Prediction Of Musculoskeletal Disorders Cases Associated With Work-Relatedness Of Semiconductor Workers Using Logistic Regression Model

• Main Author: Chai, Fong Ling
• Format: Thesis
• Language: English; English
• Published: 2020
• Subjects: Q Science (General); QA Mathematics
• Online Access: http://eprints.utem.edu.my/id/eprint/25410/1/Prediction%20Of%20Musculoskeletal%20Disorders%20Cases%20Associated%20With%20Work-Relatedness%20Of%20Semiconductor%20Workers%20Using%20Logistic%20Regression%20Model.pdf; http://eprints.utem.edu.my/id/eprint/25410/2/Prediction%20Of%20Musculoskeletal%20Disorders%20Cases%20Associated%20With%20Work-Relatedness%20Of%20Semiconductor%20Workers%20Using%20Logistic%20Regression%20Model.pdf; http://eprints.utem.edu.my/id/eprint/25410/; https://plh.utem.edu.my/cgi-bin/koha/opac-detail.pl?biblionumber=119727

Semiconductor industry is one of the main contributors to the economy of Malaysia. Semiconductor workers are generally exposed to various conditions at work that is likely to contribute to the development of musculoskeletal disorders (MSDs).Currently, there has been limited studies conducted to identify potential MSDs risk factors among semiconductor workers. Besides, there have been little studies categorize semiconductor workers’ exposures of risk factors, and there has been a paucity of predictive model developed to predict the work-relatedness of MSDs. This study was conducted to (i) identify the potential risk factors of MSDs among semiconductor workers; (ii) construct a model in predicting work-relatedness of MSDs diagnoses based on the identified risk factors and investigate the relationship between the identified risk factors and work-related MSDs cases; (iii) validate the developed predictive model and the relationship findings statistically and through face validation by experienced experts in the ergonomic field. Risk factors from the literature searches and 277 work assessment reports of workers diagnosed with MSDs were sorted and compared. A total of 16 predictors were identified from ergonomic risk factors (ERFs), work activities, confounding factors of MSDs, and duration of employment for workers reporting the first musculoskeletal symptoms (MSS). Kurskal-Wallis one-way analysis of variance testand Kendall’s tau correlation were conducted to test the significance difference of the risk factors and analyse correlation between identified risk factors and MSDs outcomes. The specific factors identified to have significant effect in predicting work-relatedness of MSDs in the developed model are ERFs such as poor posture, forceful exertion, and static posture and loading; work activities such as lifting and lowering, transferring, pushing and pulling, repairing, preventive maintenance and quality inspection; confounding factors such as age and previous injury history; and the duration of employment in reporting first MSS. Crossvalidation of the developed predictive model was conducted with a new set of test data (n=30), and the accuracy of the prediction model was measured at 86.20%. Thirty experts in the ergonomics field gave promising 80% average rating agreements on the inclusion of identified risk factors and major result findings. Improvements in future research study suggest the inclusion of psychological and environmental factors in work assessments for a more comprehensive prediction process. These outcomes may help practitioners to understand the exposure components contributing to work-relatedness of MSDs cases among semiconductor workers. Ergonomists, safety and health officers, engineers and management employees can utilize this predictive model to predict the potential risk of MSDs cases and guide the process of determining appropriate control measures and future interventions