Quality-of-Life Assessment After Primary and Revision Ear Surgery Using the Chronic Ear Survey

Objective  To measure subjective outcomes after primary and revision surgery for chronic ear disease.

Design  Prospective questionnaire-based outcome study.

Setting  Tertiary referral center.

Patients  Adults with chronic otitis media with or without cholesteatoma.

Interventions  Primary or revision surgery for chronic ear disease.

Main Outcome Measures  The Chronic Ear Survey, a disease-specific outcome survey, was administered preoperatively and at 1 year after surgery. We analyzed the total score and the activity restriction, symptom, and medical resource utilization subscale scores. Scores were averaged on the basis of the number of questions included in each category. Differences in preoperative and postoperative scores were analyzed within and between the 2 groups. We also assessed audiometry, postoperative complications, and the clinical condition of the operated-on ear.

Results  Twenty-one patients were enrolled in the primary surgery group, and 20 were enrolled in the revision surgery group. Significant improvements in the total score and each subscale score were observed in both groups at the 1-year postoperative survey. Improvements in the total score and symptom subscale scores were greater in the primary surgery group than in the revision surgery group (P < .05). The air conduction thresholds and any postoperative clinical problems were correlated with the total score and subscale scores in the primary group but not in the revision surgery group.

Conclusion  Comparable objective outcomes are achieved after primary and revision surgery for chronic ear disease, but the improvement in quality of life is greater in the primary surgery group.

The primary goal in the surgical management of chronic otitis media (COM) is eradication of infection and any accompanying cholesteatoma. Second-look or revision surgery is required when a dry and safe ear is not achieved. Thirteen percent to 18% of patients with COM who have undergone a canal wall down mastoidectomy (CWDM)^1-3 and 11% to approximately 34% of patients who have undergone a canal wall up mastoidectomy (CWUM) experience recurrence.^1-4 Until recently, the outcome measurements of surgery for chronic ear disease have focused on hearing improvement and disease control rate, which are assessed by objective measurement tools. In some cases, a discrepancy exists between the postoperative audiometric results or physical status of the ear and patient satisfaction with ear symptoms in daily life. For this reason, several trials have evaluated patient-based outcomes after ear surgery.^5,6

A disease-specific 13-item Likert scale outcome survey, the Chronic Ear Survey (CES), was introduced to evaluate the quality of life (QOL) in patients after surgery for chronic ear disease.⁷ The survey consists of the following 3 categories: an activity restriction–based subscale (AR), a symptom subscale (ST), and a medical resource utilization (MR) subscale (Table 1). This survey is the only sensitive and disease-specific QOL measurement tool that has been validated in patients with chronic ear disease. Various aspects of subjective outcomes after surgery for chronic ear disease have been described in several reports using this questionnaire.^8-11

However, the subjective outcomes after revision surgery have not been fully studied. Revision surgery for recurrent COM is more extensive, and the corresponding hearing outcomes are usually poorer than those achieved after primary surgery.¹² Moreover, because the goals of management are different, postoperative subjective satisfaction after primary and revision surgery may be also different.

The objectives of this study were to (1) measure subjective outcomes using the CES after surgery for chronic ear disease, (2) compare CES scores after primary or revision surgery, and (3) correlate CES scores with objective clinical outcomes.

The institutional review board of the Samsung Medical Center granted approval for this study. Thirty patients who underwent revision surgery for chronic ear disease and 30 patients who underwent primary surgery at a tertiary referral hospital in 2005 were enrolled in this prospective study. Eighteen patients in each group had COM with cholesteatoma and 12 patients in each group had COM without cholesteatoma. Subjects underwent CWDM or CWUM and were followed up for more than 12 months. Patients in the primary group were matched with those in the revision group with respect to the type of surgery, diagnosis, and other demographic findings. Patients with other medical conditions that could affect QOL were excluded from this study. The CES was administered preoperatively and at 1 year postoperatively, and differences in scores within and between groups were analyzed. Scoring for each 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.

We conducted pure-tone audiometry in all patients preoperatively and postoperatively, at the time of the survey. Air conduction (AC) and bone conduction (BC) thresholds at 0.5, 1.0, 2.0, and 3.0 kHz were averaged. The air-bone gap (ABG) was also calculated in affected and contralateral ears. Postoperative problems such as wound infections, exposure of bare bone in the external auditory canal, perforation of the neoeardrum, and recurrent otorrhea during the 12 months after surgery were also assessed. The final condition of the ear undergoing surgery was described as dry or draining on postoperative surveillance. If any of the postoperative problems lasting more than 1 month were noted at the 1-year postoperative surveillance visit, or if the ear continued to drain, then the patient was judged to have postoperative clinical problems.

Statistical analysis was performed using SPSS statistical software, version 13.0 (SPSS Inc, Chicago, Illinois). We used the χ² and Fisher exact tests to compare nonparametric demographic variants. We adopted the Mann-Whitney test to compare parametric mean values and used the paired t test to compare changes within groups. We evaluated correlations between CES scores and objective outcomes using Spearman rank correlation analysis. P < .05 was considered statistically significant. Unless otherwise indicated, data are expressed as mean (SD).

The initial study populations were reduced in number to 21 in the primary surgery group and to 20 in the revision surgery group at the 1-year follow-up. Three subjects were dropped from the study because the questionnaire was not completed, 2 were dropped because of missing audiometric data, and 14 were lost to follow-up. Demographic features of the enrolled patients are listed in Table 2. The mean ages were 46.9 (10.2) years in the primary surgery group and 47.7 (10.0) years in the revision surgery group. The male to female ratios were 12:9 in the primary surgery group and 9:11 in the revision surgery group. Because most of the patients were diagnosed as having primary or recurrent cholesteatoma (31 of 41 [76%]), CWDM was mainly used for mastoid procedures in the primary (17 of 21 [81%]) and revision (16 of 20 [80%]) surgery groups. The most common symptoms were otorrhea (27 of 41 [66%]) and hearing loss (12 of 41 [29%]). There were no statistically significant differences between the primary and revision surgery groups with respect to sex, age, laterality, chief symptom reported, second-stage ossiculoplasty, or the presence of contralateral ear disease (P ≥ .05).

The mean preoperative total score in the primary surgery group was 44.4 (15.7). This score improved to 77.3 (17.4) at 1 year postoperatively (P < .001; Figure 1), and 20 of 21 patients (95%) had improvement in their total outcome scores. Patient 18 in the primary surgery group, who had a slight decrease in the total score, had developed draining myringitis of the operated-on ear at the time of the survey, which might have affected the ST score. The improvements in each subscale score were 17.4 (17.7) for the AR, 39.4 (20.0) for the ST, and 33.3 (24.5) for the MR. The increases in the total score and each subscale score were significant (P < .01; Figure 2).

The mean preoperative total CES score in the revision surgery group was 50.7 (10.0). This improved to 73.7 (10.8) at 1 year postoperatively (P < .001; Figure 1). At that time, all patients exhibited improvement in their total outcome scores and significant increase in their subscale scores, including 13.7 (14.4) for the AR, 26.4 (15.8) for the ST, and 24.3 (17.5) for the MR (P < .05; Figure 2). There was a positive correlation (ρ = 0.45; P < .05) between the number of revision surgeries and the preoperative ST subscale scores.

On cross-sectional analysis, no differences were noted in the preoperative or postoperative total scores or MR subscale scores between the 2 groups. The preoperative ST subscale score was higher in the revision surgery group than in the primary surgery group. Postoperatively, the AR subscale score was lower in the revision surgery group than in the primary surgery group (Figure 1). On analysis of the changes observed after surgery, improvement in the total score was greater in the primary surgery group than in the revision surgery group (P < .03). The primary surgery group also exhibited greater increases in each subscale score, but a statistically significant increase was noted only in the ST subscale score (Figure 2).

When analyzing postoperative changes in subjective outcomes, we found that increasing age was positively correlated with changes in the AR (ρ = 0.39; P = .01) and ST subscale scores (ρ = 0.34; P = .03) and the total score (ρ = 0.35; P = .02). More postoperative improvements in subjective outcomes were obtained in elderly patients, a finding supported by the report by Nadol et al.⁷ In addition, CWDM resulted in less improvement in the MR subscale score (ΔMR = 24.4 [16.6]) than did CWUM (ΔMR = 47.5 [30.3]; P = .04). This observation can be explained by the fact that patients who undergo CWDM need frequent visits after surgery because of longer healing time.¹²

The AC thresholds and ABGs on preoperative and postoperative audiograms were not significantly different within groups or between groups (P ≥ .05). The mean preoperative AC thresholds were 65.3 (28.4) dB hearing loss (HL) in the primary surgery group and 56.8 (17.5) dB HL in the revision surgery group. One year after surgery, the mean AC thresholds were 53.8 (35.4) dB HL in the primary surgery group and 50.6 (18.1) dB HL in the revision surgery group. The mean preoperative ABGs were 31.0 (14.2) dB in the primary surgery group and 29.3 (10.4) dB in the revision surgery group. One year after surgery, the mean ABG had improved to 21.5 (16.1) dB in the primary surgery group and to 24.9 (11.0) dB in the revision surgery group. Nine patients (43%) in the primary surgery group showed AC threshold improvement of greater than 10 dB, whereas 8 patients (40%) in the revision surgery group exhibited the same improvement (Table 3). In addition, there were no significant differences in the mean AC thresholds of contralateral ears between groups, with 38.7 (29.4) dB HL in the primary group and 30.0 (22.6) dB HL in the revision group (P = .32).

The incidence of postoperative clinical problems was no different between the 2 groups, and about half of the enrolled patients in each group (10 patients each) experienced problems postoperatively (Table 4). Exposure of bare bone in the external auditory canal and recurrent otorrhea were the most frequent minor complications in both groups. A dry healthy ear was achieved in 20 patients in the primary surgery group (95%) and in 16 patients in the revision surgery group (80%) at 1 year after surgery (P = .18).

The ST subscale scores and total CES score had significant negative linear correlation with the AC thresholds in the primary surgery group preoperatively and postoperatively, but not in the revision surgery group (Figure 3). Other subscale scores exhibited negative correlation trends against the hearing thresholds, but not to a statistically significant degree (P ≥ .05). The total CES score changes were not correlated with the magnitude of improvement in AC thresholds or ABG closure in either group (P ≥ .05).

Patients with postoperative clinical problems had wider-ranging CES scores compared with patients without clinical problems (Table 5). The presence of clinical problems adversely affected the total CES score and AR subscale score in the primary surgery group (P < .01), but not in the revision surgery group.

The aims of this study were to determine the subjective outcomes after primary and revision surgery for chronic ear disease and to correlate these data with objective outcomes. Our findings indicate that improvements in subjective outcomes were greater and also correlated with clinical status in the primary surgery group. The following theories could help to explain these observations: (1) patients undergoing primary surgery are more vocal about their ear symptoms before surgery; (2) patients undergoing revision surgery are more accustomed to their clinical condition; (3) those undergoing revision surgery are more concerned about disease eradication than about functional outcomes; and (4) the pursuit of revision surgery was based more on the physician's judgment than on the patient's need.

Surgical outcomes in the setting of chronic ear disease are assessed by using physician-based measurements or patient-based outcome measurements. Physician-based measurements, which were used in most reports,¹³ include complete eradication of the disease, creation of a dry and safe ear, and restoration of hearing using appropriate middle ear reconstruction. Conversely, patient-based outcome measurements evaluate QOL as a result of surgery using validated instruments. Because subjective outcome measures do not always correspond to objective outcome measures,⁶ a comprehensive outcome assessment of both measures is required.

Recently, the importance of health-related QOL measures has been underscored in the clinical literature. For example, a number of QOL studies have been conducted in children with otitis media by using disease-specific, parent-based evaluation tools.^14-19 Likewise, the outcomes associated with myringotomy or tympanostomy tube insertion have been analyzed using some of those subjective measures. However, these measurement tools have limitations with regard to proxy use in assessment, and they are not adequate for QOL evaluation in adults with chronic ear disease. Other subjective outcome measurement tools have been adopted in several other reports: the Hearing Satisfaction Scale and its modified version,²⁰ the Hearing Handicap Inventory for the Elderly,²¹ the Glasgow Benefit Inventory,²² and the 36-item Short-Form Health Survey.²³ However, these surveys are hearing related or general health related and are not adequate for evaluating disease-specific characteristics. The CES has been validated in several reports,^7-11 and is currently the only disease-specific QOL outcome survey for chronic ear disease. The survey consists of 13 Likert scale questions subdivided into 3 categories. The AR subscale examines the effect of COM on the patient's daily life, the ST subscale evaluates the presence of symptoms such as hearing loss and drainage, and the MR subscale determines the degree to which antibiotics and physician services are used. Nadol et al⁷ reported that prior otological surgery did not predict postoperative CES outcomes; other variables such as surgical procedures, surgeon experience, demographic features (except patient age), and comorbidities did not affect the CES.

In this study, patients in primary and revision surgery groups showed improvements in QOL 1 year after surgery. Although some of the subscale scores showed significant differences at surveillance time, the mean total CES scores were no different between the primary and revision surgery groups, preoperatively or postoperatively (Figure 1). Postoperative AR subscale scores were lower in the revision surgery group than they were in the primary surgery group. This may suggest that patients who had undergone previous surgery were more cautious about their everyday activities even after undergoing curative surgery. However, the preoperative ST subscale score was significantly lower in the primary surgery group, and it was more significantly improved in this group after surgery despite the presence of comparable hearing and clinical outcomes in both groups. These findings suggest that the subjective symptoms of chronic ear disease, such as hearing loss and ear discharge, were more important and influential in the primary surgery group. Conversely, patients who underwent revision surgery may have been more adapted to their symptoms as supported by a positive correlation (ρ = 0.45; P = .045) between the number of surgeries and the preoperative ST subscale scores. This is also supported by the fact that AC thresholds were linearly correlated with preoperative and postoperative ST subscale scores in the primary surgery group, whereas hearing thresholds did not affect the total CES score or any subscale scores in the revision surgery group (Figure 3). The MR subscale scores were lower than the other subscale scores across the groups, and the differences were greater compared with those (about 60 to 80 points) from other reports.⁷ This may be attributable to the easy accessibility of medical resources in South Korea compared with other countries.

Total CES scores improved more in the primary surgery group than they did in the revision surgery group. Larger improvements in all of the subscale scores were also noted in the primary surgery group; especially with regard to the ST subscale scores (Figure 2). In other words, patients who underwent revision surgery reported less improvement in QOL compared with those who underwent primary surgery. Although postoperative CES outcomes were no different between the primary and revision surgery groups, as Nadol et al⁷ reported, the amount of improvement was greater in the primary surgery group. They reported that patients with lower total CES scores had more significant postoperative improvement, and lower individual scores were also predictive of a postoperative increase in the CES score.⁷ The greater improvement in the total CES score and ST subscale scores in the primary surgery group in our series may be partly a result of lower preoperative ST subscale scores.

Regarding objective physician-based outcomes, average hearing did not improve in either group. Because most of the patients had cholesteatomas and had undergone CWDM, and because ossiculoplasty was conducted in only half of the patients, the hearing results in our series were relatively poor for both groups (Table 2).

Because questions about hearing loss are included in the ST subscale, we expected objective hearing thresholds to affect the symptom scores in both groups. However, only in the primary surgery group did AC thresholds show linear correlations with the total CES score and the ST subscale scores. In addition, the presence of postoperative problems affected AR subscale scores in the primary surgery group but not in the revision surgery group (Table 5). As mentioned previously, these findings suggest that patients who underwent revision surgery were more adapted to their status, including their hearing loss, and were more cautious in their daily lives, regardless of the condition of the ear that underwent the operation. Collectively, objective outcome measures could not predict the subjective QOL for the revision surgery group.

Despite the prospective design of this study, it has some limitations. First, the populations initially recruited for each group were reduced by one-third because of missing questionnaire data, incomplete audiologic testing, and loss to follow-up. Second, because most of the enrolled patients underwent CWDM procedures, the applicability in the setting of common chronic ear disease is not certain. However, our findings clearly indicate that improvements in subjective outcomes are of great importance and more strongly correlated with clinical status in the primary surgery group compared with the revision surgery group.

In conclusion, the primary and revision surgery groups showed improvement in QOL 1 year after surgery. Despite comparable objective outcomes, the improvement in QOL was greater in the primary surgery group. In the primary surgery group, the total CES score and some subscale scores showed significant correlations with the objective clinical outcomes. However, subjective outcomes did not correlate with objective outcomes in the revision surgery group.

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

Submitted for Publication: December 16, 2008; final revision received August 23, 2009; accepted November 5, 2009.

Author Contributions: Drs Jung 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: Jung, Cho, S. H. Hong, Chung, and S. D. Hong. Acquisition of data: Jung, Lee, and S. D. Hong. Analysis and interpretation of data: Jung and Cho. Drafting of the manuscript: Jung, Cho, S. H. Hong, Chung, Lee, and S. D. Hong. Critical revision of the manuscript for important intellectual content: Jung and Cho. Statistical analysis: Jung. Study supervision: Cho, S. H. Hong, and Chung.

Financial Disclosure: None reported.