This study describes the toxicology testing status and results in a series of 427 victims who died on the job in Iowa between 2005 and 2009. Iowa law requires that autopsies be performed in all cases of “work and farm-related deaths unless there is an obvious natural cause of death” (Iowa Code 641-127.3(331,691)). Most, if not all, autopsies of work-related fatalities should include toxicology testing. In our study, however, only a little more than two-thirds of fatally injured workers had toxicology tests performed. Furthermore, certain types of victims were more likely to be tested than others: (1) those who died the same day of their traumatic injury and (2) workers who were less than 50 years of age. Independent of age, agricultural workers were much less likely to undergo toxicological testing than other occupations.

Prior studies report that slightly higher proportions (87–95%) of fatally injured workers underwent toxicology testing [10,11,12,14] compared with our estimate of 69%. This disparity may be due to undercounting of work-related deaths that are then investigated and tested for drugs and alcohol, particularly in studies where only one source of data exists. Most cases identified by coroners and medical examiners underestimated work-related fatalities as defined by federal agencies [15,16]. In rural states like Iowa, misclassification of work-related injuries is particularly problematic for cases of work-related deaths to farmers—a group prone to high misclassification due to the nature of the agricultural work (rural and isolated) and the demographics of farmers (older, often retired or farming as a secondary or part-time job) [14,17]. Similarly, fatally injured workers who die more than 24 h after their injury and older workers may also be more difficult to ascribe the injury as work-related. As an exception, one US national study using the Census of Fatal Occupational Injuries (CFOI) also reported lower rates of testing; toxicology reports were available for only about a fourth of all cases [13]. Our FACE data, like CFOI, is a multiple source surveillance system that likely captures work-related fatalities more reliably than single source studies [19]. Although there are factors associated with the likelihood of being tested (age, occupation, survival time), we found no evidence of selection bias. Hence, while younger workers were more likely to be tested than older workers and farmers less likely to be tested than other occupations, fatal injury cases with toxicology testing were neither more nor less likely to be positive than the untested cases. Our findings still underscore the need for complete testing of all occupational fatalities, nonetheless. In fact, our reduced sample size of tested victims (from n = 406 to n = 280) led to power limitations in our analyses and thus an inability to construct multivariable models to identify predictors for positive test status. Furthermore, sample collection sites (femoral, heart, subclavian, vitreous, antemortem, etc.) were not controlled, so the ability to assess postmortem redistribution was not able to be taken into account. Postmortem redistribution could be a factor in determining a positive or negative result if the concentration for a particular drug is near the reporting limit.

A number of findings related to the toxicology test results are noteworthy. Of those tested, 22% were positive for alcohol or other substances, an estimate just slightly higher than those reported (17%–20%) in prior research conducted in the U.S. and Australia [10,11,12,13,14]. Alcohol and cannabinoids were among the most frequently detected substances, which is also rather consistent with prior studies which have consistently identified alcohol as the top substance found in toxicology reports of fatally injured workers [9,10,11,12].

Unlike prior studies construction and maintenance workers in our study had the lowest proportion of positive test results compared with other occupations [13]. Furthermore, in all other studies previous to ours, deaths attributed to motor vehicle crashes were most frequently positive for alcohol and/or drugs compared to other external causes [10,11,12,13,14]. In our study, positive toxicology tests were more often found among deaths due to other land transport or other causes. A number of explanations are possible. First, our results are limited to the population of Iowa workers, who, in fact, may differ from workers in other settings. Second, the choice of drugs for testing was not consistent across studies. The Australian study, for example, limited its drug screenings to alcohol, cannabinoids, and methamphetamines [10]. The Texas study included drugs used for medical treatment [12]. Finally, studies used different tissue samples for testing. In our study, resulting positive toxicological testing from any testing matrix, including blood, vitreous samples and/or urine, was acceptable. Our goal was to identify the presence of any substance in all types of human samples used in toxicology testing of fatally injured workers.

A result of interest was that workers 65 years or older tested positive for substances more frequently than younger workers, although differences were not statistically significant. With the greying of baby boomers comes a growing number of older workers, who comprised about 19% of the U.S. workforce in 2009. Older workers suffer the highest rate of fatal injury compared with their younger counterparts, and are at high risk of fatal injury [27,28]. Although we cannot determine if the drugs indeed led to fatal injury, use of drugs and prescription drugs for this age group, in particular, may impair judgment and motor coordination. Hence, further research particularly with a larger sample size is needed to determine if older workers have indeed a high incidence of drug-related fatalities.

The study has a few limitations. Classification of external causes was quite broad due to our small sample of tested and positive cases. Positive toxicology test results do not necessarily mean that workers were impaired at the time of the traumatic injury [12]. In fact, residual metabolites or traces of substances remain in the body for weeks after ingestion. Toxicology tests were also conducted in at least 13 laboratories across IA and in four neighboring states which may lead to some level of misclassification due to different samples collected for testing or different reference levels used to ascertain positive test status. Finally, while we capture prevalence of drugs and alcohol in fatally injured workers, we cannot generalize our estimates to non-occupational fatalities or even populations currently active in the workforce.