The Effects of Substance Use on Workplace Injuries

Critique of Objective Measures of Substance Use
While studies that use objective measures of substance use may reduce underreporting of drug 
use via self-report, there are certain limitations to this approach as well. First and foremost, 
there is little information on how accurately a positive 
preemployment test may predict use after 
employment, and drug testing will not necessarily capture patterns of drug-using behaviors 
that may influence occupational injuries, such as binge drinking. On the other hand, drug-
testing results will capture drugs used by workers for only a limited, proximal time span and 
thus miss certain substance-using patterns that may be relevant among those who are “devi-

---

14 The Effects of Substance Use on Workplace Injuries
ance prone.” Hospital records or company referrals are also likely to underestimate the preva-
lence of substance-use disorders. According to a recent, nationally representative survey, only 
10 percent of those with a probable substance-use disorder used specialty substance-abuse 
services in the past year (Mojtabai, 2005). Finally, it is possible that such measures are subject 
to selection biases: There may be systematic differences between workers who are tested for 
alcohol and other drugs and those who are not or between which workers with substance-use 
problems access EAPs and which do not.
Fatal Occupational Injuries
Finally, the seven studies presented in Table 3.4 looked at the effect of alcohol and other drug 
use on fatalities that occur while at work, typically using postmortem toxicology reports to 
identify the presence of alcohol or other drugs in the system.
All but one of the studies in this category looked at the proportion of decedents in work-
related accidents with alcohol present in their systems (Bernhardt and Langley, 1999; Green-
berg, Hamilton, and Toscano, 1999; Harrison, Mandryk, and Frommer, 1993; Lindström, 
Bylund, and Eriksson, 2006; Lipscomb, Dement, and Rodriguez-Acosta, 2000; Sahli and 
Armstrong, 1992). Collectively, these studies estimate the presence of alcohol among occupa-
tional fatalities to range from 4 percent to 20 percent, with the majority falling between 15 
and 20 percent. These studies also vary with respect to the population that was studied (e.g., 
work-related road fatalities [Harrison, Mandryk, and Frommer, 1993] to fatalities among con-
struction workers [Lipscomb, Dement, and Rodriguez-Acosta, 2000]), as well as by the level 
of blood alcohol that indicated a positive screen. The other study (Lucas and Lincoln, 2007) 
examined fatal deaths from fishing vessels in Alaskan waters for which substance use was indi-
cated in the investigative records; over a 15-year period, the researchers estimated that 20 per-
cent of these fatalities involved alcohol.