背景：

我国约有2亿人群接触职业性危害因素，国家高度重视职业人群健康，因此，职业健康保护行动是《健康中国行动（2019—2030年）》的15大行动之一。船舶制造业作为典型的高噪声暴露行业，其工人面临较高的听力损失风险。近年来，国家出台了一系列政策和措施，明确要求“治理噪声超标”、“预防和控制重点职业病发生”。在此背景下，准确估算职业性噪声所致听力损失（Occupational Noise-Induced Hearing Loss, ONIHL）对保护与促进我国噪声作业工人的职业健康具有重要意义。然而，当前职业性噪声归因听力损失的研究存在基础数据不足、方法学研究不充分和研究数量有限等问题，需进一步在船舶制造业等高噪声行业中探索如何深入开展归因方法学研究。本研究旨在采用两种估计风险值的方法，结合两种对照场景，分别估算2020年我国船舶制造业接触噪声工人中职业性噪声归因听力损失的比例（Attributable Fraction, AF）和人数（Attributable Case, AC），以摸清该行业ONIHL现状，并比较和探讨不同暴露风险度量方法以及不同对照场景下开展职业性噪声归因研究的区别。

方法：

1.数据来源于经济普查、专项调查、监测系统和综述数据。包括2018年的第四次全国经济普查、2017-2019年的“船舶制造业现况调查”、2020年的“全国职业病危害现状调查”、2014年-2015年的“我国四省听力障碍流行现况调查”、2020年的“职业病及健康危害因素监测信息系统”以及系统综述到的公开发表文献。

2.描述我国2020年船舶制造业工人的噪声接触情况和听力损失情况。全国船舶制造业从业人数来自于“全国经济普查”，该行业的接噪率来自于“全国职业病危害现状调查”，结合二者，进而获得全国该行业的接噪人数。进一步借用“职业病及健康危害因素监测信息系统”中船舶制造业工人的年龄构成比，获得以上相关指标的分年龄别结果。再结合由“船舶制造业现况调查”获得的分年龄别听力损失患病率，最终估算出2020年全国船舶制造业接噪工人中分年龄别听力损失人数。

3.基于风险值采用二分类度量，且以普通人群为对照时（场景1），估算AF和AC：系统检索公开发表文献，尝试合并二分类风险值，即RR（Relative Risk, RR）。当系统综述无法获得该值时，利用关键的中国人群研究，通过“我国四省听力障碍流行现况调查”和“船舶制造业现况调查”分别获得普通人群和接噪工人的听力损失患病率，从而估算患病率比（Prevalence Ratio, PR）来代替RR。再基于二分类度量的AF计算公式，利用PR获得AF，并将其与接噪工人中听力损失人数结合，从而估算AC。

4.基于风险值采用剂量-反应关系度量，且以累积噪声暴露量（Cumulative Noise Exposure, CNE）的理论最小暴露水平（Theoretical Minimum-Risk Exposure Level, TMREL）为对照时（场景2），估算AF和AC：首先，仍采用系统综述方式，尝试合并剂量-反应关系风险值。当系统综述无法获得该值时，基于“船舶制造业现况调查”数据，构建Logistic回归模型，估算CNE的剂量-反应关系比值比（Odds Risk, OR）来代替RR。其次，使用“船舶制造业现况调查”数据，描述接噪工人的CNE分布情况，并拟合CNE的概率密度函数。最后，利用CNE的概率密度函数、剂量-反应关系OR和相关参数估算AF。AC计算同场景1。

结果：

1.2020年我国船舶制造业有416,461人，接噪率为28.22%，共有接噪工人117,535人，主要集中于30-49岁。接噪工人的听力损失患病率为49.24%，且患病率随年龄增长呈现上升趋势，由18-29岁的24.95%大幅上涨到50-59岁的74.89%；其中共有60,523人患有听力损失，主要分布在30-49岁。

2.基于风险值采用二分类度量，且以普通人群为对照时（场景1）的AF和AC：二分类风险值来源于典型调查。结果显示，PR值随年龄增加呈现明显下降趋势，由18-29岁的12.73大幅降至50-59岁的2.83；相应的AF值从18-29岁的92.14%下降至50-59岁的64.66%，合计AF为80.80%。合计AC为48,905人，呈现先上升再下降的趋势，主要集中在30-49岁。

3.基于风险值采用剂量-反应关系度量，以CNE的TMREL为对照时（场景2）的AF和AC：剂量-反应关系风险值同样来源于典型调查。Logistic回归分析结果表明，CNE每增加1dB(A)·年，听力损失的OR增加4.7%（OR：1.047；95%CI：1.036-1.057）。随着年龄推移，CNE的中位数逐渐增加，整体分布呈现上升趋势。AF随年龄的增长而逐渐升高，由18-29岁的49.37%增长至50-59岁的61.91%，合计AF为57.59%。合计AC为34,856人，亦呈现先上升再下降的走向，主要集中在30-49岁。

结论：

职业性噪声对噪声作业工人的听力健康造成了极大危害，船舶制造业接噪工人中，约80%（近五万）的听力损失可归因于职业性噪声。且随着暴露浓度和时长的加深，ONIHL所占的比例不断加大。听力损失工人主要集中在30-49岁年龄段，提示我国应加强针对该人群的职业性噪声防控力度以减少中国工人的健康损失。同时，在基础数据薄弱的情况下，本研究充分整合现有中国人群数据资源，提供了两种风险值估计的方法，以及两种对照场景下归因疾病负担的估算，为其它职业病和其危险因素归因疾病负担的估算提供了方法学参考。

Background:

Approximately 200 million people in our country are exposed to occupational hazards. The state places high importance on the health of the working population. Therefore, occupational health protection actions are one of the 15 major initiatives of the Healthy China Action (2019-2030). The shipbuilding industry, as a typical high-noise exposure sector, poses a significant risk of hearing loss for its workers. In recent years, the government has introduced a series of policies and measures that explicitly call for “Control noise pollution” and “Prevent and control the occurrence of key occupational diseases.” Against this backdrop, accurately estimating Occupational Noise-Induced Hearing Loss (ONIHL) is of great importance for the protection and promotion of the occupational health of workers exposed to noise in China. However, current research on occupational noise attributable to hearing loss is plagued by a lack of foundational data, insufficient methodological research, and a limited number of studies. There is a need for in-depth attributable research in high-noise exposure industries such as shipbuilding. Using two methods of exposure risk estimation and two scenarios of control populations, this study aimed to estimate the attributable fraction (AF) and the attributable case (AC) among workers exposed to noise in China's shipbuilding industry in 2020, in order to clarify the current situation of ONIHL in shipbuilding, and to compare and discuss differences in occupational noise attribution research conducted under different risk measurement methods and different control scenarios.

Methods:

Data were obtained from economic censuses, special surveys, monitoring systems and review data. Including the “Fourth National Economic Census 2018”, the “Shipbuilding Industry Cross-Sectional Survey (2017-2019)”, the “National Occupational Hazards Survey 2020”, the “Prevalence of Hearing Disorders in China (2014-2015)”, the “Occupational Disease and Hazard Monitoring System 2020”, and the published literature by systematic review.

Described the situation of noise exposure and hearing loss among workers in China's shipbuilding industry in 2020, respectively. The number of workers in the national shipbuilding industry came from the “National Economic Census 2018”, and the noise exposure rate in this industry was derived from the “National Occupational Hazards Survey 2020”, thus the number of noise-exposed workers nationwide in the industry was estimated. Further, by borrowing the age composition proportion of workers in the shipbuilding industry from the “Occupational Disease and Hazard Monitoring System 2020”, we can obtain the age-specific results for the aforementioned indicators. Combined the number of noise-exposed workers nationwide with the age-specific prevalence of hearing loss from the “Shipbuilding Industry Cross-Sectional Survey (2017-2019)”, the age-specific number of workers with hearing loss in shipbuilding industry in 2020 was estimated.

Using dichotomous measurement based on the exposure risk, the AF and the AC were estimated when the general population was a control (Scenario 1). By systematically searching published literature, try to combine the binary risk values. When the systematic review could not obtain the pooled risk, key Chinese population studies were utilized. By comparing the prevalence of hearing loss among general population from the “Prevalence of Hearing Disorders in China” and the prevalence of hearing loss among noise-exposed workers from the “Shipbuilding Industry Cross-Sectional Survey (2017-2019)”, the prevalence ratio (PR) as a substitute for the relative risk (RR) was estimated. And then based on the method for dichotomous risk factor, the AF was calculated by PR. AF was multiplied by the number of noise-exposed workers with hearing loss to obtain the AC.

Using continuous measurement based on the exposure risk, the AF and the AC were estimated when the theoretical minimum-risk exposure level (TMREL) of cumulative noise exposure (CNE) was a control (Scenario 2). First, by systematically searching published literature, try to combine the dose-response risk. When the systematic review could not obtain the pooled risk, based on the “Shipbuilding Industry Cross-Sectional Survey”, the Logistic Regression Model was used to calculate dose-response odds ratio (OR) as a substitute for the RR. Next, by using the data from the “Shipbuilding Industry Cross-Sectional Survey (2017-2019)”, described the distribution of CNE among noise-exposed workers, as well as fit the probability density function of CNE. Finally, the probability density function of CNE, dose-response OR and related parameters were used to estimate AF. AC was calculated as in Scenario 1.

Results:

In 2020, there were 416,461 people in China's shipbuilding industry, with a total noise exposure rate of 28.22%, resulting in 117,535 workers exposed to noise, mainly concentrated in the age group of 30 to 49 years old. The total prevalence of hearing loss among noise-exposed workers was 49.24%, and the prevalence showed an increasing trend with age, which increased from 24.95% in the 18-29 age group to 74.89% in the 50-59 age group. A total of 60,523 workers suffered from hearing loss, mainly between the ages of 30 and 49 years.

Using dichotomous measurement based on the exposure risk, the AF and the AC were as follows when the general population was a control (Scenario 1). Binary risk values derived from key single studies. The results showed that PR decreased significantly with age, from 12.73 in the 18-29 age group to 2.83 in the 50-59 age group. Accordingly, AF decreased from 92.14% in the 18-29 age group to 64.66% in the 50-59 age group, and the total AF was 80.80%. The total AC was 48,905, presenting a trend of first rising and then declining, mainly concentrated in the 30 to 49 years.

Using continuous measurement based on the exposure risk, the AF and the AC were as follows when the TMREL of CNE was a control (Scenario 2). The dose-response risk also came from key single studies. Logistic regression analysis showed that CNE increased by 1dB(A)·year, the OR of hearing loss increased by 4.7% (OR: 1.047; 95%CI: 1.036-1.057). The median of CNE increased gradually with age, and the overall distribution of CNE presented an upward trend. The AF gradually increased with age, rising from 49.37% in the 18-29 age group to 61.91% in the 50-59 age group, with a total AF of 57.59%. The total AC was 34,856, showing a trend of first rising and then declining, most of the cases between the ages of 30 to 49 years.

Conclusion:

Occupational noise has caused great harm to the hearing health of noise-exposed workers. About 80% (nearly 50,000) of the hearing loss of noise-exposed workers in the shipbuilding industry can be attributed to occupational noise. With the deepening of exposure concentration and duration, the proportion of ONIHL continues to increase. Workers with hearing loss are mainly concentrated in the age group of 30-49 years, indicating that China should strengthen the prevention and control of occupational noise in this population to reduce the health loss of Chinese workers. At the same time, in the case of weak basic data, this study fully integrated existing Chinese data resources, as well as provided two methods for estimating exposure risk and the estimation of attributable disease burden under two control scenarios. It will offer a methodological reference for the estimation of attributable disease burden for other occupational diseases and their risk factors.