Serafim Rômulo Silva, Bispo Lucas Gomes Miranda, da Silva Jonhatan Magno Norte, da Silva Joel Gomes
Federal University of Alagoas - Campus do Sertão - Delmiro Gouveia, Alagoas, Brazil.
Postgraduate Program in Production Engineering, Federal University of Rio Grande do Sul, Rio Grande do Sul, Brazil.
Work. 2024;78(1):119-130. doi: 10.3233/WOR-230480.
Work-related musculoskeletal disorders (WMSD) encompass a range of conditions affecting muscles, tendons, and nerves. Visual diagrams are widely used to identify symptoms and to generate musculoskeletal discomfort metrics. However, there is no consensus on the number of discomfort dimensions that can originate from self-reported musculoskeletal symptoms by individuals.
This study aimed to test the fit of WMSD symptom models from workers in two samples of different sizes.
A combination of Full-Information Item Factor Analysis (FIFA) and Item Response Theory (IRT) was utilized to analyze and test the models. The study was conducted in two samples of workers (n1 = 6944 and n2 = 420) who had their symptoms identified with the aid of a human body diagram. An analysis was conducted considering each sample's unidimensional and three multidimensional models.
The unidimensional model (general musculoskeletal discomfort), bi-dimensional model (discomfort in upper and lower body), and tridimensional model (discomfort in the upper limbs, lower limbs, and trunk) showed good values of factor loading and communalities, along with satisfactory item discrimination ability. Regardless of sample size, parameter estimation for IRT and FIFA proceeded without issues, presenting suitable fit parameters.
Three models were valid and reliable for more extensive and smaller samples. However, the tridimensional model was best for generating discomfort scores in body regions. Companies and safety professionals can use these findings to devise strategies to mitigate musculoskeletal pains based on perceived symptom locations.
与工作相关的肌肉骨骼疾病（WMSD）包括一系列影响肌肉、肌腱和神经的病症。视觉图表被广泛用于识别症状并生成肌肉骨骼不适指标。然而，对于个体自我报告的肌肉骨骼症状可能产生的不适维度数量，尚无共识。
本研究旨在检验来自不同规模两个样本的工人的WMSD症状模型的拟合度。
采用全信息项目因子分析（FIFA）和项目反应理论（IRT）相结合的方法来分析和检验模型。该研究在两个工人样本（n1 = 6944和n2 = 420）中进行，这些工人借助人体图表识别了他们的症状。针对每个样本的一维和三个多维模型进行了分析。
一维模型（一般肌肉骨骼不适）、二维模型（上半身和下半身不适）和三维模型（上肢、下肢和躯干不适）显示出良好的因子载荷和共同度值，以及令人满意的项目区分能力。无论样本大小，IRT和FIFA的参数估计都没有问题，呈现出合适的拟合参数。
对于更大规模和较小规模的样本，三种模型都是有效且可靠的。然而，三维模型最适合生成身体区域的不适评分。公司和安全专业人员可以利用这些发现来制定基于感知症状位置减轻肌肉骨骼疼痛的策略。
