Otolaryngology–Head and Neck Surgery (2008) 139, 386-390

ORIGINAL RESEARCH—OTOLOGY AND NEUROTOLOGY

Evaluation of prognostic factors and middle ear risk

index in tympanoplasty
Ercan Pinar, MD, Kerim Sadullahoglu, MD, Caglar Calli, MD, and
Semih Oncel, MD, Izmir, Turkey
Black6 introduced the surgical, prosthetic, infection, tissues,
OBJECTIVE: The aim of this study was to examine the role of and eustachian tube system (SPITE), and more recently
the prognostic factors and middle ear risk index on the success of Kartush7 developed middle ear risk index (MERI). Bec-
tympanoplasty. varovski and Kartush revised and updated MERI in 2001
STUDY DESIGN: Case series.
(Table 1). Smoking is added as a middle ear risk. Further-
SUBJECTS AND METHODS: The charts of 231 patients who
underwent tympanoplasty operations between 2002 and 2007 were
more, cholesteatoma and granulation tissue or effusion risk
reviewed. Prognostic factors such as age, sex, presence of systemic value has been increased in MERI 2001.8 The MERI com-
diseases, location and size of perforation, duration of dry period, bines the known preoperative and intraoperative risk factors
presence of myringosclerosis, presence of septal and conchal pa- for tympanoplasty prognosis into a numeric value.
thology, operation type, and status of the opposite ear and middle The aim of this retrospective study was to evaluate the
ear risk index were investigated. role of prognostic factors and the MERI on the surgical
RESULTS: The overall success rate was 74.4%. Multivariate success of tympanoplasty.
analysis was carried out on significant prognostic factors to obtain
independent variables and yielded the following results (95% CI):
size of the perforation (⬍50%) (OR:8.11), healthy opposite ear
(OR:5.64), more than 3 months dry period (OR:2.21), absence of
myringosclerosis (OR:4.01) and low middle ear risk index SUBJECTS AND METHODS
(OR:87.1).
CONCLUSION: Size of the perforation(⬍50%), healthy oppo- A total of 231 patients who underwent tympanoplasty op-
site ear, absence of myringosclerosis, more than 3 months dry erations between January 2002 and September 2007 were
period, and low middle ear risk index were found to be significant reviewed in this study. This study was approved by an
independent prognostic factors. institutional review board. The patients ranged in age from
© 2008 American Academy of Otolaryngology–Head and Neck 11 to 58 years of age. Patients younger than 16 years old
Surgery Foundation. All rights reserved. were analyzed in the pediatric age group. The other patients’
ages ranged from 17 to 58 years; they were analyzed in the

T he goals of successful tympanoplasty are the removal

of the pathosis and achievement of a mucosal-lined
middle ear cleft with an intact tympanic membrane. The
adult group. The mean follow-up of patients was 11.8
months (range, 6 months to 3 years).
Each patient underwent audiologic evaluation from 250
goal of successful tympanoplasty is to create a sound con- Hz to 8 kHz preoperatively. Hearing results were based on
ducting mechanism in a well-aerated mucosal-lined middle the most recent audiogram available at 6 months postoper-
ear cleft. It also provides the possibility of improved hear- atively. The patients were followed up at three weeks, six
ing.1 There are various prognostic factors reported in the weeks, three months, and six months postoperatively. An
literature that may influence the surgical success of tympa- intact graft in proper position and aerated middle ear at the
noplasty.2 Some studies report significance for some of end of six months was considered a success.
these factors whereas others report the contrary. The re- Variables such as age, sex, presence of systemic dis-
ported incidence of surgical success of tympanoplasty eases (diabetes mellitus and hypertension), location and
ranges from 60% to 99% in adults.3 size of the perforation, status of the opposite ear, duration
Prognosis of tympanoplasty were classified. Belluci4 de- of the dry period, presence of myringosclerosis at admis-
scribed four separate stages for prognosis of tympanoplasty. sion, presence of nasal moderate septal and inferior con-
Austin5 proposed a prognostic stratification according to chal pathology (nasal septum deviation and hypertrophy
disease categories, stage categories, and disease descriptors. of inferior turbinates), type of operation, risk of the

Received March 16, 2008; revised May 19, 2008; accepted May 28,
2008.

0194-5998/$34.00 © 2008 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved.
doi:10.1016/j.otohns.2008.05.623
Pinar et al Evaluation of prognostic factors and middle ear risk . . . 387

Table 1
obtain independent prognostic factors and to assess the
Middle Ear Risk Index 2001 relative importance.

Risk Assigned
Risk factor value risk
RESULTS
Otorrhea (Bellucci)
I - Dry 0 There were 89 male and 142 female patients in this study.
II - Occasionally wet 1 The age range of the patients was 11 to 58 years (mean,
III - Persistently wet 2
IV - Wet, cleft palate 3 30.65 ⫾ 10.97). The pediatric age group included 42 pa-
__ tients (range, 11 to 16 years of age). The adult group had
Perforation 189 patients. Intact canal wall group (ICW) (176 patients,
None 0 76.1%) included myringoplasty, tympanoplasty with atticot-
Present 1 omy, and intact canal wall technique mastoidectomy. My-
__
Cholesteatoma ringoplasty was performed in 41 patients, tympanoplasty
None 0 with atticotomy in 14 patients, and intact canal wall tech-
Present 1 nique mastoidectomy in 121 patients. Canal wall down
__ (CWD) mastoidectomy was performed in 55 (23.8%) pa-
Ossicular status (Austin/Kartush) tients. Graft success rates for tympanoplasty without mas-
0) M⫹I⫹S⫹ 0
A) M⫹S⫹ 1 toidectomy (myringoplasty, tympanoplasty with atticot-
B) M⫹S⫺ 2 omy) was higher compared with intact canal wall technique
C) M⫺S⫹ 3 mastoidectomy (P ⫽ 0.011).
D) M⫺S⫺ 4 The overall graft success rate was 74.4% (172 of 231
E) Ossicular head fixation 2 patients), and the mean air-bone gap improvement was 8.2 ⫾
F) Stapes fixation 3

Middle ear: granulations or

effusion Table 2
No 0 Significant variables and success of tympanoplasty*
Yes 2
__ Patients Success rate
Previous surgery
Variables (n ⫽ 231) No: Incidence P value
None 0
Staged 1
Revision 2 Location of
__ perforation 0.022*
Smoker Anterior 84 53 (63.0%)
No 0 Posterior 39 28 (71.7%)
Yes 2 Central 108 91 (84.2%)
Total__ MER Size of perforation 0.001*
Index ⬍50% 129 106 (82.1%)
⬎50% 102 66 (64.7%)
Status of the
opposite ear 0.000*
MERI were evaluated. We used the MERI 2001 in our 8 Healthy 139 115 (82.7%)
Abnormal 92 57 (61.9%)
study. Patients with the severe septal deviation and infe- Duration of dry
rior turbinates hypertrophy, otorrhea, previous ear sur- period 0.003*
gery on the same ear, and deceased to follow-up were More than 3
excluded from the study. months 153 122 (79.7%)
The underlay tympanoplasty technique, harvesting the Less than 3
temporalis fascia as the graft material, was used in all months 78 50 (64.1%)
Myringosclerosis 0.027*
surgical procedures. General anesthesia was used in all Presence 71 45 (63.3%)
patients. The surgical approach was retroauricular in 179 Absence 160 127 (79.3%)
patients and endaural or transcanal in 52 patients. Mastoid- Operation type 0.010*
ectomies and type of the operation (intact canal wall tech- Intact canal wall
nique and canal wall down) were, whether performed or not, group 176 139 (78.9%)
based on the pre- and intraoperative findings. Canal wall
down group 55 33 (60%)
Univariate statistical analysis was performed with the Risk of MERI 0.000*
␹-squared test for the categoric variables. A P value of less Low 102 88 (86.2%)
than 0.05 was considered significant. Multivariate logistic Medium-High 129 84 (62.3%)
regression analysis with enter method was performed to
388 Otolaryngology–Head and Neck Surgery, Vol 139, No 3, September 2008

Table 3
Multivariate analysis of the significant prognostic parameters

Variables ␤ SE Sig. Odds ratio 95% CI

Location of perforation (Central) 0.038 0.585 0.949 1.038 0.33-3.27

Size of perforation (⬍50%) 2.094 0.834 0.012* 8.115 1.58-41.60
Healthy opposite ear 1.730 0.536 0.001* 5.641 1.972-16.13
More than 3 months dry period ⫺1.519 0.529 0.004* 2.219 0.078-0.61
Absence of myringosclerosis 1.390 0.546 0.011* 4.015 1.378-11.70
Operation type (ICW group) 0.274 0.698 0.694 1.315 0.335-5.16
Low MERI score 4.467 0.861 0.000* 87.101 16.110-470.92
*Significant variables, ␤, beta-coefficient; SE, standard error; CI, confidence interval; ICW group, intact canal wall group.

13.2. Based on the univariate analysis, central perforation and tympanoplasty success. Furthermore, the mean MERI
(P ⫽ 0.022), size of the perforation (⬍50%) (P ⫽ 0.001), score was found as 7.8 ⫾ 11.8 in the canal wall down
healthy opposite ear (P ⫽ 0.000), more than 3 months dry mastoidectomy group and 4.6 ⫾ 9.2 in the canal wall-up
period of the ear (P ⫽ 0.003), absence of myringosclerosis group. The mean MERI score was significantly higher in
(P ⫽ 0.027), operation type (intact canal wall group) (P ⫽ CWD mastoidectomy group compared with ICW group
0.01) and the low MERI score (ⱕ3) (P ⫽ 0.000) were found (Table 4).
to be statistically significant prognostic factors that affect The correlation between the significant variables and the
success rate (Table 2). MERI was analyzed with Spearman’s correlation coeffi-
The other variables did not show any statistically signif- cient. The operation type and the healthy opposite ear were
icant relationship with surgical success. Multivariate anal- correlated with the MERI, but the other significant variables
ysis was carried out on these significant prognostic factors were not (Table 5). The location of the perforation was also
to assess the relative importance of the independent vari- correlated with the size of the perforation (r ⫽ 0.579; P ⫽
ables and yielded the following results: the size of the 0.00). There was no significant correlation between the
perforation (⬍50%) (OR:8.11), healthy opposite ear (OR: other significant prognostic factors.
5.64), more than 3 months dry period (OR:2.21), absence
of myringosclerosis (OR:4.01) and the low MERI score
(OR:87.1) were found as independent prognostic factors
(Table 3). The location of the perforation (P ⫽ 0.949) and DISCUSSION
the operation type (P ⫽ 0.694) were not found as indepen-
dent prognostic factors. There have been many reports that discuss the prognostic
We used the MERI 2001 to stratify patients according to factors and hearing results in tympanomastoid surgery. Age
known preoperative and intraoperative risk factors for tym- is one of the prognostic factors. In general, the success of
panoplasty. The MERI generates a numeric value that cor- pediatric tympanoplasty is slightly lower than in adults.9,10
relates to severity of disease and prognosis. A MERI of 0 to This has been traditionally attributed to the fact that children
3 represents mild disease, 4 to 6 moderate disease, and 7 to have a higher incidence of eustachian tube dysfunction.11
12 severe disease. Patients were stratified into above risk On the other hand, some authors12-14 concluded that patient
categories. 102 (44.1%) patients had low risk, 80 (34.6%) age did not influence the surgical success of tympanoplasty.
nedium risk, and 49 (21.2%) high risk. There was a statis-
tically significant association between the low MERI score
Table 5
Correlation analysis of significant prognostic factors
with the MERI
Table 4
Association of the MERI and type of the operation Correlation
Variables coefficient P value
Mean MERI
Variables score P value Location of perforation 0.070 0.356
Size of perforation 0.115 0.129
Operation type 0.03* Healthy opposite ear 0.197 0.009*
ICW mastoidectomy group 4.6 ⫾ 9.2 More than 3 months dry period ⫺0.100 0.187
CWD mastoidectomy group 7.8 ⫾ 11.8 Absence of myringosclerosis 0.052 0.494
Operation type 0.209 0.005*
*Significant value, ␹2 statistical method; ICW, Intact canal
wall group; CWD, canal wall down group. *Significant correlation.
Pinar et al Evaluation of prognostic factors and middle ear risk . . . 389

We also did not find a relationship between success rate of Furthermore, mean MERI score was significantly higher in
surgery and patient age. CWD tympanoplasties compared with ICW tympanoplas-
The location and size of the perforation have been fre- ties. Possible explanations of this significant association
quently examined in the literature. Technically anterior per- could be: 1) moderate and severe MERI indicates presence
forations are more difficult to access and place grafts. Sur- of cholesteatoma, ossicular discontinuity, and middle ear
prisingly, previous studies15,16 reported that the location of pathology; 2) smoking is one of the parameters of the
the perforation had no effect on the surgical or hearing MERI; high MERI scores may indicate smoking, which has
result. Conversely, we found that graft success rate for negative local, regional, and systemic effects on the middle
central perforations was higher compared with posterior and ear mucosa; 3) these patients also need more extensive
anterior perforations. With respect to size, Lee et al17 and surgery; CWD tympanoplasty is often required to eradicate
Onal et al18 demonstrated a significantly higher success rate the disease in patients with severe MERI.
with perforations smaller than 50%. We also found a sig-
nificant relationship between graft success and smaller per-
foration. However, no statistically significant association CONCLUSION
was found between perforation size and successful tympa-
noplasty in recent studies.15,19 The size of the perforation (⬍50%), healthy opposite ear,
Many otolaryngologists believe that dry ear is critical for more than 3 months of dry period, absence of myringoscle-
graft uptake, whereas some believe it plays little role in rosis, and low MERI scores were found as independent
success. Uyar et al20 found that there was a significantly prognostic factors and carried respectively a eight-, five-,
higher rate of graft uptake in patients who had dry ear for 3 two-, four-, and eighty-seven–fold increase on the surgical
months preoperatively. We also found a significant associ- success rate, respectively. Although the location of the per-
ation between dry ear and success of tympanoplasty similar foration and the operation type were significant variables in
to this study. We recommend performing tympanoplasty the univariate analysis, they were not found as independent
when the ears have been dry for 3 months. On the other prognostic factors in logistic regression analysis. This study
hand, there are also studies that found no statistically sig- showed that high MERI had a much higher possibility of
nificant correlation with respect to dry ear.14,18 CWD surgery and lower chance of successful tympano-
The status of the opposite ear is an important prognostic plasty. The consideration of the high MERI allows the
factor. Our study also revealed that the status of the opposite surgeon to counsel the patient more accurately before sur-
ear was a prognostic factor. Success of graft uptake was gery and provide realistic expectations. This will enable the
poorer in patients whose opposite ears were athelectatic or surgeon to design a case-specific operation strategy for each
perforated. Collins et al21 had similar findings. Merenda et patient.
al15 also reported that contralateral disease was associated
with lower success rates. Conversely, Singh et al16 found no
association between contralateral ear status and surgical AUTHOR INFORMATION
success.
From the Ataturk Training and Research Hospital, ENT Clinic, Izmir,
Effect of the presence of myringosclerosis on surgical Turkey.
success was not investigated thoroughly during assessment
Corresponding author: Ercan Pinar, MD, Sair Esref Bul. No: 80/13, 35220
of the prognostic factors. This factor was investigated pre- Alsancak-Izmir, Turkey.
viously only in Onal’s study.18 The author reviewed 74 E-mail address: [email protected].
tympanoplasties and did not find a correlation between
success and myringosclerosis. In our large series, we found
that absence of myringosclerosis increased success of tym-
panoplasty. AUTHOR CONTRIBUTION
We found that ICW tympanoplasties actually had higher Ercan Pinar, study design; Kerim Sadullahoglu, data collection; Caglar
surgical success rates of graft compared with CWD tympa- Calli, data collection; Semih Oncel, design and approval of manuscript.
noplasties. Onal et al18 did not agree with this finding in
their series. Success rates were higher in canal wall down
tympanoplasties compared with canal wall up tympanoplas- FINANCIAL DISCLOSURE
ties and their result was close to the level of significance.
We used the MERI 2001 to stratify patients according to None.
known preoperative and intraoperative risk factors for tym-
panoplasty. The relationship between MERI and surgical
success has not been investigated in previous studies. Kar- REFERENCES
tush7 introduced MERI for prognosis of tympanomastoid
surgery initially in 1994. We found that patients with low 1. Albu S, Babighian G, Trabalzini F. Prognostic factors in tympano-
MERI scores (ⱕ3) had significantly higher rates of success. plasty. Am J Otol 1998;19:136 –140.
390 Otolaryngology–Head and Neck Surgery, Vol 139, No 3, September 2008

2. Lin AC, Messner AH. Pediatric tympanoplasty: factors affecting suc- 12. Yung M, Neumann C, Vowler SL. A longitudinal study on pediatric
cess. Curr Opin Otolaryngol Head Neck Surg 2008;16:64 – 68. myringoplasty. Otol Neurotol 2007;28:353–355.
3. Inwood JL, Wallace HC, Clarke SE. Endaural or postaural incision for 13. Umapathy N, Dekker PJ. Myringoplasty: is it worth performing in
myringoplasty: does it make a difference to the patient? Clin Otolar- children? Arch Otolaryngol Head Neck Surg 2003;129:1053–1055.
yngol Allied Sci 2003;28:396 –398. 14. Albera R, Ferrero V, Lacilla M, et al. Tympanic reperforation in
4. Belluci R. Dual classification of tympanoplasty. Laryngoscope 1973; myringoplasty: evaluation of prognostic factors. Ann Otol Rhinol
83:1754 –1758. Laryngol 2006;115:875– 879.
5. Austin D. Reporting results in tympanoplasty. Am J Otol 1985;6: 15. Meranda D, Koike K, Shafiei M, et al. Tympanometric volume: a
85– 88. predictor of success of tympanoplasty in children. Otolaryngol Head
6. Black B. Design and development of a contoured ossicular replace- Neck Surg 2007;136:189 –192.
ment prosthesis: clinical trials of 125 cases. Am J Otol 1990;11:
16. Singh GB, Sidhu TS, Sharma A, et al. Tympanoplasty type 1 in
85– 89.
children: an evaluative study. Int J Pediatr Otorhinolaryngol 2005;69:
7. Kartush JM. Ossicular chain reconstruction: capitulum to malleus.
1071–1076.
Otolaryngol Clin North Am 1994;27:689 –715.
17. Lee P, Kelly G, Mills RP. Myringoplasty: does the size of the perfo-
8. Becvarovski Z, Kartush JM. Smoking and tympanoplasty: implica-
ration matter? Clin Otolaryngol Allied Sci 2002;27:331–334.
tions for prognosis and the Middle Ear Risk Index (MERI). Laryngo-
scope 2001;111:1806 –1811. 18. Onal K, Uguz MZ, Kazikdas KC, et al. A multivariate analysis of
9. Koch WM, Friedman EM, McGill TJ, et al. Tympanoplasty in chil- otological, surgical and patient related factors in determining success
dren. The Boston Children’s Hospital experience. Arch Otolaryngol in myringoplasty. Clin Otolaryngol 2005;30:115–120.
Head Neck Surg 1990;116:35– 40. 19. Denoyelle F, Roger G, Chauvin P, et al. Myringoplasty in children:
10. Emir H, Ceylan K, Kizilkaya Z, et al. Success is a matter of experi- predictive factors of outcome. Laryngoscope 1999;109:47–51.
ence: type 1 tympanoplasty: influencing factors on type 1 tympano- 20. Uyar Y, Keles B, Koc S, et al. Tympanoplasty in pediatric patients. Int
plasty. Eur Arch Otorhinolaryngol 2007;264:595–599. J Pediatr Otorhinolaryngol 2006;70:1805–1809.
11. Bluestone CD, Cantekin EI, Kenna MA, et al. Prognostic value of 21. Collins WO, Telischi FF, Balkany TJ, et al. Pediatric tympanoplasty:
eustachian tube function in pediatric tympanoplasty. Laryngoscope effect of contralateral ear status on outcomes. Arch Otolaryngol Head
1987;97:1012–1016. Neck Surg 2003;129:646 – 651.