On the Relationship of Musculoskeletal Disorder Compensation Claims to Ergonomic Factors in Manufacturing

“Surveillance is the collection, analysis, and dissemination of results for the purpose of prevention.” (Halperin, 1996). Identification of high rates of work-related morbidity or mortality serves this purpose when the health data lead back to preventable exposures. Timeliness is also essential, in order that the information can be used while it is still relevant and actionable. Workers’ compensation (WC) claims for workplace injury and illness represent one commonly used type of data for occupational health surveillance, even though these data exist for an entirely different purpose. Important advantages of using WC paid claims include that the employer and insurance company have already accepted that the incident is work-related; and that they do not require new, expensive data collection or direct access to the workplace. However, the chain of events from occupational exposure to onset of work-related morbidity to claim-filing leads through many filters (Azaroff et al., 2002; Spieler and Burton, 2012). These involve attitudes and behaviors at multiple levels, from employee decision to file a claim, to employer acceptance of the claim, to insurance company approval of work-relatedness. As a result, many injuries and disorders do not result in compensation claims and/or payments, even after they have been judged to be work-related by an employer representative (Morse et al., 2005; Boden et al., 2008). From the perspective of priority-setting for preventive efforts, under-reporting would not introduce a fatal flaw if the events that are reported represent a consistently proportional sample of those actually occurring, by the type of outcome as well as by risk factors. However, there is reason to believe that the probability of claim-filing is not generally consistent even for clinically similar events. Multiple investigations have shown that non-filing varies by demographic characteristics, employment sector and arrangement, working conditions, worker knowledge and self-efficacy, and other factors (e.g. Fan et al., 2006; Keegel et al., 2009; Groenewold and Baron, 2013; Foley et al., 2014; Qin et al., 2014). This point may be seen as distinguishing the use of WC datasets for surveillance versus computing relative risks for specific exposures. In other words, high WC claim rates may reasonably be taken as indicative of high-risk job settings, perhaps even underestimating the incidence in those jobs. However, it is less certain that low rates demonstrate actual low risk, because of the many obstacles to claim-filing and their differential impact, as noted above. If low-rate and high-rate jobs actually have more similar incidences of work-related musculoskeletal disorders (MSDs) than would appear from the claim rates, they might also have similar exposure profiles. In such a scenario we might find few differences in exposure between the two groups, not because the exposures are causally unimportant but because of nonrandom misclassification of outcome. Bao and colleagues at the Washington State program, Safety & Health Assessment & Research for Prevention (SHARP), observed that almost one-half of compensable claims and of costs were due to work-related MSDs, and that the largest number of claims arose from the manufacturing sector. Therefore, they sought to identify company-level factors which might explain those differences in claim rates (Bao et al., 2020). Few WC systems are structured so as to provide data on specific exposures sustained by the injured workers. Even if they did there would typically be no comparable data on exposures to non-injured workers, so relative risks by exposure are not readily estimated without separate information such as that available through a job-exposure matrix. Thus, to generate a basis for comparison, the authors opted to characterize ergonomic features of jobs in companies with high and low WC claim rates. They used 10 years of WC data for those rates, to obtain stable categorizations; they selected observational job evaluation methods that are well-known and within the scope of occupational ergonomics practice, to replicate what could be feasibly be expected of workplace safety and health professionals or trained union representatives. The design of this study, with a sample of jobs standing in for company-level risk, is ecological. This means that both risk factors and outcomes are quantified at the group level and exposure–response relationships are examined between the aggregated metrics. The ecological design has been criticized for potential confounding (because other possible causal factors are unknown for the individuals) and the ‘ecological fallacy’, which refers to using group-level data to make inferences on individual-level relationships between determinants and diseases (because the affected persons within a group may not be the same as the exposed persons). On the other hand, when the determinant is a group-level phenomenon, the ecological study is consistent with the posited mechanism (Schwartz, 1994). For example, if union density within a jurisdiction (rather than individual union membership) determines union strength in negotiating wages and conditions of work, then an ecological study of its effect on cause-specific mortality by geographical units (Eisenberg-Guyot et al., 2019) is a reasonable approach. Similarly, it is appropriate for evaluation of large-scale change in occupational safety policy or practice, such as efforts to reduce falls in construction (Lipscomb et al., 2014). An ecological study of specific ergonomic exposures that are sustained by some people within a workplace but not others would have limited ability to identify individuals’ MSD risk attributable to those exposures. Interindividual differences in work methods (within company) could easily bias the results toward the null hypothesis. A related concern in this study is that the jobs observed were not necessarily those from which the reported injuries arose. Conversely, company-level factors are appropriately studied with this design, to ascertain whether they have sufficient impact to influence the total claim rate within the enterprise. Thus it was highly commendable of Bao and colleagues to complement the observations of individual jobs with open-ended interviews of company and worker representatives to learn about relevant aspects of safety culture and programs. It is notable that the qualitative interview items highlighted a contrast in reliance on administrative versus engineering controls in high- versus low-risk companies. It would have been very useful to see those results in more detail, to understand which specific features of the broader work environment and culture might be associated with injury occurrence and/or support for injury reporting. Unfortunately, this study had limited statistical power. Regardless of the number of people employed by the companies studied, the use of company-level analyses necessarily relied on a ‘n’ of 16 companies to examine specific ergonomic exposures (from observations) and 32 for the company policies (interview results). Thus it is unfortunate that only differences featuring a P-value of 0.05 were considered meaningful. Several useful lessons can be derived from this article. The authors have documented that MSDs continue to represent a large proportion of work-related morbidity; that MSDs do not occur at random with respect to preventable ergonomic exposures; and that WC claims can be used to identify jobs and workplaces with excess morbidity. While some occupational risk factors may have been missed by this study, others were identified using simple observational methods that are well-known to workplace ergonomics practitioners. These methods can and should be readily employed as screening techniques. In particular, jobs that require prolonged standing, heavy lifting, high repetition and work pace, pinch forces, and job stress should be redesigned to reduce the magnitude of this epidemic among employed individuals, in the manufacturing sector and elsewhere. Last, there is evidence here that good organizational policy and practice in ergonomics may reduce MSD risk. The methods used here could be incorporated into routinely scheduled walk-throughs to identify high-risk jobs and specific exposures to be reduced. The fact that low-claim companies were more likely to use engineering controls, in contrast to administrative controls at high-claim companies, is particularly notable and should serve to strengthen the emphasis on primary prevention among practicing ergonomists. The author declares no conflict of interest relating to the material presented in this editorial. Its contents, including any opinions and/or conclusions expressed, are solely those of the author.