[Validation of the ACGIH hand activity level (HAL) model among automotive manufacturing workers in China]

Objective: To validate the applicability of the ACGIH hand activity level (HAL) model for wrist risk assessment among manufacturing workers in China, and to provide scientific evidence for optimizing occupational health risk assessment methods. Methods: From August 1 to October 1, 2024, a study was conducted using cluster random sampling combined with stratified sampling to select 66 male workers from the final assembly workshop of an automobile manufacturing enterprise as research subjects. Occupational health surveys and surface electromyography (sEMG) measurements were employed to obtain HAL and normalized peak force (NPF) data. Using the internationally recognized occupational repetitive actions (OCRA) index as the reference standard, methods such as Spearman rank correlation, weighted Kappa test, ordered logistic regression, and receiver operating characteristic (ROC) curve analysis were applied to evaluate the correlation, consistency, goodness of fit, and discriminative power of the two indicators. Results: The median age of the study subjects was 30.00 (29.00, 32.00) years old, and the median length of service was 10.00 (9.00, 10.00) years. The reported incidence of wrist pain or discomfort within the past year was 28.8% (19/66), and the incidence of hand injury was 24.2% (16/66). The HAL, NPF, and OCRA indices all showed significant positive correlations (r(s)=0.92, 0.87, P<0.001). The risk stratification method established in this study exhibited a weighted Kappa value of 0.86 (95% CI: 0.77-0.96, P<0.001) compared to the OCRA risk stratification, indicating good consistency. Multivariate ordered logistic regression analysis revealed that after controlling for confounding factors such as age, length of service, height, and body mass index, both HAL and NPF were independently positively correlated with OCRA risk levels, with adjusted OR values of 27.52 (95%CI: 3.89-195.20, P<0.001) and 18.69 (95%CI: 1.62-215.94, P=0.019), respectively. The model Nagelkerke R(2)=0.91, demonstrating good fit. ROC analysis revealed that the area under the curve (AUC) for high-risk OCRA in HAL, NPF, and their combined model was 0.94, 0.91, and 0.95, respectively. The sensitivity corresponding to the optimal cutoff values was 88.9%, 100.0%, and 94.4%, while specificity reached 97.9%, 77.1%, and 93.7%, demonstrating excellent discriminative performance. Conclusion: HAL and NPF may serve as effective quantitative indicators for repetitive or forceful wrist work risks among automotive manufacturing workers. Their risk classification method shows good consistency with OCRA and is applicable for on-site rapid screening and risk assessment of occupational carpal tunnel syndrome.