 25 (2000) 485—491. [24] M.L. Casselbrant, L.M. Brostoff, E.I. Cantekin, et al., Otitis media with effusion in preschool children, Laryngoscope 95 (1985) 428—436. [25] D. Chalmers, I. Stewart, P. Silva, A. Mulvena, Otitis media with effusion in children—the Dunedin Study, Clin. Dev. Med. 108 (1989) 1—167.