Related Experiment Video
Updated: Aug 26, 2026

Inverse Probability of Treatment Weighting (Propensity Score) using the Military Health System Data Repository and National Death Index
Published on: January 8, 2020
Methodologic issues in the use of workers' compensation databases for the study of work injuries with days away from
Arthur Oleinick1, Brian Zaidman
1School of Public Health, University of Michigan, Ann Arbor, MI 48109, USA. aoleinic@umich.edu
Background:
Case ascertainment costs vary substantially between primary and secondary data sources. This review summarizes information on the sensitivity of state administrative databases in workers' compensation systems for the ascertainment of days-away-from-work (DAFW) work injuries for use in modeling studies.
Methods:
Review of the literature supplemented by data from governmental or organizational reports or produced for this report.
Results:
Employers currently appear to provide workers' compensation insurance coverage for 98.9% of wage and salary workers. Wage and salary jobs account for approximately 90% of jobs in the United States. In industries such as manufacturing, the fraction of covered jobs is probably closer to 98%. In Minnesota, the number of DAFW cases ascertained by the Bureau of Labor Statistics' annual survey of occupational injuries and illnesses is approximately 92-97% concordant with the number of wage compensation claims for injuries producing DAFW over the period 1992-2000, once adjustments are made to permit direct comparisons of the numbers. The workers' compensation databases provide information for more than 95% of the total DAFW resulting from work injuries. Covariate estimates are unaffected by this less than 5% loss because effects appear dependent on time from injury.
Conclusions:
Statewide workers' compensation administrative databases can have substantial utility for epidemiologic study of work injuries with DAFW because of their size, using high sensitivity for case ascertainment as the evaluative criterion.
Related Concept Videos
Strategies for Assessing and Addressing Confounding
Confounding can be addressed at both the design phase of a study and through analytical methods after data...
Assumptions of Survival Analysis
Bias in Epidemiological Studies
The Availability Heuristic
Archival Research
Statistical Methods for Analyzing Epidemiological Data