BACKGROUND: The Bureau of Labor Statistics (BLS) publishes annual estimates of nonfatal OSHA recordable occupational injuries and illnesses for the US and most states based on employer-reported data collected through the Survey of Occupational Injuries and Illnesses (SOII). Washington State undertook several studies to investigate various aspects of underreporting in SOII including: the magnitude of and characteristics associated with underreporting; employers’ understanding of workplace injury and illness recordkeeping requirements, and reasons why SOII-eligible workers’ compensation claims fail to be reported to BLS.
METHODS: Findings from three studies are summarized: a) a linkage of nine years of SOII data to Washington workers’ compensation data; b) a telephone survey administered to SOII respondents across four states; and c) a study in which SOII respondents with Washington workers’ compensation claims identified as unreported in SOII (based on SOII-workers’ compensation record linkage) were interviewed and asked why the claim was omitted from the injury data they reported for SOII.
RESULTS: In 2003–2011, one in five lost time claims were likely ineligible for SOII case reporting because of non-immediate work absence. Among SOII-eligible claims, an estimated 70% were reported in SOII. Claims involving sprains or strains, employer protests, and those not eligible for work disability payments until months after the initial injury were least likely to be reported in SOII. Unreported claims were most common among large construction establishments and small educational services establishments. Claims among state and local government establishments were most likely to be reported. Claims went unreported in SOII for numerous reasons: the employer did not consider the injury work-related, incorrectly counted days of missed work, failed to maintain complete injury records, or simple data entry errors. Most SOII respondents demonstrated an inadequate understanding of the injury and illness reporting requirements, and suggested they may over-report minor cases, and under-report cases involving days of missed work.
CONCLUSIONS: Because the reasons for underreporting are varied, improvements to SOII data may require a multipronged approach. Effective recordkeeping training materials that address common recordkeeping misperceptions; increased communication between the SOII data collection staff and respondents throughout the survey period to facilitate an improved understanding of the reporting requirements; data collection forms that incorporate error checks; and enrollment of establishments in SOII for a multi-year period instead of one year to increase familiarity with recordkeeping and to capture injuries that progress in severity over multiple years may improve the accuracy of employer-reported SOII data.