Predicting the Effective Factors on Musculoskeletal Disorders among Kerman University of Medical Sciences Computer Users through Neural Network Algorithm in 2018

Introduction: In the past 20 years, computers and their workplaces have increased at both offices and houses, which consequently has led to saving in time, energy and resources. This study aimed to weight risk factors of musculoskeletal disorders among computer users through neural network. Methods: A cross-sectional study was carried out at 200 stations in Kerman University of Medical Sciences. Firstly, the factors affecting musculoskeletal disorders through ROSA were determined, and then the score for each of them was determined. Then, the final score of user's musculoskeletal disorders was determined, and after pre-processing, the prediction of the effect of factors was obtained using neural network. Data was analyzed using IBM SPSS Modeler 18.0. Results: The average of final score of ROSA, chair, telephone-monitor and mouse-keyboard were 4.36 ± 0.91, 3.67 ± 1.06, 3.68 ± 1.09 And 3.66 ± 1.18 respectively. 131 Workstation (65.5%) had a score less than 5 & 69 Workstation (34.5%) had a score equal to or greater than 5. Based on neural network algorithm Chair factor with a normalized weighting 41%; telephone-monitor factor with a normalized weighting 31% and finally mouse-keyboard factor with a weighting factor 28% were respectively effective factors on disorders caused by working with computers. Conclusions: The most normalized weight is for chair, and then the telephone-monitor and mouse-keyboard. We should include ergonomic interventions considering the effect of each factor (normalized weighting of factors) provided by neural network to decrease such disorders.