Musculoskeletal disorders continue to be a leading source of lost workdays across all industries. Common ergonomics assessment tools may include criteria extraneous to the stresses at specific companies or industries. Therefore, the creation of assessment tools, based on scientifically validated methods, with industry- or company-specific stresses may be of benefit. Privacy regulations (among other factors) may preclude the 1:1 allocation of stresses to health outcomes, prohibiting conventional epidemiological investigation and validation approaches, therefore requiring the use of aggregated data sources.
 Based on data from the ergonomic assessment tool of interest (Safety and Ergonomics Risk Assessment - SERA) and aggregated data derived from internal insurance reports a statistical method is developed to investigate; 1) the validity of the tool, 2) a prioritization of intervention targets, and 3) the derivation of threshold values for monitoring. The method involves statistical tests used in epidemiological investigations and draws on the iterative modelling approach common to machine learning.
 Using the described approach, significant results for known musculoskeletal issues, a prioritization of countermeasure and intervention targets, and threshold values for long-term monitoring were determined. The successful results indicate that the assessment tool investigated is valid. Consequently, the method is generalized such that practitioners can apply it, should they be faced with similarly structured data.