Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Sound localization wikipedia , lookup
Lip reading wikipedia , lookup
Auditory system wikipedia , lookup
Hearing loss wikipedia , lookup
Noise-induced hearing loss wikipedia , lookup
Sensorineural hearing loss wikipedia , lookup
Audiology and hearing health professionals in developed and developing countries wikipedia , lookup
The Laryngoscope C 2012 The American Laryngological, V Rhinological and Otological Society, Inc. Tympanic Membrane Retraction: An Endoscopic Evaluation of Staging Systems Adrian L. James, DM, FRCS(ORL-HNS); Blake C. Papsin, MD, FRCSC; Keith Trimble, MPhil, FRCS(ORL-HNS); James Ramsden, MA, FRCS(ORL-HNS); Nadarajah Sanjeevan, MD; Neil Bailie, PhD, FRCS(ORL-HNS); Neil K. Chadha, MPHe, FRCS(ORL-HNS) Objectives/Hypothesis: The objectives of this work were to assess inter- and intraobserver variability of different staging systems for tympanic membrane (TM) retraction using otoendoscopy in children at risk of retraction from cleft palate, to compare hearing level with stage of retraction, and to propose optimum characteristics for monitoring TM retraction with endoscopy. Study Design: Cross-sectional study. Methods: Endoscopic images of 245 TMs of children with cleft palate (mean age, 13.0 years) were assessed on two separate occasions by six observers using the Sade and Erasmus staging systems for pars tensa retraction and Tos system for pars flaccida retraction. Intra- and interobserver agreements were calculated. Extent of TM retraction was compared with hearing threshold. TMs with middle ear effusion, tympanostomy tubes, or perforation were excluded. Results: A total of 108 ear drums (44%) were rated as having pars tensa and/or flaccida retraction. Intraobserver agreement was fair to moderate (kappa ¼ 0.3–0.37, P < .001) for the different staging systems and interobserver agreement slight to moderate (0.18–0.41 P < .001). Conductive hearing loss (four-tone average air-bone gap >25 dB HL) was present in 11 ears (15%). No correlation between hearing threshold and retraction stage was found. Isolated tensa retraction onto the promontory increased hearing threshold more than retraction involving the incus (P ¼ .02; analysis of variance). Conclusions: Endoscopic image capture may provide a clear objective record of TM retraction, but current staging systems have unsatisfactory reliability when applied to such images, and retraction stage correlates poorly with hearing threshold. Modification of retraction assessment to improve validity and clinical relevance is proposed. Key Words: Tympanic membrane retraction, staging system, cleft palate, hearing loss. Level of Evidence: 2c. Laryngoscope, 122:1115–1120, 2012 INTRODUCTION Retraction of the tympanic membrane (TM) into the middle ear can cause significant conductive hearing loss, particularly as a result of ossicular erosion.1 Accumulation of desquamated keratin within the retraction pocket (i.e., From the Department of Otolaryngology–Head and Neck Surgery (A.L.J., B.C.P., N.S.), University of Toronto, Toronto, Ontario, Canada; Department of Ear, Nose, and Throat Surgery (K.T.), Royal Belfast Hospital for Sick Children, Belfast, United Kingdom; Ear, Nose, and Throat Department (J.R.), John Radcliffe Hospital, Oxford, United Kingdom; Department of Ear, Nose, and Throat Surgery (N.B.), Royal Victoria Hospital, Belfast, United Kingdom; and Division of Otolaryngology–Head and Neck Surgery (N.K.C.), British Columbia Children’s Hospital, Vancouver, British Columbia, Canada. Editor’s Note: This Manuscript was accepted for publication December 21, 2011. Presented at the 114th Triological Society Meeting, Chicago, Illinois, U.S.A., April 27– May 1, 2011. This work was performed entirely at the Department of Otolaryngology–Head and Neck Surgery, Hospital for Sick Children, Toronto, Ontario, Canada. The authors have no funding, financial relationships, or conflicts of interest to disclose. Send correspondence to Adrian L. James, DM, FRCS(ORL-HNS), Department of Otolaryngology–Head and Neck Surgery, Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, M5G 1X8, Canada. E-mail: [email protected] DOI: 10.1002/lary.23203 Laryngoscope 122: May 2012 cholesteatoma) predisposes to further infective and destructive consequences within and occasionally beyond the temporal bone. Ossicular erosion and cholesteatoma are relatively rare consequences, and are therefore hard to predict in most cases of TM retraction.2 Also unpredictably, some TM retractions resolve spontaneously.2,3 These variable outcomes make careful observation and monitoring an important part of the management of TM retraction. Several staging systems have been developed and are widely used for this purpose.2,4–6 However, there have been few attempts to validate these systems by making repeated assessments with the same or multiple observers,7 which is of concern because many characteristics of these systems have the potential to interfere with their reliability. For example, subjective criteria, such as extent of bone erosion in pars flaccida retraction (Tos stage 3 vs. 44) or depth of pars tensa retraction (Erasmus stage 3 vs. 46), may be prone to variability in interpretation, as may assessment of adherence of the TM to the promontory (Sade stage 3 vs. 42). Alternative interpretations have lead to changes in the descriptions of some systems.8,9 Such discrepancies, coupled with limitations in accuracy of recall by the observer,10 may also confound reliability. Furthermore, consensus on correlation between stage of retraction and disease process or hearing threshold has not been established.6 James et al.: Evaluation of Ear Drum Retraction 1115 Although endoscopy does not allow depth perception as readily as the stereoscopic view from a microscope, otoendoscopic images are now commonly used to record and monitor TM retraction. The validity of applying conventional staging systems to such images has not been widely assessed. To determine an optimal system for clinical and research classification of TM retraction on endoscopic TM images, we assessed the validity of existing systems2,4,6 and compared them with audiometric measures. Analysis of these findings and additional descriptors of TM retraction were used to identify reliable and clinically useful indicators for management of TM retraction. The study sample was selected from a consecutive series of children with cleft palate to provide a clinically representative range of TM retraction severity from a population at increased risk of retraction. Approval for this study was granted by the hospital’s research ethics board. MATERIALS AND METHODS Children with cleft palate have a routine multidisciplinary review at our regional cleft palate referral center at approximately five yearly intervals following repair of their cleft at around 1 year of age. A consecutive cross-sectional sample of 143 children attending this clinic over the course of a year was selected for this study. All patients received a pure-tone audiogram and digital photography of both TMs with a 0 4-mm or 2.7-mm otoendoscope, after microsuction of meatal wax and keratin debris if necessary. The otoendoscope was positioned and focused to include a close-up view of the entire TM if possible. Images were displayed on a dedicated PC with an image capture card and stored on a hard disk drive for later retrieval. Demographic details of each patient were collected. Ears with tympanostomy tubes in situ, TM perforation, or a history of any other surgery to the TM were excluded. Each ear was assigned a random study number, and the TM image was stored for viewing on a computer monitor. The complete series was reviewed by an experienced pediatric otologist and dichotomized into two groups: A) those with clearly normal and B) those with questionable or certain retraction. Images of ears with concomitant conductive hearing loss from group A, plus a random sample of around 20% of the remaining normal images from group A, were combined with those from group B to comprise a study sample for further assessment. This study sample was assessed on two separate occasions by six otolaryngologists with an interest in otology (two otologists, two pediatric otologists, and two pediatric otolaryngologists, all fellowship trained). All assessments were blinded to hearing and demographic details. The TMs were assigned to a stage of retraction with the systems developed by Sade2,8 and the Erasmus group6 for the pars tensa, and Tos for the pars flaccida.4 A modified version of the Sade system was used for this study, because the sole distinguishing factor in the original description between Sade stages 3 and 4 is whether the TM retraction is adherent. This important dynamic distinction cannot be made with a single endoscopic image. Therefore, a later modification of the staging system was used based on diagrammatic representation of stage 4, in which the TM is retracted more deeply to line the mesotympanum.8 A copy of the illustrations and descriptions was provided to the observers on both occasions to reduce error from misapplication of the systems. A value of zero was given in the absence of any visible retraction. Following analysis of these findings, a list of characteristics of pars tensa retraction was compiled in an attempt to Laryngoscope 122: May 2012 1116 identify more clinically relevant features of retraction that could be assessed reliably. Six observers twice reinspected a subset of 43 TMs representing the range of retractions in the larger sample and rated the pars tensa according to these criteria: (0) normal; (1a) retraction, no contact to middle ear structures; (1b) retraction touching promontory; (2a) retraction touching incus; (2b) retraction enveloping incus; (3a) retraction eroding incus (partial); (3b) retraction eroding incus (complete); (4) retraction disappears out of site; (5) keratin accumulation medial to annulus (potentially removable permeatus); (6) granulation tissue at retraction; (7) cholesteatoma (keratin accumulation that cannot be cleaned permeatus). The intra- and interobserver reliability for each staging systems was the primary outcome measure, which was determined by calculating the kappa statistic. The kappa statistic values were interpreted for reliability using the criteria described by Landis and Koch.11 The mean degree of hearing loss for each stage of each system was then calculated using the mode average stage for the samples. Hearing loss was defined as a mean four-tone average air-bone gap (ABG)12 of >25 dB HL. Data from ears with middle ear effusion were excluded. For the purposes of this analysis, a retracted TM was defined as a staging score above zero by three or more observers on one or more occasions. The Spearman rank order correlation was then used to compare the hearing threshold and stages to explore any significant differences in hearing between stages. RESULTS The median age of the 143 children reviewed was 13 years (range, 4–18 years). There were 41 ears excluded for having a perforation (eight), tympanostomy tube in situ (21), previous tympanoplasty or tympanomastoidectomy (three), or unclear photograph of the pars tensa (nine), giving a total sample of 245 ears. Of these, 76 (31%) were found to have a retraction of pars tensa. The rate of pars flaccida retraction was similar at 73 ears (32%) in which the pars flaccida was adequately photographed. A total of 137 (56%) had no sign of retraction. The extent of retraction according to each staging system is shown in Table I. Reliability of Retraction Assessment Table II outlines the inter- and intraobserver agreement of the different systems for assessing the extent of TM retraction. It can be seen that the pars tensa systems were more reliable than the system used for pars flaccida retraction. Kappa scores were similar for the Erasmus and Sade systems for assessing extent of pars tensa retraction, although results for the latter are categorized more favorably as showing moderate agreement. The multicomponent characterization of retraction outlined in the Materials and Methods section was applied to a subset of the TM images, therefore the findings are not directly comparable, but it does give substantial agreement between observers. TM Retraction and Hearing Loss After excluding six ears with middle ear effusions, nine of 72 (13%) pars tensa retractions were found to be associated with conductive hearing loss (ABG >25 dB HL), compared with seven of 71 (10%) pars flaccida James et al.: Evaluation of Ear Drum Retraction TABLE I. Severity of Retractions by Staging System. Pars Tensa Multicomponent* Pars Tensa Sade8 Pars Tensa Erasmus6 Pars Flaccida Tos4 0 169 169 169 158 1 32 34 32 38 1b 2 13 9 — 9 — 13 — 21 Stage 2b 2 — — — 3 3b 3 2 23 — 18 — 10 — 4 7 6 8 4 5 6 4 3 — — 5 — — — 1 4 5 — 0 245 0 245 0 245 14 245 Cholesteatoma Not visible Total Ears with middle ear effusions are included. *See the Materials and Methods section for a description of the multicomponent characterization of retraction. retractions (P ¼ 0.05; v2 test). It can be seen in Figure 1 that the severity of hearing loss generally increases with the extent of retraction with all three staging systems. However, there was no significant difference between adjacent stages in any system except between Erasmus stages 1 and 2 (i.e., retraction not in contact with middle ear structures vs. retraction in contact with promontory) (P < .001; Spearman rank order correlation). Analysis of the different characteristics of pars tensa retraction in relation to hearing loss showed that retraction onto the promontory alone (11 ears) was associated with a hearing loss of 30 dB (mean ABG), retraction onto the incus and promontory (seven ears) a loss of 18 dB, and onto the incus alone (eight ears) a loss of 17dB. The hearing loss with retraction on to the promontory alone was significantly worse than when the TM was in contact with the incus (P ¼ 0.02; one-way analysis of variance). The number of ears with other characteristics of retractions was too small to make further subgroup analysis appropriate. DISCUSSION Classification systems for assessing the extent of TM retraction do not appear to be very reliable when applied to endoscopic images. Intra- and interobserver assessment is only slight to fair for pars flaccida retraction, and fair to moderate for retraction of the pars tensa. Even though such systems are widely used, they have not been extensively validated for use with conventional otoscopy, microscopy, or otoendoscopy. One previous report has documented unsatisfactory interobserver agreement for the Sade2 and Tos4 classifications using selected images of TM retraction and a large number of observers with varying experience levels.7 In the current study, levels of agreement remained suboptimal despite the specialist interest of the observers. A contributing factor is likely to be the limitation of assessment of the depth of retraction with a static two-dimensional endoscopic image compared to the dynamic, or especially microscopic, view available during clinical practice.13 Limitations in the design of the classifications that may contribute are discussed below. In addition to this lack of reliability, the clinical relevance of these systems is limited by the lack of any direct correlation with priorities in management, namely to 1) prevent suppurative complications from cholesteatoma; 2) treat erosive complications, most commonly permanent conductive hearing loss; 3) prevent erosive complications; 4) treat hearing loss; and 5) prevent hearing loss. Hearing threshold is more important to the patient than the appearance of their TM and must be included as part of the assessment of any retraction. The only feature of retraction noted to be associated with significant hearing loss was retraction onto the promontory alone, not touching the incus. This is represented by Erasmus stage 2. A further limitation of the Sade and Erasmus systems is that three separate states of retraction (onto incus alone, promontory alone, or both incus and promontory) are grouped into two stages; (i.e., Sade: [2] incus alone, [3] promontory with or without incus; Erasmus: [2] promontory alone, [3] incus with or without promontory). Although not relevant to hearing, justification for the lower ranking of isolated promontory retraction in the Erasmus system can be given by greater risk of permanent hearing loss from ossicular erosion or cholesteatoma development from posterosuperior retraction than from central TM retraction. Finally, neither system allows monitoring of small changes in extent of depth, size, adherence, or bone erosion, which further limits their usefulness. Small differences can be seen and potentially monitored more readily with serial endoscopic images. TABLE II. Inter- and Intraobserver Agreement for Repeated Assessments of Tympanic Membrane Retraction Using Different Staging Systems. Interobserver Agreement Intraobserver Agreement Kappa Agreement P Value Kappa Agreement P Value Pars flaccida, Tos 0.18 Slight <.001 0.30 Fair <.001 Pars tensa, Erasmus6 0.35 Fair <.001 0.37 Fair <.001 Pars tensa, Sade8 Pars tensa, multicomponent* 0.41 0.78 Moderate Substantial <.001 <.001 0.35 0.5 Moderate Fair <.001 <.001 Staging System 4 Level of agreement as defined by Landis and Koch.11 *The multicomponent characterization is described in the Materials and Methods section. Laryngoscope 122: May 2012 James et al.: Evaluation of Ear Drum Retraction 1117 Fig. 1. Box plots of hearing loss versus stage of retraction for the Sade and Erasmus classifications of pars tensa retraction and Tos classification of pars flaccida retraction. No significant difference in hearing between stages, except Erasmus stage 1 and 2 (Spearman rank order correlation). ABG ¼ air-bone gap. The importance of adherence of a retracted tympanic membrane to middle ear structures is recognized in the classification proposed by Sade.2 Observation of a single image per TM did not allow assessment of adherence in this study. Even in the clinical setting, this characteristic can be difficult to assess reliably, especially in children as cooperation can be limited. The clinical significance, and indeed validity of assessment of adherence, might depend on the method of assessment (Siegel speculum, Valsalva maneuver, insufflation by nitrous oxide anesthesia [the necessary duration of which will vary according to pneumatization14] or ease of surgical dissection at tympanoplasty). The significance of this finding may also depend on whether the retracted segment is partially or totally adherent to the promontory and/or incus. Finally, as demonstrated in Figure 2, the prognostic significance of adherence remains unclear, as retractions that had formerly appeared adherent may subsequently lateralize spontaneously. This study concentrated on assessment of pars tensa retraction, because that causes the majority of acquired pediatric cholesteatoma in our practice15 and others.16 In addition to assessing the validity of currently used staging systems, we studied other characteristics of TM retraction that might indicate deterioration, based on the clinical imperative to intervene before ossicular erosion occurs to optimize the chance of maintaining good hearing.1,15 The extent to which the retraction involved the long process of incus (touching/enveloping/eroding/complete erosion) was hypothesized to be important for monitoring the progression to permanent hearing loss. Factors that might be considered more immediate precursors of cholesteatoma (granulation tissue/keratin accumulation), which could be termed precholesteatoma, were also noted. Although substantial agreement was achieved on interpretation of these characteristics, the study was not adequately powered for subgroup analysis of their effect on hearing. Fig. 2. Otoendoscopic images of the right tympanic membrane of child with tympanic membrane retraction. (a) Pars tensa retraction appearing adherent to promontory, long process of incus, capitulum of the stapes and stapedius tendon. The long process of incus is partially eroded. (b) Same ear 10 months later showing lateralization of the tympanic membrane. The formerly retracted segment is now only in contact with the long process of incus. This improvement occurred spontaneously, without treatment. Laryngoscope 122: May 2012 1118 James et al.: Evaluation of Ear Drum Retraction TABLE III. Proposed Factors for the Assessment of Tympanic Membrane Retraction in Relation to the Planning of Reconstructive Surgery. Retraction with no keratin accumulation; normal hearing No boney contact in ME Retraction touching promontory Retraction touching incus Retraction touching stapes Whole surface of retraction not visible Retraction with impairment of keratin migration;* normal hearing Keratin trail on adjacent canal wall or drum Dry sheet or crust in retraction pocket Moist keratin or granulation tissue in pocket Retraction causing or contributing to ABG; no keratin accumulation Intact ossicular chain† Long process incus eroded (partial or complete) Stapes crus erosion (with or without incus erosion) Retraction with associated ABG and impaired keratin migration Cholesteatoma (i.e., uncleanable accumulation of desquamated keratin) *Keratin deposit can be removed completely without surgical intervention. † May not be ascertainable without surgical exploration. ME ¼ middle ear; ABG ¼ air-bone gap. We conclude that monitoring of TM retraction for guidance in timing of surgical management cannot be made with simple staging systems applied to endoscopic images of TM retraction and question whether these systems are significantly more valid when used with microscopy. Similarly, drawing of TM retractions has been shown to be unreliable.17 For individual patients, comparison of serial otoscopic images is likely to be a more reliable monitoring system to indicate progression or resolution of retraction (D Pothier, personal communication, thesis in preparation). For the purposes of guiding timing of TM reconstruction, a simple approach is proposed based on the detection of keratin accumulation and hearing loss. These are the important factors that influence surgical decision making.18 It is hypothesized that surgical reconstruction for TM retraction is more strongly indicated when keratin migration becomes impaired, increases more so if hearing loss occurs, and more still if both occur, hence the four retraction groups shown in Table III. A distinction must be made between accumulation of desquamated keratin, which can be cleaned from a retraction completely (in clinic or even under anesthesia), and cholesteatoma, in which the accumulated keratin cannot be removed without surgery. It has been proposed that cholesteatoma be added as the ultimate stage of the retraction process6 and is listed here for that reason, although the addition is somewhat superfluous when considering retraction management, as cholesteatoma is a separate entity. Classification systems with multiple stages become too cumbersome for routine use, but on the basis of experience from this study, additional potential subgroups are listed in Table III as suggestions for future research. Although they may be indicators of clinical severity, our Laryngoscope 122: May 2012 findings suggest it may not be possible to assess them reliably. Ossicular erosion is included in accordance with previous recommendations.12 Even though the current study shows retraction onto the promontory is associated with worse hearing than onto the incus, ossicular disruption is considered more likely to cause persistent hearing loss after spontaneous or surgical lateralization of the TM. Large and longitudinal studies would be required to validate the reliability and prognostic significance of all these subgroups. The difficulty clinicians have in accurate recollection of existing four-stage systems10 indicates that a system with more components could not be implemented reliably in standard clinical practice. However, in this research setting where details of the systems were available for reference, we found multicomponent assessment very feasible with good inter- and intraobserver reliability. Clearly, the decision to offer reconstructive TM surgery cannot be based on such a classification alone, but must also take into account the status of the contralateral ear and other patient factors. Reliability of assessment of pars flaccida retraction was much more susceptible to intra- and interobserver variability than assessment of the extent of pars tensa retraction. The subtle difference between a normal and slightly retracted pars flaccida and the subjective assessment of extent of scutum erosion are considered to be the principal limitations. We do not consider that clinical assessment with an otoscope or microscope would make these judgments any more reliable than with an endoscopic image. The absence of any correlation between stage of retraction and hearing loss is less surprising than with pars tensa retraction given its smaller role in hearing function. Our cross-sectional study design does not make it possible to determine whether any prognostic significance can be attached to this scale. Further characterization of pars flaccida retraction assessment would appear to be warranted. Children with cleft palate were selected for this study because of the greater prevalence of tympanic membrane retraction in this population. Nearly half of the ears were found to have some degree of retraction, compared with retraction rates in a normal pediatric population of around 6% pars tensa and 8% pars flaccida.19 The consecutive cross-sectional sample from a cleft palate clinic should provide a less-skewed representation of the severity of TM retraction in the normal population than sampling of an otologic clinic, in which severe retractions would likely be overrepresented. As only a small proportion of our sample had conductive hearing loss, investigation of a population with known retraction would facilitate investigation of the association between retraction and hearing loss. The staging systems evaluated in this study were not developed specifically for children with cleft palate. It is conceivable that the nature of TM retraction may differ in other populations so as to render our observations less applicable. CONCLUSION The reliability of staging systems for tympanic membrane retraction in current use is unsatisfactory James et al.: Evaluation of Ear Drum Retraction 1119 when applied to endoscopic images. The different stages do not allow monitoring of small changes in extent of retraction, or provide adequate distinction between factors associated with conductive hearing loss, ossicular erosion, or early formation of cholesteatoma. More reliable and clinically relevant measures of TM retraction are required. Acknowledgments The authors are grateful to Homira Osman and Andrea K. Suthakaran for entering audiometric data. BIBLIOGRAPHY 1. Borgstein J, Gerritsma TV, Bruce IA. Erosion of the incus in pediatric posterior tympanic membrane retraction pockets without cholesteatoma. Int J Pediatr Otorhinolaryngol 2008;72:1419–1423. 2. Sade J, Avraham S, Brown M. Atelectasis, retraction pockets and cholesteatoma. Acta Otolaryngol 1981;92:501–512. 3. Pars tensa and pars flaccida retractions in persistent otitis media with effusion. Otol Neurotol 2001;22:291–298. 4. Tos M, Poulsen G. Attic retractions following secretory otitis. Acta Otolaryngol 1980;89:479–486. 5. Charachon R, Gratacap B, Vuarnet J. Classification and surgical treatment of fibroadhesive otitis. Am J Otol 1985;6:305–310. 6. Borgstein J, Gerritsma TV, Wieringa MH, Bruce IA. The Erasmus atelectasis classification: proposal of a new classification for atelectasis of the middle ear in children. Laryngoscope 2007;117:1255–1259. 7. Pothier DD. The Sade and Tos staging systems: not adequately reliable methods of staging retraction of the tympanic membrane? Clin Otolaryngol 2009;34:506–507. Laryngoscope 122: May 2012 1120 8. Sade J, Ar A. Middle ear and auditory tube: middle ear clearance, gas exchange, and pressure regulation. Otolaryngol Head Neck Surg 1997; 116:499–524. 9. Sade J, Amos AR. The Eustachian tube. In: Ludman H, Wright T, eds. Diseases of the Ear. London: Arnold; 1998:334–352. 10. Pothier DD, Cooke C, Ahmed SB. Limited knowledge and recall of common staging systems: can we rely on what the clinical record tells us about retraction of the tympanic membrane? Clin Otolaryngol 2009;34:274. 11. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–174. 12. Committee on Hearing and Equilibrium guidelines for the evaluation of results of treatment of conductive hearing loss. American Academy of Otolaryngology–Head and Neck Surgery Foundation, Inc. Otolaryngol Head Neck Surg 1995;113:186–187. 13. Bhutta MF, Haggard M. Re: The Sade and Tos staging systems: not adequately reliable methods of staging retraction of the tympanic membrane? Clin Otolaryngol 2010;35:68; author reply 69. 14. Alper CM, Kitsko DJ, Swarts JD, et al. Role of the mastoid in middle ear pressure regulation. Laryngoscope 2011;121:404–408. 15. James AL. Hearing preservation in pediatric cholesteatoma surgery. Paper presented at: 37th Annual Meeting of the Society for Ear, Nose and Throat Advances in Children (SENTAC), December 3–6, 2009, in Salt Lake City, UT. Audio-Digest Otolaryngology: Advances in the Treatment of Children. Available at: http://www.audio-digest.org/pages/htmlos/ 35152.2.6868546997516489959. Accessed January 20, 2012. 16. Hamilton J, Rhagava N. Comparison of the characteristics of cholesteatoma at the time of surgery in children and adults. In: Ozgirgin ON, ed. Surgery of the Ear–Current Topics. Antalya, Turkey: Rekmay; 2009: 187–188. 17. Pothier DD, Awad Z. Can we accurately interpret drawings of ears with retractions? Clin Otolaryngol 2007;32:42–46. 18. Luntz M, Avraham S, Sade J. The surgical treatment of atelectatic ears and retraction pockets in children and adults. Eur Arch Otorhinolaryngol 1991;248:400–401. 19. Maw AR, Hall AJ, Pothier DD, Gregory SP, Steer CD. The prevalence of tympanic membrane and related middle ear pathology in children: a large longitudinal cohort study followed from birth to age ten. Otol Neurotol 2011;32:1256–1261. James et al.: Evaluation of Ear Drum Retraction