Is Otitis Media With Effusion Almost Always Accompanying Cleft Palate in Children?: The Experience of 319 Asian Patients

Transcription

1 The Laryngoscope VC 2011 The American Laryngological, Rhinological and Otological Society, Inc. Is Otitis Media With Effusion Almost Always Accompanying Cleft Palate in Children?: The Experience of 319 Asian Patients Yi-Wei Chen, MD; Kuo-Ting Philip Chen, MD; Po-Hung Chang, MD; Jen-Liang Su, MD; Chi-Che Huang, MD; Ta-Jen Lee, MD Objectives/Hypothesis: To evaluate the incidence of concomitant otitis media with effusion (OME) in children with cleft palate in Taiwan and the reliability of preoperative tympanometry. Study Design: Individual prospective cohort study. Methods: We included 319 patients who underwent palatoplasty for cleft palate from 2005 to All received tympanometry 1 day before surgery, and myringotomy was performed before palatoplasty. Grommet was inserted if there was OME, the amount and content of which were graded and recorded. All data, including the tympanometry results, were analyzed for statistical significance. Results: The incidence of OME accompanying cleft palate during palatoplasty was 71.92% in Asian patients, which was lower than in previous studies. The content of OME was serous fluid in 47.8%, mucoid in 33.1%, and mucopus in 19.1%. Type-B tympanogram had high sensitivity (0.956) in all age groups. However, its specificity was poor in children younger than 9 months (0.375), fair in children aged 9 to 14 months (0.582), and good in children older than 14 months (0.857). Conclusions: The incidence of OME accompanying cleft palate has decreased in Asian patients in the past 5 years. Tympanometry is not a very reliable tool for evaluating OME in children with cleft palate younger than 14 months, especially those younger than 9 months. Meticulous examination of the middle ear during palatoplasty is necessary for making a definite diagnosis and deciding on the use of a grommet. Key Words: Otitis media with effusion, cleft palate, tympanometry, sensitivity, specificity. Level of Evidence: 1b Laryngoscope, 122: , 2012 INTRODUCTION Since Paradise et al. indicated that otitis media with effusion (OME) was almost universally present in children with cleft palate in 1969, 1 various incidences of OME have been reported. However, the precise incidence could be established based on the result of myringotomy. In 1988, Grant et al. designed a multicenter prospective study of 55 children with cleft palate in London and revealed OME in 97% of these patients. 2 Valtonen et al. also reported that 98% of cleft palate patients had accompanying OME found in 51 cases during the period 1983 to 1993 in Finland. 3 The incidence of cleft palate also shows ethnic variations. It is generally thought that populations of Asian and Native North American descent have the highest incidence, with Caucasian populations having intermediate incidence, and African populations having the lowest From the Department of Otolaryngology, Chang Gung Memorial Hospital, Taipei, Taiwan (Y.-W.C., P.-H.C., J.-L.S., C.-C.H., T.-J.L.); Chang Gung University, Taoyuan, Taiwan (Y.-W.C., K.-T.P.C., P.-H.C., J.-L.S., C.-C.H., T.-J.L.); and Department of Craniofacial Center (K.-T.P.C.), Chang Gung Memorial Hospital, Taipei, Taiwan. Editor s Note: This Manuscript was accepted for publication July 25, The authors have no funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Ta-Jen Lee, MD, Department of Otolaryngology, Chang Gung Memorial Hospital, 5 Fu-Shin Street, Kueishan, Taoyuan, 333, Taiwan. yiwei0819@gmail.com DOI: /lary incidence. 4 However, there is no evidence-based study showing the precise incidence of OME accompanying cleft palate in Asian patients or any ethnic difference between Eastern and Western populations. Thus, this large-scale prospective study has been designed to determine the incidence and nature of OME in children with cleft palate in Taiwan in the past 5 years. In addition, tympanometry has been the most common and widely accepted tool for assessing the presence of fluid in the middle ear. Type-B tympanogram has high sensitivity in predicting OME and good specificity in children aged 1 to 15 years. 5 But in children younger than 1 year old, the results may be considerably different. Hence, this study also aimed to examine the reliability of type-b tympanogram in predicting OME in patients with cleft palate for preoperative evaluation. MATERIALS AND METHODS From June 1, 2005, to May 31, 2010, patients with cleft palate diagnosed and treated in the craniofacial department of Chang-Gung Memorial Hospital were included. Patients received cheiloplasty at the age of 3 months and palatoplasty at 1 year old. Cerumen was removed as much as possible, and tympanogram (AT235h Impedance audiometer; Interacoustics, Assens, Denmark) was then performed 1 day prior to palatoplasty. All of the patients provided written informed consent. An ENT doctor first performed middle ear surgery with general anesthesia, followed by palatoplasty by a plastic surgeon. Residual cerumen in the external auditory canal was

2 TABLE I. Sensitivity, Specificity, and Predictive Values of Type-B Tympanogram. Sensitivity of type-b tympanogram for OME ¼ Specificity of type-b tympanogram for OME ¼ Positive predictive values of type-b tympanogram for OME ¼ Negative predictive values of type-b tympanogram for OME ¼ OME ¼ otitis media with effusion. Fig. 1. Age at myringotomy (in months). cleared at the beginning, and the condition of the eardrum was recorded as retracted, dull, bulging, or normal. In ears with type-b tympanogram, a minimal myringotomy wound was created first in the inferior portion of the eardrum. If OME existed, the wound was lengthened and the effusion content cleared. A 1.0-mm grommet (Rock pediatric vent tube, fluoroplastic, 1-mm internal diameter, Gyrus ACMI, Inc. Southborough, MA) was then inserted. In ears with type-a or type-c tympanogram, fine-needle aspiration of the middle ear was performed to evaluate any middle ear effusion. Grommet insertion was only done if there was a positive finding. The amount and content of OME were also recorded. The content of effusion was classified as serous, mucoid, or mucopus, and the amount was categorized into grades 1 to 3 (grade 1 [little]: effusion not filling the middle ear cavity; grade 2 [moderate]: effusion filling the middle ear cavity; and grade 3 (full): effusion filling the middle ear cavity and making the eardrum bulge). One surgeon performed all of the middle ear surgeries and evaluations. Statistical analyses included the v 2 test with Fisher exact tests to compare the categorical presence or absence of pathologic findings in cross tables. P <.05 was regarded as statistically significant. The SPSS for Windows 10.0 statistical packet program (SPSS Inc., Chicago, IL) was used for data analysis. RESULTS In the past 5 years, 319 patients received palatoplasty with middle ear surgery (Fig. 1). There were 165 males and 154 females, and their mean age at surgery was 12.3 months. Three patients with unilateral microtia and one with narrowing of the external auditory canal were excluded. Of the 634 ears included in the analysis, 456 ears had OME and 178 had dry contents. The incidence of OME accompanying cleft palate was 71.92% in surgery. There were 178 ears without OME that received palatoplasty, of which 77 had type-a tympanogram, 72 had type B, and 29 had type C. Mean age was 11.8 months. OME was noted in 456 ears, of which 11 showed type-a tympanogram, 436 showed type B, and nine showed type C. The mean age was 12.4 months. The sensitivity, specificity, positive predictive value, and negative predictive value of type-b tympanogram in patients with cleft palate are shown in Table I. Sensitivity was and specificity was When sensitivity and specificity were calculated according to patient age (Fig. 2), sensitivity was approximately 0.95 in each age group, but specificity was extremely low in patients younger than 9 months old and high in those older than 14 months. The content of OME was serous fluid in 47.8% of ears, mucoid in 33.1%, and mucopus in 19.1%. The amount of OME was graded as little, moderate, and full in 21.5%, 63.4%, and 15.1% of ears, respectively. The contents and amount of OME in each type of tympanogram are shown in Figure 3 and Figure 4. However, 25.1% of the patients had different contents of OME in both ears. DISCUSSION Infants with unrepaired cleft palate reportedly have a universal incidence of a middle ear effusion, which is associated with a functional obstruction of the Eustachian tube as a result of anatomic and physiologic variations of the tensor veli palatini muscle. 6 Doyle et al. also assessed Eustachian tube function in children with cleft palate before and after palatoplasty and reported that 70% had dysfunction characterized as constriction. 7 Asian populations have the highest birth prevalence of oral-facial clefts in the world. 8 However, the incidence of accompanying OME is lower than in other countries, as shown by the current study. It is not clear whether differences in prevalence are due to differences in gene frequencies, geographic clustering, or environment. In studies of patients with obstructive sleep apnea syndrome, differences of craniofacial and upper airway structures are noted in various ethnic groups. 9 The Chinese have significantly small maxilla and mandibles, Fig. 2. Sensitivity and specificity of type B tympanogram according to age. M ¼ months. 221

3 Fig. 3. The content of otitis media with effusion in each type of tympanogram. more severe mandibular retrognathism, increased total and upper facial heights, and steeper and shorter anterior cranial bases than Caucasians. 10 Therefore, interethnic cephalometric differences may account for the lower incidence of accompanying OME in this study. On the other hand, data from other studies have been collected 20 years ago, and medicine has made vast advances in recent decades. The role of preventive medicine is interesting. Pneumococcal vaccination has been proven to largely reduce the rate of acute otitis media and related complications from any cause in recent years, 11 and the influenza vaccine has been effective in reducing acute otitis media and OME episodes in 6- to 60-monthold day-care children. 12 Moreover, early and aggressive treatment reduces the incidence of complications resulting from acute infectious diseases. However, in children with a cleft palate, the clinical picture of OME from birth is very different from that found in children without a cleft palate. 2 OME in the cleft palate population is often treated as a different clinical entity altogether, and its relation to infection is still controversial. 2,13 We have a study under way investigating the relationship between vaccination and OME in patients with cleft palate. We believe that the role of vaccination in OME of cleft palate children will be clarified in the future. There is debate regarding the use of grommets in patients with cleft palate. Robson et al. performed a retrospective study in 1992 and concluded that there was no difference in speech development between cleft palate patients treated with grommets for OME and those Fig. 4. The amount of otitis media with effusion in each type of tympanogram. 222

4 untreated. 14 Sheahan et al. also indicated that some sequelae, like tympanic membrane scarring, atelectasis and perforation, and the development of chronic otitis media were found in patients with cleft palate who received grommet insertion for OME. 15 On the other hand, Valtonen et al. believed that early tympanostomy ensured normal hearing during the critical years of language, speech, and cognitive development and maintained the development of an aerated mastoid. 3 Greig et al. indicated that overall, parents were pleased with improvements in speech, language, and hearing after grommet insertion. 16 In general, most studies suggest grommet insertion as a treatment of choice for OME in children with cleft palate. But in those without OME, tympanostomy is seldom used as a preventive measure. The tool used for evaluating OME before surgery is important. Tympanometry is the best clinical test for the presence or absence of OME. Watters et al. evaluated 501 children (age range: 11 months to 15 years) receiving surgery for possible OME and showed that type-b tympanogram had high sensitivity (0.91) and good specificity (0.79) in predicting OME. 5 In the current study, there is also high sensitivity (0.956), but specificity is only fair (0.596). The main reason for this difference is the age distribution of the study group. The average age of patients in the current study is 12.3 months, obviously younger than those of Watters et al. Although tympanometry has proved to be important for detecting effusion and other middle ear problems, it produces differently shaped tympanograms in infants compared to adults and older children Alaerts et al. indicated that conventional 226-Hz tympanometry is more appropriate if used in children from the age of 9 months. For children younger than 9 months, 1,000-Hz tympanometry is necessary because the conventional 226-Hz tympanometry is not so reliable. 20 The current study shows similar results. When tympanometry was analyzed in terms of age, the specificity of type-b tympanogram was poor (0.375) in children younger than 9 months. On the other hand, specificity was fair (0.617) in children older than 9 months (P ¼.06 between these two groups). Furthermore, children older than 14 months had good specificity (0.857), which was identical to the study by Watters et al. Thus, patients with cleft palate can be divided into three age groups when they receive tympanometry for evaluating OME. Although the sensitivity of the type-b tympanogram is almost the same between these three age groups, specificity is poor (0.375) in children younger than 9 months, fair (0.582) in children between 9 and 14 months, and good (0.857) in children older than 14 months. The difference in specificity has statistical significance (P <.05). Taking predictive value into consideration, the positive predictive value of a type-b tympanogram for OME is in cleft palate patients older than 14 months. This drops to and in those aged between 9 and 14 months and those younger than 9 months, respectively. Narrowing of the external ear canal, difficulties in clearing cerumen, and uncooperative children may account for differences in specificities and positive predictive values of type-b tympanogram. When children are younger, the aforementioned factors will be more aggravated and affect the results of tympanometry. Although type-a tympanogram had high specificity (0.975) and positive predictive value (0.875) in our study, there were still 11 ears that had type-a tympanogram but had OME. Gates et al. have suggested that a shallow type-a curve is not incompatible with an effusion, and this may explain some of the ears with type-a tympanogram that did prove to have OME. 21 Furthermore, we found that in ears with OME having normal tympanogram, 73% (8 of 11) had serous content and 64% (7 of 11) had little effusion. In ears with OME, when we combined both factors into consideration, little serous effusion was noted in 54.5% (6 of 11) of ears with type-a tympanogram, but the same type of effusion was only found in 15.8% (69 of 436) of ears with type-b tympanogram (P <.01). Therefore, some ears with OME, while having a normal tympanogram, may reflect a small amount of effusion that is usually composed of serous content. Because the current study shows that the tympanogram is not so reliable for evaluating OME in patients with cleft palate younger than 14 months, we should pay more attention to deciding about the use of the grommet during palatoplasty. Because of the poor specificity of type-b tympanogram of cleft palate patients younger than 9 months, a meticulous examination of the middle ear condition in the operating room when patients receive palatoplasty is necessary to determine the necessity of grommet insertion. In ears with type-b tympanogram, a minimal wound of myringotomy should be created first to ensure the existence of OME. The wound may be lengthened later and a ventilation tube inserted. In patients between 9 and 14 months old, even though the specificity is fair, the same procedure is also suggested because there are still as many as 15% of ears with type-b tympanogram without OME in this age group. Preoperative explanations should include the uncertainty of tympanometry and the necessity of middle ear examination intraoperatively to determine the need for a grommet. Good communication with the parents before surgery regarding this condition is very important. CONCLUSION In this study, the incidence of OME was 71.92% in Asian patients with cleft palate, which is lower than in previous studies. Remarkable advances in medications in recent years, as well as ethnic differences between Eastern and Western populations, may contribute to this difference. Tympanogram is not a very reliable tool for evaluating OME in patients with cleft palate younger than 14 months, especially in those younger than 9 months. Meticulous examination of the middle ear during palatoplasty is necessary for determining the diagnosis and for deciding on the use of a grommet. BIBLIOGRAPHY 1. Paradise JL, Bluestone CD, Felder H. The universality of otitis media in 50 infants with cleft palate. Pediatrics 1969;44:

5 2. Grant HR, Quiney RE, Mercer DM, Lodge S. Cleft palate and glue ear. Arch Dis Child 1988;63: Valtonen H, Dietz A, Qvarnberg Y. Long-term clinical, audiologic, and radiologic outcomes in palate cleft children treated with early tympanostomy for otitis media with effusion: A controlled prospective study. Laryngoscope 2005;115: Gorlin RJ, Cohen MM, Hennekam RC. Syndromes of the Head and Neck. (Oxford Monographs on Medical Genetics.) Oxford, UK: Oxford University Press; Watters GWR, Jones JE, Freeland AP. The predictive value of tympanometry in the diagnosis of middle ear effusion. Clin Otolaryngol 1997;22: Muntz HR. An overview of middle ear disease in cleft palate children. Facial Plast Surg 1999;39: Doyle WJ, Reilly JS, Jardini L, Rovnak S. Effect of palatoplasty on the function of the Eustachian tube in children with cleft palate. Cleft Palate J 1986;23: Cooper ME, Ratay JS, Marazita ML. Asian oral-facial cleft birth prevalence. Cleft Palate Craniofac J 2006;43: Villaneuva ATC, Buchanan PR, Yee BJ, Grunstein RR. Ethnicity and obstructive sleep apnea. Sleep Med Rev 2005;9: Liu Y, Lowe AA, Zeng X, Fu M, Fleetham JA. Cephalometric comparisons between Chinese and Caucasian patients with obstructive sleep apnea. Am J Orthod Dentofacial Orthop 2000;117: Eskola J, Kilpi T, Palmu A, et al. Efficacy of a pneumococcal conjugate vaccine against acute otitis media. N Engl J Med 2001;344: Ozgur SK, Beyazova U, Kemaloglu YK, et al. Effectiveness of inactivated influenza vaccine for prevention of otitis media in children. Pediatr Infect Dis J 2006;25: Sheahan P, Miller I, Sheahan JN, Earley MJ, Blayney AW. Incidence and outcome of middle ear disease in cleft lip and/or cleft palate. Int J Pediatr Otorhinolaryngol 2003;67: Robson AK, Blanshard JD, Jones K, et al. A conservative approach to the management of otitis media with effusion in cleft palate children. J Laryngol Otol 1992;106: Sheahan P, Blayney AW, Sheahan JN, Earley MJ. Sequelae of otitis media with effusion among children with cleft lip and/or cleft palate. Clin Otolaryngol 2002;27: Greig AVH, Papesch ME, Rowsell AR, Phil D. Parental perceptions of grommet insertion in children with cleft palate. J Laryngol Otol 1999; 113: Keith RW. Impedance audiometry with neonates. Arch Otolaryngol 1973; 97: Sprague BH, Wiley TL, Goldstein R. Tympanometric and acoustic-reflex studies in neonates. J Speech Hear Res 1985;28: Holte L, Margolis RH, Cavanaugh RM. Developmental changes in multifrequency tympanograms. Audiology 1991;30: Alaerts J, Luts H, Wouters J. Evaluation of middle ear function in young children: Clinical guidelines for the use of 226- and 1000-Hz tympanometry. Otol Neurotol 2007;28: Gates GA, Avery CA, Cooper JC, Hearne EM, Holt GR. Predictive value of tympanometry in middle ear effusion. Ann Otol Rhinol Laryngol 1986; 95: