? ? The CDC has identified musculoskeletal disorders (MSDs) as a Research Theme within the Health Protection goal of Healthy People in Healthy Places. After decades of epidemiologic research, consensus is emerging that 1) forceful exertions, 2) awkward postures, and 3) repetitive motions of the hand and wrist are risk factors for upper extremity MSDs [3, 4]. Recent efforts have been made to pool these three risk factors into a single metric or scale to estimate UEMSD risk. The Strain Index (SI) is a highly reliable, widely used tool that was developed for this purpose [5-13]. The SI outcome value is an ordinal numerical score. Higher SI scores are associated with greater UEMSD health outcomes, although no published studies have compared SI scores to incident UEMSDs. Despite its wide acceptance, there is also considerable evidence that the current SI cut-off scores overestimate risk. Our expectations are that more appropriate risk categories can be established. The long-term of the investigators is to prevent UEMSDs among manufacturing workers by implementing ergonomic interventions. The main rationale for conducting this study is to better assess the SI as an exposure assessment tool by using more powerful epidemiological methods and more precise alternate estimates of exposure to physical risk factors.
The specific aims of this study are: 1) For UEMSD outcomes among manufacturing workers, separately model a) associations between the SI using the original job risk categories and incident UEMSD outcomes and b) associations between the SI using the empirically derived quartile job risk categories and incident UEMSD outcomes; and 2) Compare the adequacy of fit of a) models that regress UEMSD outcomes on exposure estimated with the SI to b) models that regress UEMSD outcomes on exposure estimated with separate highly reliable and precise measures of force, repetition and posture. This project requires additional analyses of existing video exposure and health outcome data from a three year prospective cohort study of manufacturing workers in Iowa (n=387). All SI evaluations will be conducted independently by two investigators and then the final SI score will be determined by consensus. Multinomial logistic regression will be used to test both specific aims.
For Specific Aim 1, incidence data for upper extremity musculoskeletal outcomes (symptom free, symptom positive, or disorder positive) in the dominant arm will be regressed on exposure estimated by 1) the original SI job risk categories, and 2) the proposed empirically derived SI risk categories.
For specific aim 2, the Akaike Information Criterion (AIC) will be used to test the adequacy of fit of the aforementioned multinomial logit model will be compared to an alternate multinomial logit model that will estimate exposure using separate measures of force (surface electromyography), repetition (Hand Activity Level), and posture (% time spent in awkward wrist postures). ? ? ?
| Meyers, Alysha R; Gerr, Fredric; Fethke, Nathan B (2014) Evaluation of alternate category structures for the Strain Index: an empirical analysis. Hum Factors 56:131-42 |
