Drug Testing and Workplace Accidents
Paul Rountree, M.D.
Over the past twenty years the use of workplace drug tests have become commonplace. Both
Hanson1 and Zwerling2 have presented excellent overviews of the history of drug and alcohol
testing for industry in this country. At present approximately 35 million drug tests are performed
each year at a direct cost exceeding one billion dollars. 3 This money is distributed among
numerous parties within an enormous drug testing industry, and recipients include laboratories,
third party administrators, medical review officers (MROs), substance abuse professionals (SAPs),
specimen collectors, and others. Indirect costs of this effort are rarely contemplated and involve
decreased productivity as a result of time lost for testing. Tests are typically performed before
employment, after accidents, for suspicious behaviors, in a random fashion, or for follow-up of
individuals with a history of drug use. It is estimated that companies with drug-testing requirements
employ half of the American workforce. The vast majority of employees believe that these tests
deter drug use. Most also believe that drug tests reduce accidents and product defects.4
There has always been a close relationship between safety concerns and workplace drug
screening programs. Over the years several spectacular accidents have been associated with
drugs. Proof of drug use was noted, for example, after the 1981 crash of an EA-6B Prowler on the
USS Nimitz, the 1987 wreck of a Conrail/Amtrak train in Maryland, and the 1989 environmental
disaster resulting from the Exxon Valdez incident. These episodes have highlighted concerns
about job performance by impaired workers. Intuitively it would seem that drug use is associated
with workplace accidents, and many drug-testing advocates have offered estimates of very high
injury and fatality rates among involved employees,5, 6, 7 but most scientific authorities agree that
there is little data regarding such a relationship.2,8, 9,10,,11
Few studies have been published about drugs and their association with on-the-job injuries
because of several barriers. Accident rates of drug users are rarely compared with matched
groups of non-users. Unlike a blood alcohol test, correlations between a positive urine test, a
drug’s pharmacologic effect, and related levels of impairment are generally unknown. Studies
about vehicular accidents note that 40-80% of drivers who have positive drug tests also are
intoxicated with alcohol, 12 but many times employees involved in accidents on the job are not
tested for alcohol or other legal drug use. In addition, several investigators have noted that drug
and alcohol use is significantly lowered among work injury patients compared to non-work related
injury victims. 13,14 Reports about drug use and accidents rarely determine causality or include
consideration about variables such as time of day, weather conditions, or other circumstances that
may be involved.
Personality traits obviously influence behavior. Sensation seeking activities have been associated
with a number of risky jobs, sports, sexual activity, drug and alcohol use, reckless driving, and
driving under the influence.15 Drug use may simply be a marker of a risk-taking behavior. One
study of post-office workers, for example, noted a significantly increased relative risk of accident
and injury among cigarette smokers, when compared to the non-smokers.16
Published Studies about the Relationship between Accidents and Drug Tests
In Harris County, Texas, Lewis and Cooper reported on 196 fatal work-related injuries occurring
between 1984-85. Of this group alcohol was detected in 13.3% of cases examined, and drugs
capable of altering physiologic function in 7% of cases examined. Only one individual, however,
was noted to have a positive illicit drug test.17 In 1989 Taggart described a dramatic decline in
“personal injuries” associated with fewer positive tests in a railroad industry following the institution
of drug and alcohol testing in 1984.18
A prospective study of 2,537 new post office workers in Boston was reported in 1990 and revealed
that those who tested positive for either cocaine or marijuana had a relative risk of workplace injury
of 1.85. Those positive for marijuana had a relative risk for accidents of 1.55, while those positive
for cocaine had a relative risk for accidents of 1.59.19 The post office employee study did not
evaluate alcohol use. A later similar study of 5,465 post office job applicants at multiple sites failed
to demonstrate a significant association between drug-test results and injury and accident
A study of 459 workplace deaths conducted in Alberta, Canada revealed 10 workers who tested
positive for cannabis, and 40 who tested positive for alcohol. Fifty tested positive for either
prescription or non-prescription drugs. 21 A retrospective review of deaths in New Mexico noted
that information about the presence of alcohol or drug use was available in 530 of 613 occupational
deaths. The presence of alcohol was detected in 13.6% of deaths, and 6.4% of these were legally
intoxicated. Drugs of abuse were present in 4.5% of the deceased, but 1.3% of these were also
positive for alcohol. 22 Another Canadian study of 470 occupational fatalities that occurred between
1986 and 1989 in Ontario noted the presence of alcohol in 2% of those tested, and the presence of
cannabis in 17% of those tested, but no other drugs of abuse were identified.23
Investigators reviewed claims of work-related accidents and injuries from 1984 until 1988 at a
company with 48 facilities in Wisconsin. Twelve utilized drug testing and 36 did not. Accident
rates were identical, but post-accident testing in 3 facilities was associated with a decrease in
claims when compared to the pre-testing period, or with facilities only using pre-employment
In an attempt to analyze the effect of the federal Department of Transportation (DOT) regulations
regarding drug testing Swena and Gaines reviewed fatality rate per 100 million vehicle miles and
performed regression analysis for the period from 1984 until 1989. They were unable to establish
an association between the initiation of drug testing and decreases in fatal truck accidents. 25
A study focusing on the construction industry used questionnaires from 69 companies who
responded of 405 companies (17%) selected. The majority of participants represented larger
companies. The workers’ compensation experience rating modification factors (MODs) were noted
to be lower in companies utilizing drug tests. Additionally the average injury claim rate declined
after institution of drug testing. 26
Alcohol Use and Workplace Accidents
Trent has reviewed several studies relating alcohol use and occupational injury. 13 He has
concluded that the relationship between work-related fatality and alcohol use is small, and that
published studies of non-fatal injury and alcohol use suggest only a weak to moderate effect. He
concluded that alcohol use poses a much greater risk for non-work-related injury than that for work-
Stallones and Kraus published an excellent review of studies of alcohol-related occupational
injuries. They concluded that the extent of the problem remains unknown. 27 They noted reasons
for difficulty in the assessment of a relationship, but stated that estimates suggested alcohol was a
factor in 3-4% of occupational injuries. Dawson studied the association between the frequency of
heavy drinking and occupational injury, and noted an increased risk for light, moderate, or heavy
daily drinking. 28
Lipscomb et al reviewed 3955 deaths that occurred between 1988 and 1994 among construction
workers in North Carolina. There were significant differences between work-related and non-work-
related fatalities. Fifty-seven (57) percent of the non-work-related victims were impaired by
alcohol, compared with 5 percent of the victims with work-related injuries. 14
Veazie and Smith studied a group of 8569 respondents in the 1989 annual interview of a
population (age 24-32) from the National Longitudinal Survey of Youth. They reported that
common occupational injuries (excluding sprains and strains) may not be strongly associated with
alcohol dependence in this group, and suggested that confounding with other risk factors may
explain the association between being a heavy drinker and occupational injuries in this population.
Eighteen years after the institution of the drug free workplace there has been an incredible growth
in workplace drug testing. In addition there is ample evidence that the great majority of employers
and workers support these programs. 4,30 Unfortunately there remains little scientific evidence of a
substantial association between workplace accidents, injury, and drug or alcohol use.
The Substance Abuse and Mental Health Services Administration (SAMHSA) has written a
proposal that is predicted to expand the use of workplace drug tests, and this proposal was
recently released for public comment. Details of the proposal are available on the SAMHSA
website (http://workplace.samhsa.gov/ResourceCenter/DT/FA/GuidelinesDraft4.htm). The new
proposal will allow three additional types of drug tests (sweat, oral fluid, and hair). These
modalities are expected to cost more than urine drug tests, and will require additional training and
experience in new collection procedures. Medical review officers will also require additional
training. The American Civil Liberties Union (ACLU) has suggested that it now costs $77,000 to
find one drug user in the current federal government drug testing program.31 Perhaps because of
the costs of testing, some reports suggest that employers are beginning to utilize drug tests less
Several reports have suggested that injury claims decline after the institution of post-accident
testing, and anecdotal information suggests that workers’ compensation insurance companies are
encouraging the use of this practice. In Texas it is common for insurance carriers to deny claims if
a post-accident test is positive.
It is critical to remember that the original intent of drug testing was to discourage drug and alcohol
use in the workplace. There appears to be little proof that workplace drug tests have reduced drug
use in America. Few good studies have been performed regarding drug and alcohol as a cause of
a significant portion of workplace accident and injury. At some point employers are likely to
question the “return-on-investment” and insist on strategies that minimize interruption of the work
schedule while focusing on work groups with greatest risk or likelihood of testing positive. These
groups would include those exhibiting suspicious behaviors and those involved in follow-up testing
after a history of drug abuse.
Drug or inappropriate alcohol use may simply be a marker of risk-taking behavior. As a rule, men
more frequently use drugs than do women. In addition, single people are more likely to use drugs
than are married individuals. Some industries have high rates of drug use. According to a report
released by the Substance Abuse and Mental Health Services Administration (SAMHSA), for
males these occupations included those in the entertainment industries, athletes, those in food
services (including bars), and construction workers. Among women the occupations with the
highest rates of illicit drug use were in food services, social work, and the legal professions.
The average rates of positive tests in pre-employment and random testing programs seem small
enough to justify a decrease in the frequency of pre-employment and random testing for many
industries. Ultimately, the use of drug tests should be related to the prevalence of drug use in a
specific worksite in order to gain the greatest return on investment.
1 Hanson M. Overview on drug and alcohol testing in the workplace. Bul on Narcotics XLV(2):3-44
2 Zwerling C. Current practice and experience in drug and alcohol testing in the workplace. Bul on
Narcotics XLV (2): 155-96 (1993).
3 Anonymous. Fed panel says value of workplace drug testing unproven. Forensic Drug Abuse
Advisor 7(1):6 (1995).
4 Institute for a Drug-Free Workplace Gallup Survey.
5 Quayle D. American productivity. The devastating effect of alcoholism and drug abuse. Am
Psychologist: 454-8 (Apr 1983).
6 Bureau of Labor Statistics. BLS links alcohol or drug abuse to one in eight workplace deaths.
Alcoholism Report 23(3):7 (Mar 1995).
7 Bensinger PB. Drugs in the workplace. Behavioral Sciences and the Law 3(4)441-53 (1985).
8 Feinauer DM. The relationship between workplace accident rates and drug and alcohol abuse:
the unproven hypothesis. Labor Studies Journal:3-15 (Winter, 1990).
9 Kaestner R, Grossman M. Wages, Workers’ compensation benefits, and drug use: indirect
evidence of the effects of drugs on workplace accidents. AEA Papers and Proceedings:85(2):55-60
10 Substance abuse: injuries and accidents.
11 Kaestner R. Illicit drug use and labor market outcomes: a review of economic theory and its
empirical implications. J Drug Issues 28(3): 663-81 (1998).
12 Lowenstein SR, Koziol-McLain J. Drugs and traffic crash responsibility: a study of injured
motorists in Colorado. J Trauma 50(2):313-320 (2001).
13 Trent RB. Emergency room evidence on the role of alcohol intoxication in injury at work in the
U.S. Safety Science14:241-52 (1991).
14 Lipscomb HJ, Dement JM, Rodriguez-Acosta R. Deaths from external causes of injury among
construction workers in North Carolina, 1988-1994. Appl Occup Environ Hyg 15(7):569-80 (2000).
15 Zuckerman M, Kuhlman, DM. Personality and risk taking: common biosocial factors. J
Personality 68:(6) 999-1029 (2000).
16 Ryan J, Zwerling C, Orav EJ. Occupational risks associated with cigarette smoking: a
prospective study. Am J Public Health 82(1):29-32 (1992).
17 Lewis RJ, Cooper SP. Alcohol, other drugs, and fatal work-related injuries. J Occup Med 31:23-8
18 Taggart RW. Results of the drug-testing program at Southern Pacific Railroad. In Drugs in the
workplace: research and evaluation data. Rockville, MD. National Institute on Drug Abuse: 3-13
19 Zwerling C, Ryan J, Orav EJ. Costs and benefits of pre-employment drug screening. J Am Med
Assoc 264:2639-43 (1990).
20 Normand J, Salyards SV, Mahoney JJ. An evaluation of pre-employment drug testing. J Applied
Psychology 75:629-39 (1990).
21 Alleyne BC, Stuart P, Copes R. Alcohol and other drug use in occupational fatalities. J Occup
Med 33(4):496-500 (1991).
22 Fullerton L, Olson L, Crandall C, Sklar D, Zumwalt R. Occupational injury mortality in New
Mexico. Ann Emerg Med 26(4): 447-54 (1995).
23 Shannon HS, Hope L, Griffith L, Stieb D. Fatal occupational accidents in Ontario, 1986-1989. Am
J Ind Med 23(2): 253-64 (1993).
24 Feinauer DM, Havlovic SJ. Drug testing as a strategy to reduce occupational accidents: A
longitudinal analysis. J Safety Research 24:1-7 (1993).
25 Swena DD, Gaines Jr, W. Effect of random drug screening on fatal commercial truck accident
rates. Internat J Drug Testing 2:1-13 (1999).
26 Gerber JK, Yacoubian Jr. GS. An assessment of drug testing within the construction industry. J
Drug Education 32(1): 53-68 (2002).
27 Stallones L, Kraus JF. The occurrence and epidemiologic features of alcohol-related
occupational injuries. Addiction 88:945 (1993).
28 Dawson DA. Heavy drinking and the risk of occupational injury. Acid Anal Prev 26(5):655-65
29 Veazie MA, Smith GS. Heavy drinking, alcohol dependence, and injuries among young workers
in the United States labor force. Alcohol Clin Exp Res 24(12):1811-19 (2000).
30 Howland J, Mangione TW, Lee M, Bell N, Levine S. Employee attitudes toward work-site alcohol
testing. J Occup Environ Med 38(10):1041 (1996).
31 Anonymous. Drug testing: a bad investment – executive summary.
http://www.aclu.org/DrugPolicy/DrugPolicy.cfm?ID=9218&c=79 (Sep 30, 1999).
32 Meisler A. Drug testing’s negative results. Workforce Management: 35-40.
http://www.workforce.com/section/06/feature/23/53/63/index.html (Oct 2003).
33 Anonymous. 2001 AMA Survey: Medical Testing. American Management Association.
34 Office of Applied Studies: Drug use among U.S. workers: Prevalence and Trends by Occupation
and Industry. http://www.samhsa.gov/oas/work1996/ch4.htm