聽力損失之效果研究

Outcome Study in Hearing Loss

趙庭寬(Ting-Kuang Chao)

指導教授 ： 陳秀熙

國立臺灣大學/公共衛生學院/預防醫學研究所/博士(2008年)

https://doi.org/10.6342/NTU.2008.00086

摘要

前言：
聽力損失會造成社會隔離與溝通障礙。在造成突發性聽力損失的疾病中，突發性耳聾仍有許多爭議性；雖然許多預後因子曾被報告過，但是卻沒有非常深入性的研究。除此之外，考慮所有可能的預後因子來對於突發性耳聾恢復做預測的研究很少，而且有許多未能考慮周詳之處。另一方面，老年性聽障是困擾許多人的疾病，雖然隨年齡增長聽力衰退的特性為人所熟知，但是仍舊缺乏對其自然病史的量化研究；除此之外，使用助聽器是目前老年性聽障的主要復健方法，但是對於其經濟效益評估仍有許多必須研究的部份。

目的：
(1) 評估突發性耳聾相關之預後因子，尤其針對變頻耳聲傳射，評估其是否可以作為突發性耳聾預後的客觀聽覺指標。 (2) 建立突發性耳聾恢復之預測模式，提供為臨床之應用。 (3) 發展量化老年性聽障聽力退化的適當模式；並且建立新式決策方式及機率性分析模式，作為評估老年人配戴助聽器之成本效益分析。

方法：
(1) 共108位突發性耳聾患者納入研究，使用依時性Cox回歸模式評估變頻耳聲傳射以及其他可能之預後相關因子。兩次變頻耳聲傳射平均值將其歸類為大於等於6分貝以及小於6分貝，並且使用評估恢復時間前3日之最近資料做為分析。 (2) 以200位突發性耳聾患者之臨床資料建立預測模式。首先以lognormal accelerated failure time model預測自接受治療後到出現有意義之聽力恢復之恢復時間；其次使用cure rate model預測長期恢復之比例。使用貝氏分析，藉由WinBUGS程式以預測恢復時間以及長期恢復比例之中位數及95%可信區間。 (3) 在老年性聽障，我們將聽力衰退建立為四階段馬可夫模式，使用不同文獻之統合分析結果可以獲得其衰退速度及預測不同階段之機率值；對於助聽器之成本效益分析，則使用機率性分析及蒙地卡羅模擬，藉以獲得不同成本時符合成本效益之機率分配曲線。

結果：
(1) 結果顯示變頻耳聲傳射較佳為突發性耳聾較佳預後之顯著指標，在單因子分析(recovery rate ratio = 3.626, 95%CI = 2.119~6.205, p<0.0001)及多因子分析(recovery rate ratio= 2.94, 95%CI = 1.537~5.624, p=0.0011)皆具有顯著意義；年紀較輕、發作時具有較佳聽力者、以及具有正常聽覺腦幹反應與前庭誘發肌電位者會有較好之預後；另外，發作時之聽力圖形式也是有意義的預後指標。 (2) 預測模式之結果與觀察資料比較後發現，此預測模式具有相當好之內在效度。兩種預測模式在三星期內皆有相當正確之預測結果，但長期恢復率預測結果則以cure rate model較佳。我們將所有預測指標不同組合之預測結果，整理於同一個Access檔案中，以利於臨床之應用。 (3) 在老年性聽障，聽力衰退之速度隨頻率增高而增加；男性除了90歲以上之外，在所有的頻率以及其他年齡層皆有較快的衰退速度；雙耳比較時，左耳在老年初期較右側有較快之衰退速度，但之後兩耳之衰退速度相當；使用所得之速率值可以獲得各個聽力階段的機率值。而助聽器佩帶之成本效益分析結果顯示，獲得一額外之聽力相關生活品質調整後人年，所需增加之花費在女性為NT$ 461,154元，在男性為NT$ 329,010元。不同成本時符合成本效益之機率分配曲線顯示，當願付價格為NT$ 400,000元時，機率值在女性為52%，男性為64%；當願付價格為NT$ 700,000元時，機率值達到穩定水準，在女性為67%，男性為78%。

結論：
部分耳部疾病造成之聽力損失可以在治療後得到改善，突發性耳聾即為一典型的例子。此篇研究的第一部分結果，可以提供一個適當的統計模式，作為評估這類病患的預後因子以及治療方式；預測模式的開發則可為臨床應用時提供很好的參考。另一方面，有一些耳部疾病的聽力損失為漸進性，以目前的治療並無法將聽力恢復，老年性聽障即為其中之代表，預防及復健為最重要的課題。藉由此研究可以了解聽力惡化的自然史，並可以提供適當之模式藉以評估危險因子，作為未來研究之基礎。使用多階段模示考慮不同聽力階段之影響，以此研究之結果顯示，在老年性聽障者，助聽器配戴確實為符合成本效益之復健方式，並可作為政策決策者分派健康資源之參考。

關鍵字

痊癒率模式 ； 變頻耳聲傳射 ； 馬可夫鏈 ； 預測 ； 老年性聽障 ； 生活品質 ； 突發性聽力損失

並列摘要

Background: Hearing impairments would cause social isolation and difficulties in communications. Among diseases that cause sudden hearing deterioration, idiopathic sudden sensorineural hearing loss (ISSHL) is still a controversial topic. Although several prognostic factors had been reported, there is lacking of in-depth analysis. Besides, the prediction of recovery in ISSHL with the incorporation of all putative factors has been barely investigated. Only few studies have covered this theme. In addition to ISSHL, most people suffered from age-related hearing loss (ARHL) with aging. The progressive type has been well documented but the disease natural history is poorly understood. In addition, the mainstay treatment for such kind of hearing loss relies on hearing aids fitting. However, the efficacy of such assistance is also lacking of systematic evaluation.

Objective: (1) To evaluate the prognostic factors associated with ISSHL, especially focused on the distortion-product otoacoustic emissions (DPOAE), an objective auditory indicator as a potential prognostic factor for ISSHL patients. (2) To establish a predictive model for recovery of ISSHL for clinical application. (3) To develop an appropriate model in quantifying the progression of hearing in ARHL and to establish a new model in performing a probabilistic analysis of cost-effectiveness for hearing aids fitting in elders.

Methods: (1) One hundred and eight patients with ISSHL were included. Time-dependent Cox regression model was used in evaluation of DPOAE and other prognostic factors. The average of two series of DPOAE intensity corrected for the noise level in 8 frequencies was coded as dichotomous as U>U6 dB or <6 dB. We selected the most updated DPOAE at 3 days or more before the assessment of recovery into the analysis. (2) Two hundred patients with ISSHL were recruited to build up the predictive model. First, the log-normal accelerated failure time model was applied to investigate the time duration from treatment to a significant recovery of the hearing. Second, the cure rate model was utilized to predict the long-term cumulative recovery rate by taking into account the extent of susceptibility to recovery. Bayesian approaches of both models with WinBUGS program were applied in predicting the median and 95% credible interval of the recovery time or long-term cumulative recovery rate. (3) In ARHL, we modeled the disease process for hearing loss as a four-state continuous-time Markov process. The pooled estimates were obtained and compared with each other. With the estimates, we can also predict all transition probabilities between states. Then a decision model run with different hearing states was used in evaluation of cost-effectiveness for hearing aids fitting. Probabilistic approach with Monte Carlo simulation was performed to produce the cost effectiveness acceptability curve of showing the probabilities of being cost-effective given a series of willingness to pay (WTP).

Results: (1) In ISSHL, the results showed that a better DPOAE amplitude was a significant good prognostic indicator both in univariate analysis (recovery rate ratio = 3.626, 95%CI = 2.119~6.205, p<0.0001) and multivariate analysis (recovery rate ratio= 2.94, 95%CI = 1.537~5.624, p=0.0011). Younger age (age<= 40 years), better initial pure tone audiometry (< 65 dB), normal auditory brainstem response and vestibular evoked myogenic potential represented a better prognosis. The configuration of initial audiometry was also a significant prognostic factor. (2) Predicted results were compared with the observed values and the model showed a good internal validity. However, both models have equivalent predictive accuracy within three weeks after onset of ISSHL but the cure rate model had a better long-term prediction than the lognormal model. The overall results predicted by different combinations of covariates were summarized and organized in an Access program file that is convenient for clinical application. (3) In ARHL, the progression of hearing loss increased with the ascending frequencies across all age groups. Males had significantly faster progression rates in all frequencies and age groups except for the age group of 90 years or older. In comparison between ears, the progression of hearing would be slightly faster in left ears initially in early elder life and did not show any difference in further aging and later hearing declines. With the pooled estimates of progression rates, the probabilities of hearing deterioration could be obtained. With regard to the cost-effectiveness of hearing aids fitting, the incremental cost for an additional HQALY gain in women and men were NT$ 461,154 and NT$ 329,010 respectively. Using acceptability curve, the probability of being cost-effective increased up to 52% in women and 64% in men given a WTP of NT $ 400,000. It would reach plateau with 67% in women and 78% in men given a WTP of NT $ 700,000.

Conclusions: In some kinds of ear diseases, hearing loss might be reversible after treatment. ISSHL is a typical type that treatment might improve the prognosis. The first part of our study could offer an appropriate statistical model in evaluating prognostic factors and different treatment protocols for ISSHL patients. The development of the predictive models could offer a good reference for clinical consultations. On the other side, in some diseases, hearing loss is progressive and can’t be restored with present treatments and techniques. ARHL is a representative type that prevention and rehabilitation are the major issues. In this study, we realize the natural course of the hearing progression and could offer a model in evaluating the effects of risk factors. The establishment of an appropriate model in quantifying the hearing progression in ARHL would be helpful in future researches. By using a multi-state model with different hearing states, hearing aid fitting is demonstrated as a cost-effective strategy in the rehabilitation for hearing-impaired elders. The results could help policy makers in allocating the health resources more appropriately.

並列關鍵字

Cure rate model ； Distortion product otoacoustic emissions ； Markov chains ； Prediction ； Presbycusis ； Quality of life ； Sudden hearing loss

參考文獻

3. Laird N, Wilson WR. Predicting recovery from idiopathic sudden hearing loss. Am J Otolaryngol 1983;4:161-4.

4. Fetterman BL, Saunders JE, Luxford WM. Prognosis and treatment of sudden sensorineural hearing loss. Am J Otol 1996;17:529-36.

5. Park HM, Jung SW, Rhee CK. Vestibular diagnosis as prognostic indicator in sudden hearing loss with vertigo. Acta Otolaryngol Suppl 2001;545:80-3.

6. Kiris M, Cankaya H, Icli M, Kutluhan A. Retrospective analysis of our cases with sudden hearing loss. J Otolaryngol 2003;32:384-7.

7. Sakashita T, Minowa Y, Hachikawa K, Kubo T, Nakai Y. Evoked otoacoustic emissions from ears with idiopathic sudden deafness. Acta Otolaryngol Suppl 1991;486:66-72.