Background Noise is a serious issue in the workplace. Scientific assessment and risk management of noise-exposed workers is one of the important ways to reduce the risk of hearing loss.

Objective To evaluate hearing loss of noise-exposed workers at a plastic product enterprise based on the noise risk assessment model (Acoustics—Estimation of noise-induced hearing loss, GB/T 14366—2017/ISO 1999: 2013) modified using multiple population databases.

Methods The working environment of a plastic product enterprise and its 308 noise-exposed workers were selected as study subjects. Workplace noise sound pressure level and individual noise exposure level were measured respectively, a questionnaire survey and pure tone audiometry test were conducted for each noise-exposed worker, and unadjusted binaural high frequency threshold average (UNBHFTA) and age-corrected binaural high frequency threshold average (BHFTA) were calculated. Combined with personal information and in accordance with the Guidelines for risk management of occupational noise hazard (WS/T 754-2016) risk assessment and risk classification of the hearing threshold shift of noise-exposed workers were carried out. The ISO 1999: 2013 model was used and modified based on the four population databases recommended by the same standard. The average high frequency hearing threshold of each noise-exposed worker was predicted respectively, and compared with the UNBHFTA results.

Results The sound pressure level M (P₀, P₁₀₀) of four noise-exposed workstations in this enterprise was 87.9 (82.1, 92.9) dB(A), and the sound pressure level of 45 noise sources was 87.3 (70.2, 117.3) dB(A), mainly including extruders, packaging machines, vacuum conveyor and mixing machines. The age M (P₀, P₁₀₀) of 308 noise-exposed workers in the enterprise was 34.0 (25.0, 57.0) years, and the working experience M (P₀, P₁₀₀) was 8.0 (0.5, 25.5) years. Six workers reported BHFTA>25 dB, and 269 reported BHFTA>10 dB. According to WS/T 754-2016, 197 workers were at very high-risk level due to high frequency threshold shift of either ear drifting by 10 dB, and 33 workers were at high-risk level. In addition, there were significant statistical differences in the noise risk assessment results of each workstation across the four databases (P<0.001). The high frequency average hearing threshold estimates of 308 noise-exposed workers based on databases A, B.1, and B.2 were significantly different from the actual pure tone audiometry results (P<0.01), while the predicted results based on database B.3 were not statistically significant different from the UNBHFTA (P>0.05).

Conclusion The employees exposed to noise in this enterprise are at a high risk of hearing loss. The noise risk prediction and assessment model adopted by GB/T 14366-2017/ISO 1999: 2013 may report varied results by using different databases, and a population database should be selected for model modification according to the characteristics of the population. The assessment model based on database B.3 (unscreened and unremoved occupational noise exposure history) has great potential to be used for noise risk prediction, assessment, and management of the same type of workers.