Construction and application of an occupational health risk grading method based on risk matrix and dynamic weighting: a case study of footwear manufacturing enterprises.

BACKGROUND: Occupational hazard factors in the footwear manufacturing industry are complex and encompass benzene series compounds, ketones, esters, dust, noise, and other agents. Existing mainstream occupational health risk assessment methods-including the Singapore semi-quantitative risk assessment method (Ministry of Manpower method, hereafter referred to as the MOM method), the exposure ratio method specified in GBZ/T 298-2017 (hereafter referred to as the exposure ratio method), the United States Environmental Protection Agency (EPA) inhalation risk assessment method (hereafter referred to as the EPA method), and the International Council on Mining and Metals (ICMM) occupational health risk assessment model (hereafter referred to as the ICMM method)-each have specific applicable scenarios and limitations. No single method can achieve precise risk grading and dynamic management. OBJECTIVE: To construct a novel occupational health risk grading method that integrates the strengths of four mainstream methods, based on a risk matrix and dynamic weighting, and to validate its applicability using footwear manufacturing enterprises as a case study. METHODS: Four footwear manufacturing enterprises (A, B, C, D) were selected, and on-site measurements of chemical toxicants, dust, and noise in the workplace were conducted. Risk assessments were performed using the MOM method, the exposure ratio method, the EPA method, and the ICMM method. A risk ratio normalization approach was applied to align the risk levels from all methods into a unified five-level system. On this basis, a risk matrix was constructed, and a combined weighting approach integrating the entropy weight method and the analytic hierarchy process (AHP) was employed to determine the dynamic weights of each method under different hazard scenarios. A comprehensive risk index (CRI) was constructed and graded. Consistency among methods was analyzed using Spearman's rank correlation, weighted Kappa coefficient, Fleiss' Kappa coefficient, and Wilcoxon signed-rank test. RESULTS: Maximum benzene concentrations reached 3.41 mg/m3 (Enterprise B) and 5.7 mg/m3 (Enterprise C, exceeding the limit); methyl ethyl ketone peaked at 175.0 mg/m3 (Enterprise D). The conventional methods showed moderate agreement (Fleiss' Kappa = 0.455). CRI demonstrated substantial agreement with each method (κ_w = 0.625-0.796) and significantly moderated ICMM overestimation (Z = -2.873108, p = 0.004). CRI classified Enterprise C gluing as Level V (CRI = 1.00) and Enterprise D benzene as Level II (CRI = 0.28). Dust yielded a maximum CRI of 0.42 (medium risk). Noise CRI reached 0.60 (medium risk, Level III). Sensitivity analysis confirmed robustness of α, β, γ coefficients. CONCLUSION: The constructed grading and classification method can more objectively and comprehensively reflect the actual occupational health risks in footwear manufacturing enterprises, providing a scientific basis for formulating differentiated and precise management strategies.