There is currently insufficient effective treatment for hearing loss in the elderly since it is a hidden disease whose damage is irreversible. It is crucial to establish early warnings, screenings and interventions. As of now, there are few studies carried out on the assessment of hearing loss in the elderly based on risk factors at home and abroad. There are no standardized measurement tools or perfect scales.

Utilizing smart medicine and the WeChat platform, investigate high-risk factors of hearing loss in the elderly, develop screening software for hearing loss in the elderly, and explore the screening and management modes of hearing loss in the elderly.

Based on cross-sectional survey, five community health service centers in Pudong New Area of Shanghai were obtained. The tudy was performed from April to December 2019 to investigate the distribution of risk factors for hearing loss in the elderly in the community, and conditional logic was applied, and receiver operation charateristic curve (ROC curve) were used for risk stratification. The screening software for hearing loss in the elderly has been developed using JavaScript language during the period January and June 2020. Verification and evaluation of the screening software were performed between July 2020 to March 2021.

The study involved 401 elderly peoplein across-sectional design. Multivariate Logistic regression analysis revealed that aging〔OR=1.100, 95%CI (1.037, 1.166) 〕, noise history〔OR=3.886, 95%CI (1.077, 14.022) 〕, non-light diet〔OR=2.445, 95%CI (1.127, 5.305) 〕, hypertension〔OR=1.8393, 95%CI (1.015, 3.330) 〕, diabetes〔OR=4.310, 95%CI (1.817, 10.225) 〕and hyperuricemia〔OR=3.174, 95%CI (1.030, 9.779) 〕were independent risk factors (P<0.05) . A total of 18 factors (male, overweight/obesity, living alone, widowed/divorced, noise history, family history of deafness, non-light diet, no exercise habits, smoking, drinking, wearing headphones, hypertension, diabetes, hyperlipidemia, cardiovascular and cerebrovascular diseases, hyperuricemia, hypothyroidism, ototoxic drug use history) were included in the analysis as a result of the difference analysis and literature review. According to the ROC curve, combined scores of risk factors can predict hearing loss in the elderly with an area under the curve (AUC) of 0.777〔95%CI (0.721, 0.833) 〕, and the cut off value is 3.5. According to this study, a cumulative risk factor score of 4 defined the threshold for predicting hearing loss in the elderly. The elderly were then stratified into those with low risk of hearing loss (<4) and those with high risk of hearing loss (≥4) . The software for screening elderly hearing loss is developed on the WeChat platform. There are four parts in total: risk factors stratification assessment, screening version of the Hearing Impairment Scale (HHIE-s) for the elderly, general conclusions, and health education. From July to December 2020, a total of 78 elderly people were recruited to evaluatethe hearing loss screening software, with a completion rate of 55.1% (43/78) .A mode ratecor relation exists between cumulative risk factor scores and HHIE-s scale score (r_s=0.360, P=0.018) . From January to March, 2021, a suitability evaluation questionnaire with 8 single item questions was administered to 106 general practitioners to determine the suitability of hearing loss screening software for the elderly. According to the 5-level Likert scale, the proportion of respondents who answered "completely agree" to each question is 85.8% (91/106) , 81.1% (86/106) , 71.7% (76/106) , 68.9% (73/106) , 68.0% (72/106) , 59.4% (63/106) , 15.1% (16/106) and 14.2% (15/106) respectively.

There are 18 risk factors for hearing loss in the elderly. The screening software for hearing loss in the elderly based on WeChat platform has a good effect in early warning of hearing loss in the elderly, which provides a medical basis for screening for hearing loss in the elderly. Hearing loss screening software can support real-time data transmission, optimize, and integrate the hierarchical assessment system of risk factors, HHIE-s and WeChat platform for self-health management. The screening software of hearing loss for the elderly offers general practitioners a new way to manage and control hearing loss provides a new way of hearing loss management and control for general practitioners at the grassroots.It is appropriate and enforceable.

老年听力损失发展隐匿，损伤不可逆转，目前尚无治愈方法。如何进行早期预警、筛查、干预至关重要。目前国内外基于危险因素进行老年听力损失风险评估的研究较少，没有相应的测量工具及完善的量表。

基于智慧医疗，依托微信平台，分析老年听力损失高危因素，开发老年听力损失筛查软件，探讨老年听力损失的筛查及管理模式。

2019年4—12月，基于横断面调查获得上海市浦东新区5家社区卫生服务中心的老年听力损失危险因素分布，采用条件逻辑方式及受试者工作特征曲线（ROC曲线）进行危险分层评估。于2020年1—6月依托微信平台，采用JavaScript语言编写，开发老年听力损失筛查软件。2020年7月至2021年3月，对筛查软件进行验证及适宜性评价。

本研究采用便利抽样法共纳入401位老年人，多因素Logistic回归分析显示年龄大〔OR=1.100，95%CI（1.037，1.166）〕、噪音史〔OR=3.886，95%CI（1.077，14.022）〕、非清淡饮食〔OR=2.445，95%CI（1.127，5.305）〕、高血压〔OR=1.839，95%CI（1.015，3.330）〕、糖尿病〔OR=0.249，95%CI（1.817，10.225）〕、高尿酸血症〔OR=3.174，95%CI（1.030，9.779）〕是老年听力损失的独立危险因素（P<0.05）。结合差异性分析及文献检索共纳入18项危险因素（男性、超重/肥胖、独居、丧偶/离异、噪音史、失聪家族史、非清淡饮食、无锻炼习惯、吸烟、饮酒、佩戴耳机史、高血压、糖尿病、高脂血症、心脑血管疾病、高尿酸血症、甲状腺功能减退症、耳毒性药物使用史），危险因素累计得分预警老年听力损失的ROC曲线下面积为0.777〔95%CI（0.721，0.833）〕，最佳截断值为3.5。本研究将危险因素累计得分界值定义为4，将老年人群分层为听力损失低危（<4）和听力损失高危（≥4）。基于微信平台开发老年听力损失筛查软件，共包含危险因素分层评估、老年听力障碍量表筛查版（HHIE-s）评估、总体结论、健康宣教4个板块。2020年7—12月，共招募78位老年人参与老年听力损失筛查软件测试，完成配合率为55.1%（43/78）。危险因素累计得分与HHIE-s得分呈正相关（r_s=0.360，P=0.018）。2021年1—3月，共106位全科医生参与老年听力损失筛查软件适宜性评价问卷调查，包含8个单项选题。采用5级李克特量表编制，各题选择完全同意的比例分别为85.8%（91/106）、81.1%（86/106）、71.7%（76/106）、68.9%（73/106）、68.0%（72/106）、59.4%（63/106）、15.1%（16/106）、14.2%（15/106）。

与老年听力损失有关的危险因素共18项，基于微信平台的老年听力损失筛查软件预警老年听力损失的效果较好，为构建老年听力损失筛查提供了循证医学依据。老年听力损失筛查软件能实现数据实时传输，将危险因素分层评估体系、HHIE-s和微信平台进行优化融合，使得听力损失的自我健康管理落到实处。老年听力损失筛查软件为基层全科医生提供了听力损失管控新方式，具有适宜性和执行度。