Factors Associated With Quality of Life After Ear Surgery in Patients With Chronic Otitis Media

Objective To assess the factors affecting health-related quality of life (HRQOL) after ear surgery in patients with chronic otitis media (COM).

Design Prospective questionnaire-based outcome study.

Setting Tertiary care institution.

Patients The study population comprised 205 patients aged 21 to 67 years referred for ear surgery owing to COM with or without cholesteatoma between November 2009 and March 2011.

Interventions The ear surgery for COM included tympanoplasty, with or without mastoidectomy. The Korean version of the Chronic Ear Survey (K-CES), a disease-specific outcome survey, was administered before and 12 months after surgery.

Main Outcome Measures The primary outcome measures were the total score and subscale score, including an activity restriction-based subscale, symptom subscale, and medical resource utilization subscale. The scores were analyzed with disease- and patient-related factors affecting health-related quality of life.

Results Of the 205 patients enrolled, complete data from 156 patients were available for this study. The K-CES score in patients with COM improved after ear surgery in total and in all subscales of the K-CES (P < .001). In univariable analysis, poor improvement in total K-CES score was more likely to occur after ear surgery in patients with postoperative complications, diabetes mellitus, hearing loss as a chief complaint, lower air conduction threshold in postoperative audiometry, and a high level of education and in those who underwent canal wall down mastoidectomy (P < .05). Multivariable analysis using multiple linear regression revealed that the occurrence of complications, presence of diabetes mellitus, level of education, and postoperative air conduction thresholds were independent factors for poor improvement in total K-CES score (P < .05).

Conclusion Recognition of factors that have significant relation with subjective outcomes may help surgeons to identify patients who are prone to have a lower satisfaction and provide useful information in preoperative counseling.

Chronic otitis media (COM) is a common disease, defined as a chronic inflammation in the mucosa of the middle ear and/or mastoid, which affects 2% to 3% of the global population.^1-3 Although the widespread use of antibiotics has decreased the incidence of COM and COM-related complications, surgery remains a main strategy in the curative treatment of COM, especially in patients with cholesteatoma.

Because the main purposes of surgical intervention in COM are to remove the intractable pathologic tissue and restore the normal functions of the middle ear as a sound conductor and the mastoid cavity as an air reservoir, the clinical outcome measurements of ear surgery have been based on disease control rate and audiologic result.^1,4

Health-related quality of life (HRQOL) has attracted growing interest over the past 20 years in the evaluation of ear surgery. Although relatively few studies have specifically evaluated HRQOL measures in patients with COM, some studies have evaluated HRQOL using patient-based outcome measurement tools.

Until recently, studies of HRQOL regarding otitis media have focused mainly on childhood otitis media using validated evaluation tools such as the OM-6 (Otitis Media–6 [6-item quality of life survey]) and OMO-22 (Otitis Media Outcome–22 [22-item questionnaire]), which are not adequate for adults with COM.^5-10 Studies involving adults have been conducted using various measurement tools including the Hearing Satisfaction Scale and its modified version,¹¹ the Hearing Handicap Inventory for the Elderly,¹² the Glasgow Benefit Inventory,¹³ and the 36-item Short-Form Health Survey.¹⁴ These tools are not COM specific and have limitations, since they are based mainly on hearing result and general health.

Compared with these surveys, the Chronic Ear Survey (CES), introduced by Nadol et al,¹ is a sensitive and disease-specific HRQOL measurement tool that has been validated in several reports.^1,4,15-18 A Korean version (K-CES) has been validated.¹⁹ Although new COM surveys that can reflect psychological impact such as anxiety and depression have been developed,²⁰ uniform studies using the CES could provide more reliable information about the subjective outcomes of COM.

Most studies using the CES have focused on the factors directly related to COM itself. Few studies have investigated personal factors like socioeconomic status and comorbidity. Moreover, the high rate of loss to follow-up and subsequent small size of study populations has made it difficult to analyze all the available factors.

The present study was designed to measure subjective outcomes using the CES and to analyze disease and patient-related factors affecting the HRQOL in a large series of patients with COM who underwent ear surgery.

The study was approved by the institutional review board of the Samsung Medical Center. A total of 205 patients who underwent ear surgery for COM at a tertiary referral hospital from November 2009 through March 2011 were included in this prospective survey. Data collection was completed in March 2012. Inclusion criteria included all of the following: patients whose preoperative diagnoses were COM with or without cholesteatoma, defined as a disease of the middle ear and/or mastoid with irreversible mucosal damage or infection lasting more than 3 months¹; patients who had undergone tympanoplasty with or without mastoidectomy; and age 18 years or older. Patients with other medical conditions that could affect HRQOL were excluded.

At the 1-year follow-up, the initial study population was reduced to 156 patients owing to missing questionnaire data (n = 10), incomplete audiologic data (n = 14), and loss to follow-up (n = 25). The K-CES, a disease-specific 13-item Likert scale outcome survey, consists of the following 3 categories: an activity restriction–based (AR) subscale, a symptom (ST) subscale, and a medical resource utilization (MR) subscale (Figure). The K-CES was performed preoperatively and at 12 months after surgery. Scoring for each K-CES question was normalized to a scale of 0 to 100, with 100 being the highest score. The total scores and sums of subscale scores were averaged on the basis of the number of questions included in each category.^1,4 Audiologic tests and clinical examinations were performed in all patients preoperatively and 12 months after surgery. In pure-tone audiometry, air conduction and bone conduction thresholds at 0.5, 1.0, 2.0, and 3.0 kHz were averaged. The air-bone gap was also measured in both ears. In clinical examination, endoscopic and microscopic evaluations were conducted for routine ear, nose, and throat examination and postoperative problems, such as wound infection, bare bone exposure, perforation, retraction, and discharge from the neoeardrum. If any of the postoperative problems lasting more than 1 month were noted, the patient was judged to have postoperative complication(s). Demographic data were collected preoperatively from the medical record.

The paired t test or Wilcoxon signed rank test was used to compare changes within a group. Univariable analyses (independent sample t test, Mann-Whitney test, 1-way analysis of variance, Kruskal-Wallis test, and Spearman rank correlations) were initially performed to determine the variables significantly associated with the K-CES score. Multivariable analysis using multiple linear regression model were then performed, including significant (P < .05) variables in the univariable analyses or clinically important variables, to determine factors that were independently predictive of the outcomes. P <.05 (2-sided) was considered statistically significant. Statistical analysis was performed using SPSS version 18.0 (SPSS Inc) statistical software.

The demographic data are summarized in Table 1. The study population comprised 156 patients who underwent ear surgery for COM and were followed up for at least 1 year. There were 95 women (60.8%) and 61 men (39.1%), and the median age at the time of surgery was 50.5 years (interquartile range [IQR], 42.1-59.0 years). Seventy-five patients (48.0%) had right-side COM and 81 (51.9%) had left-side COM. The most common symptom was otorrhea in 85 patients (54.5%), and 36 patients (23.1%) had active discharge at the time of operation. Forty-four patients (28.2%) had COM with cholesteatoma. Tympanoplasty without mastoidectomy, canal wall up mastoidectomy, and canal wall down mastoidectomy was conducted in 52 patients (32.3%), 72 patients (46.2%), and 34 patients (21.8%), respectively. There were 26 revision cases (16.7%) and 130 primary cases (83.3%). Postoperative complications occurred in 20 cases (12.8%).

Both the total K-CES score and all 3 subscales exhibited significant improvement (Table 2). The median preoperative total CES score was 69.1 (IQR, 53.8-77.1). This score improved to 92.4 (IQR, 86.7-96.9) at 1 year postoperatively (P < .001). The mean (SD) improvement in each subscale score were 19.6 (21.3) for the AR subscale, 27.5 (16.6) for the ST subscale, and 20.0 (IQR, 6.7-26.7) for the MR subscale.

The significant improvement in the air conduction thresholds and reduction of air-bone gaps were observed after ear surgery (Table 2). The median preoperative air conduction threshold was 47.5 dB hearing level (HL) (IQR, 35.0-60.0 dB HL) and improved to 30.0 dB HL (IQR, 20.0-50.0 dB HL) 1 year after surgery. The median preoperative air-bone gap was 25.0 dB (IQR, 20.0-40.0 dB) and improved to 10.0 dB (IQR, 5.0-25.0 dB).

To determine the factors predicting subjective outcomes, univariable analyses were performed with the difference of the K-CES score between preoperative and postoperative surveys.

Table 1 summarizes factors having significant univariate associations (P < .05), with the difference of total K-CES score representing HRQOL outcome. For disease-related factors, better improvement occurred in patients with otorrhea as a chief complaint and canal wall up mastoidectomy. For patient-related factors, the improvement of total scores decreased as the level of education increased. The presence of diabetes mellitus (DM) and postoperative complications were also associated with worse outcomes. However, multivariable analysis including sex, chief complaint, active discharge, cholesteatoma, surgery type, revision, occurrence of complications, alcohol use, presence of DM, marriage, hospital days, and postoperative air conduction thresholds demonstrated that postoperative air conduction thresholds, the occurrence of complications, presence of DM, and level of education were independent factors that had significant associations with improvement in total K-CES score (P < .05). Although the postoperative air conduction threshold was not associated with outcomes in univariable analysis, statistical significance was observed in multivariable analysis (Table 3).

This study was a large-scale study and a more comprehensive evaluation of general factors affecting HRQOL in patients with COM compared with many other studies on HRQOL that used various questionnaires focused on the effects of the disease itself. In our prospective study of HRQOL after ear surgery, we found that HRQOL was influenced by both patient- and disease-related factors.

Disease-related factors previously investigated include laterality, duration, activity of the disease, history of otological surgery, presence of cholesteatoma, occurrence of complications, and hearing status. The results of this study implicated complications and postoperative air conduction threshold as independent variables that can influence the CES score.

Nadol et al¹ reported a strong correlation between the preoperative and postoperative hearing results and the CES scores. Our previous analysis⁴ also demonstrated that air conduction thresholds show linear correlations with the total K-CES score and the ST subscale scores in the primary surgery group, but not in the revision surgery group. A larger-scale analysis in this study also showed that postoperative air conduction thresholds had significant correlation with the change in the total K-CES scores in multivariable analysis. Interestingly, in the analysis of subscale, the hearing results correlated with the AR subscale scores, rather than the ST subscale score, in which the questions about hearing loss (Figure, ST subscale item numbers S1 and S5) are included. Item number 3 in the AR subscale (Figure), which asks about the restriction of social activity by hearing loss, may have a large influence in patients who experience improvement in hearing.

Patients with cholesteatoma were reported to have moderate improvement in total K-CES score but very little improvement in the AR subscore.¹ Our study also showed little improvement in the AR subscore for patients with COM with cholesteatoma (P < .003) compared with those with COM without cholesteatoma (Table 1). However, the improvement in total K-CES score was not different between COM with and without cholesteatoma in multivariable analysis.

With regard to the activity of COM, in the present study, significant improvement was observed only in the MR subscale. In another study,¹ patients with frequent reactivation showed the greatest improvement in total scores and AR and ST subscores. Considering the easy accessibility to medical services in South Korea, draining from the ear may have been easily controlled and thus rarely restricted daily activity.

We also analyzed patient-related factors including patients' sex, age, history of alcohol and smoking, marriage status, presence of DM, level of education, and body mass index with regard to the subjective outcomes. Education level and presence of DM were independent variables that influenced the K-CES score. Previously, studies on socioeconomic factors contributing to HRQOL in otology focused mainly on hearing impairment,^21,22 and little research on management of inflammatory disease has been undertaken.

In contrast to the report that the presence of comorbidity did not predict postoperative CES outcome,¹ improvement in the present study was worse in patients with DM in total K-CES score and AR and ST subscales. Diabetes mellitus is well known to be firmly associated with impaired wound healing.^23-26 Altered activation and decreased number of macrophages may impair or delay wound healing.^27,28 In a study on tympanic membrane healing using animal models, diabetic rats showed a significant delay in the healing of tympanic membrane perforation.²⁹ With regard to the socioeconomic status, the present study revealed a significant association between education level and improvement of total K-CES score. We can speculate that patients with a higher level of education may expect better postoperative outcomes by more available medical information and, subsequently, will show relatively poor satisfaction for the results of surgery.

There are some limitations of this study. First, because the data analyzed in this study were collected from a single institution, generalization of the study results may be difficult. Second, although a large number of patients was included for final analysis, selection bias, such as including only compliant patients who were followed up for more than 1 year, may have influenced the results.

In conclusion, our results suggest that the occurrence of complications, worse postoperative air conduction thresholds, presence of DM, and higher level of education may be useful factors in predicting worse subjective outcomes in patients with COM who are undergoing ear surgery. Recognition of relevant factors may help surgeons to identify the patients who are prone to have a lower subjective outcome and provide useful information in preoperative counseling.

Correspondence: Yang-Sun Cho, MD, Department of Otolaryngology–Head and Neck Surgery, Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 Irwon-dong, Gangnam-gu, Seoul 135-710, Korea (yscho@skku.edu).

Submitted for Publication: May 14, 2012; accepted July 2, 2012.

Author Contributions: Drs Choi and Cho had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Choi and Cho. Acquisition of data: Choi, Cho, Chung, and Hong. Analysis and interpretation of data: Choi, Cho, N. J. Lee, and J. Lee. Drafting of the manuscript: Choi and Hong. Critical revision of the manuscript for important intellectual content: Cho, N. J. Lee, J. Lee, and Chung. Statistical analysis: Choi. Obtained funding: Cho. Administrative, technical, and material support: Cho, N. J. Lee, and J. Lee. Study supervision: Cho.

Funding/Support: This work was supported by grant 10031764 from the Strategic Technology Development Program of Ministry of Knowledge Economy and grant SS100022 of Seoul R&D Program.