Bexelius Christin, Honeth Louise, Ekman Alexandra, Eriksson Mikael, Sandin Sven, Bagger-Sjöbäck Dan, Litton Jan-Eric
Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Nobels väg 12A, SE-171 77 Stockholm, Sweden.
J Med Internet Res. 2008 Oct 21;10(4):e32. doi: 10.2196/jmir.1065.
Hearing impairment is most accurately measured by a clinical pure-tone audiogram. This method is not suitable for large-scale, population-based epidemiological studies as it requires that study participants visit a clinic with trained personnel. An alternative approach to measuring hearing ability is self-estimation through questionnaires, but the correlation to clinical audiometric tests varies.
To evaluate an Internet-based hearing test pilot compared to a question about self-estimated hearing and the feasibility of using an Internet-based hearing test and an Internet-based questionnaire in a population of 560 members of the Swedish Hunters' Association in the age group 20-60 years.
An invitation was mailed to the participants in March 2007 together with the URL to the study Web site, a personal username, and a password. The Web site included the questionnaire, the hearing test, and instructions for participating in the study. The hearing test resembles a clinical audiogram presenting 6 tones between 500 and 8000 Hz. Tones are presented between 0 and 60 dB, and the participant responds to the tones by pressing the space bar. The hearing test requires headphones and is based on JAVA programming. Before the participant can start the hearing test, it has to be calibrated against a reference person with good hearing between 15 and 35 years of age.
After 5 months, 162 out of 560 (29%) had answered the questionnaire, out of which 88 (16%) had completed the hearing test. Those who actively declined participation numbered 230 out of 560 (41%). After removing duplicates and hearing tests calibrated by unreliable reference data, 61 hearing tests remained for analysis. The prevalence of hearing impairment from the Internet-based hearing test was 20% (12 out of 61), compared to 52% (32 out of 61) from the self-estimated question. Those who completed the hearing test were older than the non-participants, and more had headphones (P = .003) and the correct version of the JAVA program (P = .007) than those who only answered the questionnaire.
Though an Internet-based hearing test cannot replace a clinical pure-tone audiogram conducted by a trained audiologist, it is a valid and useful screening tool for hearing ability in a large population carried out at a low cost.
听力障碍最准确的测量方法是临床纯音听力图。这种方法不适用于大规模的基于人群的流行病学研究，因为它要求研究参与者前往配备训练有素人员的诊所。测量听力的另一种方法是通过问卷调查进行自我评估，但与临床听力测试的相关性各不相同。
在年龄在20 - 60岁的560名瑞典猎人协会成员中，评估基于互联网的听力测试试点与自我估计听力问题的比较，以及使用基于互联网的听力测试和基于互联网的问卷的可行性。
2007年3月向参与者邮寄了邀请函，同时附上研究网站的网址、个人用户名和密码。该网站包括问卷、听力测试以及参与研究的说明。听力测试类似于临床听力图，呈现500至8000赫兹之间的6个音调。音调在0至60分贝之间呈现，参与者通过按空格键对音调做出反应。听力测试需要使用耳机，并且基于JAVA编程。在参与者能够开始听力测试之前，必须针对一名15至35岁听力良好的参考人员进行校准。
5个月后，560名中有162名（29%）回答了问卷，其中88名（16%）完成了听力测试。主动拒绝参与的有560名中的230名（41%）。在去除重复项和由不可靠参考数据校准的听力测试后，剩下61项听力测试用于分析。基于互联网的听力测试得出的听力障碍患病率为20%（61名中的12名），而自我估计问题得出的患病率为52%（61名中的32名）。完成听力测试的人比未参与者年龄更大，并且拥有耳机（P = 0.003）和正确版本JAVA程序（P = 0.007）的人比仅回答问卷的人更多。
虽然基于互联网的听力测试不能替代由训练有素的听力学家进行的临床纯音听力图，但它是一种以低成本在大量人群中进行听力能力筛查的有效且有用的工具。
