Acta Oto-Laryngologica
ISSN: 0001-6489 (Print) 1651-2251 (Online) Journal homepage: http://www.tandfonline.com/loi/ioto20
Comparison of cartilage graft and fascia in type
1 tympanoplasty: systematic review and meta-
analysis
Tao Yang, Xuewen Wu, Xiaofei Peng, Yanni Zhang, Shaobing Xie & Hong Sun
To cite this article: Tao Yang, Xuewen Wu, Xiaofei Peng, Yanni Zhang, Shaobing Xie & Hong Sun
(2016) Comparison of cartilage graft and fascia in type 1 tympanoplasty: systematic review and
meta-analysis, Acta Oto-Laryngologica, 136:11, 1085-1090, DOI: 10.1080/00016489.2016.1195013
To link to this article: https://doi.org/10.1080/00016489.2016.1195013
Published online: 16 Jun 2016.
Submit your article to this journal
Article views: 293
View related articles
View Crossmark data
Citing articles: 2 View citing articles
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ioto20
ACTA OTO-LARYNGOLOGICA, 2016
VOL. 136, NO. 11, 1085–1090
http://dx.doi.org/10.1080/00016489.2016.1195013
REVIEW ARTICLE
Comparison of cartilage graft and fascia in type 1 tympanoplasty: systematic
review and meta-analysis
Tao Yanga,b*, Xuewen Wua,b*, Xiaofei Pengc, Yanni Zhanga,b, Shaobing Xiea,b and Hong Suna,b
a
Department of Otolaryngology Head and Neck Surgery, Xiangya Hospital, Central South University, Changsha, Hunan, PR China; bProvince
Key Laboratory of Otolaryngology Critical Diseases, Xiangya Hospital, Central South University, Changsha, Hunan, PR China; cDepartment of
Rheumatology and Immunology, The Second Xiangya Hospital of Central South University, Changsha, Hunan, PR China
ABSTRACT ARTICLE HISTORY
Conclusions: Tympanoplasty using cartilage grafts has a better graft take rate than that using tempor- Received 23 February 2016
alis fascia grafts. There are no significant differences between cartilage grafts and temporalis fascia Revised 15 May 2016
grafts for hearing outcomes. Contrary to the sliced cartilage sub-group, full-thickness cartilage grafts Accepted 19 May 2016
generate better hearing outcomes than temporalis fascia grafts. Published online 14 June
2016
Objective: Tympanic membrane perforation can cause middle ear relapsing infection and lead to hear-
ing damage. Various techniques have been applied in order to reconstruct the tympanic membrane. KEYWORDS
Recently, cartilage grafts and temporalis fascia grafts have been widely used for tympanic membrane Cartilage; fascia; hearing
closure. A systemic review and meta-analysis was carried out based on published retrospective trials loss; tympanic membrane;
that investigated the efficacy of cartilage grafts and temporalis fascia grafts in type 1 tympanoplasty. tympanoplasty
Both graft take rates and mean AIR-BONE-GAP gains were analyzed.
Methods: Cochrane Library, PubMed, and Embase were systematically searched. After a scientific investi-
gation, we extracted the relevant data following our selection criteria. Odds ratio (OR) of graft take rates
and mean difference (MD) of AIR-BONE-GAP gains were calculated within 95% confidence intervals.
Results: Eight eligible articles with 915 patients were reviewed. The pooled OR for graft take rate was
3.11 (95% CI ¼1.94–5.00; p ¼ 0.43) and the difference between the two groups was significant, which
means that the cartilage grafts group got a better graft take rate than the temporalis fascia grafts
group. The pooled MD for mean AIR-BONE-GAP gain was 1.92 (95% CI ¼ 0.12–3.95; p < 0.000 01) and
the difference was not significant. However, in the full thickness cartilage grafts sub-group, the pooled
MD for mean AIR-BONE-GAP gains was 2.56 (95% CI ¼1.02–4.10; p ¼ 0.14) and the difference was sig-
nificant, which means that the full thickness cartilage grafts sub-group got a better hearing outcome
than the temporalis fascia grafts group. On the contrary, the pooled MD of sliced cartilage grafts sub-
group was 0.12 (95% CI ¼ 0.44–0.69; p ¼ 0.61) and there was no significant difference between the
sliced cartilage grafts and temporalis fascia group.
Introduction or sliced cartilage ( 0.5 mm) harvested from tragus and
concha [11–13]. Our department has used both temporalis
Chronic otitis media is a kind of inflammatory disease that fascia and cartilage as closure grafts for many years, and
can cause permanent pathologic change to tympanic mem- we would like to acquire more reliable clinical evidences for
brane and other middle ear structures. Tympanic membrane different surgical techniques in order to treat Type 1
perforation can cause middle ear relapsing infection and lead Tympanoplasty.
to hearing damage. Various grafts have been used for closure The purpose of this review was, therefore, to compare car-
of tympanic membrane [1–6], and the temporalis fascia (TF) tilage to TF grafts in primary type 1 tympanoplasty patients,
and cartilage have been the most common used tissues based on both anatomical and audiologic outcomes. In add-
currently. ition, we also would like to collect information about the use
TF has lots of advantages such as easy access, harvesting of two different surgical techniques for cartilage grafts (full
easily, abundant quantity, and relatively transparent and thin thickness and sliced cartilage grafts).
nature [7,8], but its low elastic modulus makes it much more
susceptible to different pressure variations [9]. Cartilage,
however, is pliable and can resist deformation because of its Materials and methods
high elastic modulus. Moreover, the cartilage can receive Types of studies
nutrition through diffusion [10], which may improve its
grafts acceptance rate. In recent years, more and more We included all the retrospective studies comparing cartilage
researchers have reported their works by using full thickness to temporalis fascia grafts in type I tympanoplasty.
CONTACT Hong Sun shjhaj@vip.163.com Department of Otolaryngology Head and Neck Surgery, Xiangya Hospital, Central South University, Changsha,
Hunan 410008, PR China
*These authors contributed equally to the paper.
ß 2016 Acta Oto-Laryngologica AB (Ltd)
1086 T. YANG ET AL.
Table 1. Search syntax. (b) Intervention: Type I tympanoplasty; (c) Comparison:
Database Search syntax Hits Fascia or cartilage grafts were used in the surgery; (d)
Cochrane #1 MeSH descriptor: [Tympanoplasty] explode 4 Outcome: Graft take rate and AIR-BONE-GAP, of which the
all trees 90
#2 Ear, Middle:ti,ab,kw in Trials 1721
mean follow-up time should be more than 1 year; and (e)
#3 Ears, Middle:ti,ab,kw 1815 Design: Retrospective studies.
#4 Middle Ears:ti,ab,kw 1815
#5 Middle Ear:ti,ab,kw 1815
#6 Tympanic Cavity:ti,ab,kw 27 Definition of outcome measures
#7 Cavities, Tympanic:ti,ab,kw 27
#8 Cavity, Tympanic:ti,ab,kw 27 Post-operative surgical success was defined as an intact tym-
#9 Tympanic Cavities:ti,ab,kw 27
#10 Tympanum:ti,ab,kw 16 panic membrane without any perforation, retraction, or lat-
#11 ‘reconstruction’:ti,ab,kw 3769 eralization for at least 12 months after the operation. The
#12 Surgical Procedures,
Reconstructive:ti,ab,kw 820
average air-bone-gap was calculated at 0.5, 1, 2, and 3 kHz
#13 Reconstructive Surgical according to the guidelines of the Korean Otologic Society
Procedure:ti,ab,kw 820 (2006) [14] and hearing gain was defined as the difference
#14 Surgical Procedure,
Reconstructive:ti,ab,kw 820
between the post- and pre-operative AIR-BONE-GAP.
#15 Procedure, Reconstructive
Surgical:ti,ab,kw 820
#16 Cosmetic Reconstructive Surgical Data collection and extraction
Procedures:ti,ab,kw 36
#17 Reconstructive Surgical Procedures, Two authors (Tao Yang and Xiaofei Peng) independently
Esthetic:ti,ab,kw 3
#18 Reconstructive Surgical Procedures,
extracted the following data: first author, year of publication,
Cosmetic:ti,ab,kw 36 number of patients, patient characters, study design, experi-
#19 Plastic Surgery:ti,ab,kw 737 mental group (cartilage grafts), control group (TF grafts),
#20 Cosmetic Surgery:ti,ab,kw 731
#21 Esthetic Surgery:ti,ab,kw 153
definition of graft take rate and mean AIR-BONE-GAP, and
#22 #2 or #3 or #4 or #5 or #6 or #7 or #8 the outcomes data. Extracted data were listed in a standar-
or #9 or #10 1839 dized Excel (Microsoft Corp) file and were also checked by
#23 #11 or #12 or #13 or #14 or #15 or #16
or #17 or #18 or #19 or #20 or #21 5504 another author (Yanni Zhang). Any disagreements were
#24 #22 and #23 33 resolved by discussion and consensus.
#25 #1 or #24 120
#26 Fascia:ti,ab,kw 588
#27 Cartilage:ti,ab,kw 1201 Risk of bias assessment
#28 #26 and #27 18
#29 #25 and #28 4
Pubmed ((((Tympanoplasties[Title/Abstract]) OR 71
We use the JBI systematic review protocol [15] to assess the
‘Tympanoplasty’[Mesh])) AND risk of bias. It includes selection bias, performance bias,
‘Fascia’[Mesh]) AND ((‘Cartilage’[Mesh]) OR detection bias, attrition bias, and reporting bias. No funnel
Cartilages[Title/Abstract])
Embase #4 #1 AND #2 AND #3 25 plot was employed for asymmetry, because fewer than 10
#3 ‘fascia’/exp studies were ultimately included.
#2 ‘cartilage’/exp
#1 ‘tympanoplasty’/exp
Date of search: December 18, 2015. Statistical analysis
Statistical analysis was generated using Review Manager 5
Search strategy
(Cochrane Collaboration, Oxford, UK) and SPSS for win-
The Cochrane Central Register of Controlled Trials, dows Version 17.0 (SPSS Inc., Chicago, IL). The fixed-effects
PubMed, and EMBASE were searched by two different model was employed when statistical heterogeneity did not
authors (Tao Yang and Xiaofei Peng) following the PICOS exist (p-value was larger than 0.1) or when the I2 value was
(patient, intervention, comparison, and outcome) principles, less than 0.4; otherwise, the random-effects model was used.
which were listed in Table 1. In addition, the references list Results are presented using the RRs (relative risks), MD
of identified articles was checked manually in order to (mean difference) and 95% CIs. P-values less than 0.05, RR
include other potentially eligible trials. This process was per- not including 1 and MD not including 0 were significant.
formed iteratively until no additional studies could be
identified. Results
Search results and study descriptions
Selection criteria
Our initial search identified 100 articles, of which 83
The following inclusive selection criteria was applied: (a) appeared unique (Figure 1). Abstract and full-text screening
Population: All the patients were primary cases with a sub- yielded eight articles which were eligible for solving our clin-
total perforation (> 50%) in which the ossicular chain was ical problems. All the articles’ information and data were
intact and no mastoid surgeries were performed., excluding listed in Table 2. Finally, there are 467 patients in the cartil-
patients with atelectasis, cholesteatoma or mastoid surgery; age group and 448 patients in the temporal fascia group, and
ACTA OTO-LARYNGOLOGICA 1087
100 records idenfied through database search
(25 from EMBASE, 71 from PUBMED, 4 from
COCHRANE)
18 records removed because of duplicates
83 tle/abstracts screened
53 excluded on basis of tle and abstract
8 reviews, leers or cases
11 not relevant exposure or outcome
7 not found arcles
27 earlier than 1990s
30 full-texts retrieved for further detailed evaluaon
21 excluded due to:
13 not retrospecve studies
8 insufficient data for quantave analysis
8 studies provided sufficient data for meta-analysis
Figure 1. Flow chart for article selection.
they are from different counties. The patients’ age ranged twenty-four patients are available in the thin cartilage grafts
from 8–75 years old. They were operated on by using cartil- sub-group and the pooled MD was 0.12 (95%
age or temporal fascia grafts. CI ¼ 0.44–0.69; p ¼ 0.61; Figure 3).
The cartilage was cut to less than 0.5 mm thick in three of Then we excluded the study of Vashishth [18], which
the eight studies [13,16,17] and the other five articles used may have moderate bias risks. The pooled MD of the full-
full-thickness cartilage to repair the perforation. Both the thickness cartilage grafts sub-group changed to 2.56 (95%
graft take rate and mean AIR-BONE-GAP gain were meas- CI ¼ 1.02–4.10; p ¼ 0.14; Figure 3).
ured for at least 1-year follow-up.
Discussion
Assessment of risks of bias in included studies
Summary of main results
Eight eligible studies were assessed for risk of bias following
the JBI systematic review protocol (Table 3). Seven of the This article is a systematic review and meta-analysis of eight
eight trials were found to have lower bias risks (score above retrospective studies. It evaluates the efficacy of cartilage graft
6) and one trial with moderate bias risk (score 4 6). Three compared to temporalis fascia graft for type I tympanoplasty.
different people assessed the eight articles independently. The present meta-analysis showed superiority of cartilage
grafts for tympanic membrane closure after at least 1 year
follow-up. For hearing outcome, the full-thickness cartilage
Graft take rate sub-group indicated a better average air-bone-gap gain com-
Eight studies with 915 patients are available for analysis of pared to the temporalis fascia. However, no relevant differ-
graft take rate. The pooled OR (odds ratio) for graft take ence was found between the sliced cartilage grafting sub-
rate was 3.11 (95% CI ¼ 1.94–5.00; p ¼ 0.43; Figure 2). In the group and temporalis fascia group.
full-thickness cartilage grafts sub-group, five studies with 380
people are available for analysis. The pooled OR for graft Comparison with other reviews
take rate was 3.73 (95% CI ¼ 1.90–7.32; p ¼ 0.53; Figure 2).
In total, 535 patients were included in the thin cartilage sub- The main findings of our meta-analysis seems to contradict
group. The pooled OR for graft take rate was 2.52 (95% the meta-analysis of Lyons et al. [19], which was performed
CI ¼ 1.28–4.95; p ¼ 0.21; Figure 2). to appraise the effectiveness of one-piece composite cartilage
grafting compared to temporalis fascia in type 1 tympano-
plasty. Our analysis included different types of cartilage graft-
Mean air-bone-gap gain ing such as one piece, palisade, and sliced techniques. We
Eight studies with 868 patients were available for analysis of also focused on the difference between full-thickness cartilage
mean AIR-BONE-GAP gains. The pooled MD was 1.92 (one-piece or palisade) and sliced cartilage ( 0.5 mm) com-
(95% CI ¼ 0.12–3.95; p < 0.000 01; Figure 3). In the full- pared with temporalis fascia graft. In addition, due to our
thickness cartilage grafts sub-group, five studies with 344 selection criteria, the mean follow-up duration for TM clos-
people are included and the pooled MD was 3.30 (95% ure and hearing outcomes should be 1 year. Therefore, we
CI ¼ 1.06–5.55; p ¼ 0.002; Figure 3). Three hundred and excluded most of the articles with 3-month follow-ups which,
1088 T. YANG ET AL.
Table 2. Characters and extracted data of the eight included articles.
Cartilage Temporalis fascia
Author year Types of cartilage Graft taken rate Mean ABG gain (dB) Graft taken rate Mean ABG gain (dB)
K. Cagdas [17] 2007 Sliced 22/23 8.3 ± 5.3 21/28 10.5 ± 10.0
Cem Ozbek [10] 2008 Thick 21/21 14.71 ± 2.47 17/24 14.2 ± 9.49
Kazim Onal [11] 2012 Thick 36/39 14.49 ± 3.59 27/41 11.63 ± 3.47
Joo Yeon Kim [16] 2012 Sliced 64/83 9.78 ± 15.25 22/31 9.71 ± 8.94
Esra Sozen [9] 2012 Thick 40/43 9.4 ± 3.8 29/36 5.7 ± 2.8
Sule Demirci [12] 2014 Thick 23/25 12.1 ± 6.8 21/25 13.1 ± 9.6
A. Vashishth [18] 2014 Thick 27/30 21.67 ± 6.73 50/60 13.24 ± 6.58
M. M Khan [13] 2015 Sliced 200/203 24.28 ± 2.79 157/167 24.12 ± 2.71
Table 3. JBI critical appraisal checklist for comparable cohort study.
Selection bias Confounding Objective Sufficient fol- Loss follow-up Reliable Appropriate
Author Representative control factors control criteria low-up time control measurement analysis Risk of bias
Ashish þ þ þ / þ / moderate
Vashishth [18]
Cem Ozbek [10] / / þ þ þ þ þ Iow
Esra Sozen [9] þ / / þ þ þ þ þ Iow
Joo Yeon þ þ / þ þ þ þ þ Iow
kim [16]
K. Cagdas þ / þ þ þ þ þ þ Iow
Kazikdas [17]
Kazim Onal [11] þ þ / þ þ þ þ þ Iow
M. M. Khan [13] þ þ þ þ þ þ þ þ Iow
Sule Demirci [12] þ / þ þ þ þ þ Iow
þ, qualification; , disqualification; /, not clear; one þ gets 1 score.
Figure 2. Forest plot of graft taken rate for the eight included articles.
in turn, were included in previous meta-analysis. Based on we followed the PRISMA (preferred reporting items for sys-
our experience, the hearing outcome should be followed for tematic reviews and meta-analyses) guideline. Three different
at least 1 year to obtain a stable result. Those may explain authors separately searched the database and we discussed
why we made a different conclusion. objectively each of the included articles. We attempted to
contact the corresponding author or other institutions when
a study was not available in full text. All the patients in the
Quality of evidence
eight studies were primary cases, in which the patients had
For the present meta-analysis, in an attempt to demonstrate intact ossicular chain and dry ears for at least 4 weeks, and
robust results, we employed rigorous inclusion criteria and no mastoid surgery was performed. The patients were from
ACTA OTO-LARYNGOLOGICA 1089
Figure 3. Forest plot of mean air-bone-gap for the included articles.
three different countries (Turkey, India and Korea) and con- achieved by both full-thickness and thin cartilage graft.
tained both children and adults, which can well reflect the However, when the cartilage was cut thinner than 0.5 mm,
whole population characteristics. We carefully scrutinized all the original structure was destroyed and its rigidity declined.
studies to make sure we include the patients’ unique data As a result, the sliced cartilage may be prone to shrinkage,
and two outcomes: graft take rate and average air-bone-gap leading to worse hearing outcomes. Further inquiry should
gain. Then we applied sub-group analysis and sensitivity ana- be made using a large number of sliced cartilage tympano-
lysis to assess the heterogeneity. When we excluded the high- plasty patients over a longer follow-up duration with the
risk article written by Vashishth et al. [18], the I2 changed evaluation of prognostic factors.
from 77% to 45% and the heterogeneity lowered. In addition,
the total I2 was 81%, which indicates the two surgical techni-
Potential biases in review
ques may lead to heterogeneity.
One potential limitation of the present meta-analysis is that
all the included articles were retrospective studies, which
Implications for clinical practice and research
may cause bias. Although we found one RCT study by
The present meta-analysis’s results were parallel to some pre- Mauri et al., we finally excluded the article because the
vious clinical studies [20]. The cartilage-perichondrium com- Cochrane handbook for systematic reviews did not recom-
posite graft can receive nutrition by diffusion and can resist mend to combine evidence from randomized trials and non-
deformation by pressure variations due to its stiffness. Those randomized studies, which may increase the risk of bias. The
reasons may explain the better temporalis membrane closure risk of bias was assessed with JBI criteria [15]. However, any
1090 T. YANG ET AL.
proposed tool would have difficulty to validate, and realistic Mechanics and Otosurgery, Dresden, Germany, September
assessments are eventually open to subjectivity. Furthermore, 19–22, 1996, Karl-Bernd H€ uttenbrink (Ed.). Bibliothek der
HNO-Universit€atsklinik, Dresden, Germany, 1997. ISBN 3-
there may be a publication bias towards the reporting of
86005-193-8. Int J Pediatr Otorhinolaryngol 1999;48:182–3.
spuriously inflated effects and positive trends in smaller and [8] Levinson RM. Cartilage-perichondrial composite graft tympano-
retrospective studies. plasty in the treatment of posterior marginal and attic retraction
pockets. Laryngoscope 1987;97:1069–74.
[9] Sozen E, Orhan Ucal Y, Tansuker HD, Uslu Coskun B, Yasemin
Conclusions Korkut A, Dadas B. Is the tragal cartilage necessary for type 1
tympanoplasties? J Craniofac Surg 2012;23:e280–3.
Tympanoplasty using cartilage grafts has a better graft take [10] Ozbek C, Ciftci O, Tuna EE, Yazkan O, Ozdem C. A compari-
rate than that using temporalis fascia grafts. There are no son of cartilage palisades and fascia in type 1 tympanoplasty in
significant differences between cartilage grafts and temporalis children: anatomic and functional results. Otol neurotol
fascia grafts for hearing outcomes. Contrary to the sliced car- 2008;29:679–83.
[11] Onal K, Arslanoglu S, Songu M, Demiray U, Demirpehlivan IA.
tilage sub-group, full-thickness cartilage grafts generate better Functional results of temporalis fascia versus cartilage tympano-
hearing outcomes than temporalis fascia grafts. plasty in patients with bilateral chronic otitis media. J Laryngol
Otol 2012;126:22–5.
[12] Demirci S, Tuzuner A, Karadas H, Acikgoz C, Caylan R, Samim
Acknowledgements EE. Comparison of temporal muscle fascia and cartilage grafts
in pediatric tympanoplasties. Am J Otolaryngol 2014;35:796–9.
This study was supported by the National Key Basic Research Program [13] Khan MM, Parab SR. Comparative study of sliced tragal cartil-
of China (973 Program), Grant NO. 2014CB943003 and National age and temporalis fascia in type I tympanoplasty. J Laryngol
Natural Science Foundation of China (NSFC), Grant no. 81300819 and Otol 2015;129:16–22.
Grant NO. 81170912. The funders had no role in study design, data col- [14] Kim HJ. Classification and hearing result reporting guideline in
lection and analysis, decision to publish, or preparation of the chronic otitis media surgery. Korean J Otolaryngol Head Neck
manuscript.
Surg 2006;49:2–6.
[15] Wing Sze Tang YLC, Koh SSL. The effectiveness of physical
restraints in reducing falls among adults in acute care hospitals
Disclosure statement and nursing homes: a systematic review. JBI Database System
The authors report no conflicts of interest. The authors alone are Rev Implement Rep 2010;8:3–26.
responsible for the content and writing of the paper. [16] Kim JY, Oh JH, Lee HH. Fascia versus cartilage graft in type I
tympanoplasty: audiological outcome. J Craniofac Surg
2012;23:e605–8.
[17] Kazikdas KC, Onal K, Boyraz I, Karabulut E. Palisade cartilage
References
tympanoplasty for management of subtotal perforations: a com-
[1] Sheehy JL, Anderson RG. Myringoplasty. A review of 472 cases. parison with the temporalis fascia technique. Eur Arch
Ann Otol Rhinol Laryngol 1980;89:331–4. Otorhinolaryngol 2007;264:985–9.
[2] Zini C, Sanna M, Bacciu S, Delogu P, Gamoletti R, Scandellari [18] Vashishth A, Mathur NN, Choudhary SR, Bhardwaj A. Clinical
R. Molded tympanic heterograft. An eight-year experience. Am J advantages of cartilage palisades over temporalis fascia in type I
Otol 1985;6:253–6. tympanoplasty. Auris Nasus Larynx 2014;41:422–7.
[3] Marquet J. Reconstructive micro-surgery of the eardrum by [19] Lyons SA, Su T, Vissers LET, Peters JPM, Smit AL, Grolman W.
means of a tympanic membrane homograft. Preliminary report. Fascia compared to one-piece composite cartilage-perichondrium
Acta Otolaryngol 1966;62:459–64. grafting for tympanoplasty. Laryngoscope 2015. Doi: 10.1002/
[4] Goodhill V. Tragal perichondrium and cartilage in tympano- lary.25772. [Epub ahead of print].
plasty. Arch Otolaryngol 1967;85:480–91. [20] Iacovou E, Vlastarakos PV, Papacharalampous G, Kyrodimos E,
[5] Gerber MJ, Mason JC, Lambert PR. Hearing results after pri- Nikolopoulos TP. Is cartilage better than temporalis muscle
mary cartilage tympanoplasty. Laryngoscope 2001;110:1994–9. fascia in type I tympanoplasty? Implications for current surgical
[6] Rizer FM. Overlay versus underlay tympanoplasty. Part I: histor- practice. Eur Arch Otorhinolaryngol 2013;270:2803–13.
ical review of the literature. Laryngoscope 1997;107:1–25. [21] Mauri M, Lubianca Neto JF, Fuchs SC. Evaluation of inlay
[7] Puria S. Middle Ear Mechanics in Research and Otosurgery: butterfly cartilage tympanoplasty: a randomized clinical trial.
Proceedings of the International Workshop on Middle Ear Laryngoscope 2001;111:1479–85.
