Upload
scott-macdonald
View
213
Download
1
Embed Size (px)
Citation preview
Pergamon
Accid. Anal. and Prev., Vol. 30, No. 1, pp. 21-27, 1998 0 1998 Elsevier Science Ltd
All rights reserved. Printed in Great Britain OOOI-4575/98 $19.00 + 0.00
PII: s0001-4575(97)00058-4
COMPARISON OF LIFESTYLE AND SUBSTANCE USE FACTORS RELATED TO ACCIDENTAL INJURIES AT
WORK, HOME AND RECREATIONAL EVENTS
SCOTT MACDONALD*, SAMANTHA WELLS and SHELLEY LOTHIAN
Addiction Research Foundation, Social and Evaluation Research, The Gordon J. Mogenson Building, 100 Collip Circle, Suite 200, London, Ontario, Canada N6G 4X8
(Received 8 April 1997)
Abstract-The purpose of this study was to examine whether risk factors vary for injuries that occur at work, at home, at recreational events and at multiple settings. Three major types of factors were investigated: lifestyle factors such as sleep, stress and exercise; substance use; and demographic characteristics. Data were obtained from a household survey of 882 Ontario adults. In order to determine whether different factors were related to different kinds of accidental injuries, x2 tests were conducted among five injury groups: no injuries; work; home; recreational; and multiple injury episodes (i.e. at least two separate injury episodes in two different settings). The first set of comparisons were conducted between those with zero injuries and each of the aforementioned four injury groups. Those with multiple injury episodes were significantly more likely than those with no injuries to be single, under the age of 30, and have lifestyle problems (i.e. sleep problems, financial problems and a desire for counselling) and substance use/problems (i.e. cigarettes, alcohol, licit and illicit drugs). For comparisons between each of the work, home and recreational injury groups and the no injury group, eight factors were significant altogether, but no single factor was significantly related to more than one injury group. Contrasts between all combinations of injury group pairs (i.e. excluding the no injury group) indicated that risk factors for those with multiple injury episodes differed significantly from those with home and recreational injuries for several characteristics. This exploratory study provides some support that risk factors may vary, depending on
the injury group. 0 1998 Elsevier Science Ltd. All rights reserved
Keywords-Accidents, Injuries, Lifestyles, Substance use
INTRODUCTION
Some studies have been conducted to identify risk factors for injuries occurring at work, home and recreational events. However, few researchers have compared risk factors for accidental injuries across these different settings. Rather, most studies have focussed on one type of injury, making comparisons
across studies problematic. The purpose of this study is to assess empirically whether demographic, sub- stance use and lifestyle factors vary for injuries at work, at home, at recreational events and at multiple injury episodes. Multiple injury episodes refer to individuals with two or more injuries in the past year that occurred in at least two different settings (i.e.
work, home or recreational events).
*Corresponding author. Tel: 001 519 858 5000; fax: 001 519 858 5199; e-mail: [email protected]
PREDICTORS OF WORK, HOME AND RECREATIONAL INJURIES
Alcohol For work injuries, the relationship between alco-
hol problems (especially problem drinking and alco-
hol dependence) and job injuries has been examined
in two reviews of the literature [see Stallones and
Kraus ( 1993); Webb et al. ( 1994) for reviews]. Both
reviews indicate that few studies exist and most have
severe methodological weaknesses such as a lack of control groups. The reviews concluded that alcohol
was likely involved in < 11% of work injuries.
For injuries occurring in the home or during
recreational activities, little research has been con-
ducted on relationships with alcohol consumption and alcohol problems. In a case-control study of
three emergency rooms in Mexico, positive BACs
were not significantly related to home injuries (Borges
et al., 1994). For injuries that occur during recrea-
tional activities, research on alcohol consumption is
21
22 S. MACDONALD et al
sparse and limited to specific recreational activities, such as water sport activities [see Mackie (1978)].
Some studies have been conducted that examine alcohol involvement in different types of accidents. Walsh and McLeod (1983) in a study from Scotland, found that alcohol was involved in 13% of home injuries, 5% of sport injuries and 41% of injuries that occurred during other leisure activities. In a study from California, of those injured and reporting to an emergency room, 12% had positive BAC readings for home injuries, 35% for injuries at the home of another and 0% for injuries at work (Cherpitel, 1994). Additionally, wide variations have been identified in the percentage of people with positive BACs for other injury categories such as falls, cuts, burns and machine injuries (Redfern et al., 1988; Honkanen, 1977, 1993); however, these categories refer more to the specific nature of the injury rather than where they occurred.
Drugs For work injuries, in some studies injury rates
have been compared for those self reporting drug use versus no drug use. For example, findings from two studies indicated that among a sample of employed adults, drug users (i.e. respondents reporting past use of drugs) were significantly more likely to be involved in injuries on the job (Hingson et al., 1985; Macdonald, 1995). In the study by Macdonald (1995) illicit drug use became non-significant when con- trolling for other variables, suggesting the relationship may be spurious. However, licit drug use remained a significant predictor of work injuries when controlling for other variables. Recently, Bross et al. (1992) showed that chemically dependent workers from a manufacturing plant in the U.S. had significantly more injuries than a control group. Other workplace studies have compared accident rates of drug-positive and drug-negative employees, determined by pre- employment drug tests. Some researchers found that employees who tested positive for certain substances were more likely to have work injuries (Zwerling et al., 1990), whereas other researchers failed to find significant differences (Crouch et al., 1989; Normand et al., 1990).
For home and recreational injuries, little research is available on the relationship between drug use and injuries. One study was found comparing benzodiaz- epine use for different injury types. Girre et al. (1988) found benzodiazepines to be present in a high fre- quency of household accidents, but rarely in work accidents.
Other lifestyle factors For work injuries, the role of various lifestyle
factors have been examined, such as stress and slew
problems. Recently, Savery and Wooden (1994) con-
ducted a study of 1400 employees from 61 workplaces and found that stressful life events and daily hassles contributed to work injuries. Several studies have linked sleep problems with work-related motor vehi- cle accidents (Colquhoun, 1976; Folkard and Monk, 1979; Richardson et al., 1989). The relationship between lifestyle factors and injuries occurring at home and recreational events has not been well
researched.
Limitations of existing studies
Unfortunately, very little can be concluded from existing studies regarding the differential importance of risk factors for work, home and recreational injury groups. Few studies have compared the relationship between possible risk factors and injuries across different injury groups and very little research exists for multiple injury episodes. Furthermore, existing research has been carried out in different countries, such as Scotland, Mexico, Canada and the U.S. and the importance of risk factors is likely partially attrib- utable to cultural factors. For example, cultures with a high per capita consumption of alcohol will be more likely to have alcohol related injuries than
countries with a low per capita consumption of alcohol. Furthermore, measures of risk factors vary from study to study. In general, studies have focused on different sets of variables, and comparisons across different types of injuries are rare, which has meant that comparability across studies is limited.
OBJECTIVE
The primary objective of the present study was to assess whether risk factors vary for people with no injuries, work injuries, home injuries, recreational injuries and those with multiple injury episodes.
METHODS
Sample selection
The sample for this study was obtained from all telephone directories across Ontario. The number of cases selected from each telephone directory was proportional to the actual population for each region. A systematic sampling technique was used to select cases from telephone directories. The sampling interval was determined by dividing the number of pages in each book by the number of cases sampled from each book. A random number table was used to obtain the initial page number and location on ;he page.
Comparison of lifestyle and substance use factors related to accidental injuries 23
Questionnaire distribution and collection Questionnaires were mailed to 3300 households
in June and July of 1992. Enclosed with the question-
naire was a letter describing the purpose of the study, a stamped self-addressed envelope for return and a stamped postcard. Respondents were asked not to place their names on questionnaires so that infor- mation obtained would remain anonymous. Approximately 2 weeks after the questionnaires were
mailed, subjects were phoned to encourage their participation and answer questions about the study.
The response rate From the original sample of 3300 individuals,
355 were ineligible to participate in the study (i.e. retired, not working, deceased, too young). Additionally, 477 packages were returned because the
occupant had moved (185) the address was incom- plete (161), or the address was unknown (48). For 83 cases, it was unclear as to why the packages had been returned because no information indicating the reason was provided.
Questionnaires were delivered to the homes of 2468 individuals. Of these, 882 participated in the
study, producing an effective response rate of 35.7%; using the original sampling frame (i.e. 3300) the response rate is 26.7%. Mail surveys often produce a similar response rate, especially when the question- naire is lengthy and somewhat complicated. Of these participants, 57 cases were excluded from the analyses because they were not employed at the time of the study.
Measurement Respondents were asked to indicate the number
of injuries requiring professional medical attention they had in the prior year at work, at home and during recreational activities outside the home. Several variables measuring drug and alcohol use were also included in the questionnaire. Respondents were asked if they used licit drugs (i.e. sleeping pills, stimulants or tranquillizers) or illicit drugs (i.e. cocaine, marijuana, LSD, heroin or solvents) in the previous year. They were asked to complete the Shortened Michigan Alcoholism Screening Test (SMAST) and the Drug Abuse Screening Test (DAST) to assess problem drinking and drug use (Selzer et al., 1975; Skinner, 1982). Respondents also reported the number of standard drinks consumed in the prior week and whether they had quit drinking or using drugs. A number of other questions were also included, such as demographic characteristics and lifestyle problems. All of these variables were dichotomized because they were either highly skewed or only ordinal level measurements.
RESULTS
Description of the sample Demographic characteristics of the sample
(gender, marital status, age and major occupational categories) were compared to the population of employed adults in Ontario. The sample was not perfectly representative of the population (compari- sons for each variable were statistically significant). For instance, the sample contained more males
(66.0% versus 54.2%), more people in older age groups (36.2% versus 28.7% that were 45 and older), and slightly more people in managerial, administra- tive and professional occupations. Nonetheless, the sample contained diverse groups in terms of occupa- tion and industry breakdowns.
In the previous year, 37 people reported work injuries, 21 had injuries at home, 39 had injuries during recreational activities, and 19 people had multiple injury episodes (in more than one setting).
The percentage of respondents from this study report- ing drug use were as follows: sleeping pills (6.8%),
tranquillizers (4.1%), pep pills or stimulants (1.9%), marijuana or hashish (8.6%) and cocaine (1.7%). These percentages were close to figures found in other Canadian surveys (Giesbrecht et al., 1992; Health and Welfare Canada, 1988a,b). For example, corre- sponding figures for Canada’s Health Promotion Survey were sleeping pills 8.2%, tranquillizers 6.4%, stimulants 1.9%, marijuana 5.6%, and cocaine 0.9% (Health and Welfare Canada, 1988a). Similarly, the
percentage of respondents reporting consumption of 29 or more drinks in the week preceding the survey was 2.9 compared with 2.4 reporting 28 or more drinks in Canada’s Health Promotion Survey (Health and Welfare Canada, 1988a). In terms of composite measures, 8.9% reported using any illicit drug over the past year, 11.0% used licit drugs, 33.0% scored above zero on the DAST scale and 19.8% showed indications of an alcohol problem according to SMAST. Associations among substance use variables were examined to assess their reliability. Statistically significant relationships were found between the DAST scale and illicit drug use (p < 0.0001 ), between DAST and licit drug use (p<O.OOOl), and between the SMAST scale and weekly alcohol consumption
(p<0.0001).
Analyses comparingfactors associated with d@erent injury groups
In order to determine whether different factors were related to different types of accidental injuries, x2 tests were conducted. Each possible risk factor was dichotomized and cross tabulated with the five injury categories: no injuries; work; home; recrea-
24 S. MACDONALD eta1
tional injuries; and multiple injury episodes (i.e. two or more injury locations) producing a test of signifi- cance for each 2 x 5 table (see Table 1). If this overall test of association was significant, post hoc explor- atory analyses (i.e. 2 x 2 x2 tests) were conducted to the assess which pairs of injury groups contributed significantly to the overall test of association. When
significant associations were found (p < 0.05), per- centages of people with each possible risk factor were presented in Table 1 for each group. For example, 54% of those with no injuries (zero) sometimes had trouble getting to sleep or staying awake, compared to 73% reporting the same characteristic who had a work injury in the prior year. The overall x2 tests
(2 x 5 tables) were significant for 19 of the 30 variables
analysed.
Zero injuries versus work, home, recreational and
multiple injury episode groups
The next set of comparisons were carried out between those with zero injuries and each of the four injury groups. Numerous risk factors (16 of 30) were significant for those with multiple injuries compared to no injuries. Those with multiple injury episodes were significantly more likely than those with no injuries to be single (63% versus 31%) and under the age of 30 (47% versus 12%). In terms of lifestyle factors, those with multiple injury episodes were more likely to report sleep problems, financial problems and a desire for counselling. Substance use was especially prevalent among those with multiple inju- ries. They were significantly more likely than those
with zero injuries to smoke cigarettes, report alcohol problems, and report use of stimulants, tranquillizers, marijuana and LSD or other hallucinogens. The separate comparisons between those with zero injuries and those from the other injury groups (i.e. work, home and recreational injuries) produced eight of 90 possible significant comparisons. Interestingly, no single factor was significant for more than one
injury type.
Work, home and recreational injury groups compared
The next set of comparisons was carried out to
determine whether risk factors vary between each pair of the three injury groups: work; home; and recreational injuries. Only two variables, namely edu- cation and exercise, differed significantly across the three major injury types (i.e. excluding those with zero or multiple injuries). Education levels were sig- nificantly higher for those who had injuries at home as compared with those who had injuries at work. The percentage of respondents who exercised at least once a week was significantly higher among those
with recreational injuries compared to those with work injuries.
Multiple injury episodes versus work, home and
recreational injury groups The last set of comparisons was conducted
between those with multiple injury episodes and each
of the three other injury groups (i.e. work, home and recreation). First, the multiple injury group did not differ from the work injury group for any variable. For comparisons between the multiple injury group and home injuries, three variables were significant. In terms of demographic characteristics, those with multiple injury episodes were significantly more likely to be under 30 years old (i.e. 47%) than those with home injuries (5%). About 63% with multiple injury episodes were single compared to 32% with home injuries. As well as those with multiple injuries were significantly more likely than those with home injuries to have used marijuana and any illicit drug in the
prior year. Four variables were significant for compar- isons between multiple injuries and recreational inju- ries. In terms of lifestyle factors only one variable was significant: those with multiple injuries were significantly more likely to report sleeping problems (84%) than those with recreational injuries (54%). The multiple injury group also was significantly more likely to have used stimulants and report a drug use problem than the recreational injury group.
LIMITATIONS
Several limitations should be noted in this study. One limitation is the use of self report data on drug and alcohol use, which is known to be subject to under-reporting. Rush et al. (1993) found that self reported alcohol consumption accounted for < 50% of actual alcohol sales. It is very likely that many subjects failed to report their drug use and others underestimated the extent of their use.
More specific information on factors related to each injury was not available, such as the severity of the injury or the type of professional medical atten- tion. Information was not obtained that indicated whether people were tired or under the influence of drugs or alcohol at the time of their injuries and no distinction was made between minor cuts or bruises and severe injuries, such as broken bones. Other variables not included in this study might better explain some types of injuries, such as lack of experi- ence, which likely decreases with age. Similarly, a greater propensity to take risks might play a causative role in injuries but was not included as a variable in this study. As well, detailed information on home environments and the recreational activities in which
Tab
le
1.
Ove
rall
prob
abili
ty
valu
es
for
the
asso
ciat
ions
be
twee
n di
ffer
ent
vari
able
s an
d ty
pes
of
inju
ries
, se
para
te
com
pari
sons
be
twee
n si
gnif
ican
tly
diff
eren
t in
jury
gr
oups
(p
~0.
05)
and
perc
enta
ges
with
ea
ch
char
acte
rist
ic
for
sign
ific
ant
com
pari
sons
be
twee
n ea
ch
vari
able
an
d th
e in
jury
gr
oups
id
entif
ied
(i.e
. 2
x 2
tabl
es)
Zer
o ve
rsus
w
ork,
ho
me,
re
crea
tion
al,
Wor
k,
hom
e,
and
recr
eatio
nal
Mul
tiple
inj
urie
s ve
rsus
wor
k,
and
mul
tiple
inj
urie
s in
juri
es c
ompa
red
hom
e an
d re
crea
tiona
l in
juri
es
Zer
o Z
er0
Zer
o Z
ero
Wor
k W
ork
HO
me
Wor
k H
0me
Rec
. O
vera
ll pr
obab
ility
va
lue
x2 f
or
vers
us
vers
us
vers
us
vers
us
vers
us
vers
us
vers
us
vers
us
vers
us
vers
us
each
2 x
5 ta
ble
wor
k ho
me
rec.
m
ultip
le
hom
e re
c.
rec.
m
ultip
le
mul
tiple
m
ultip
le
Var
iabl
es
Dem
ogra
phic
fa
ctor
s Si
ngle
A
ge u
nder
30
yea
rs
Mal
e L
ower
ed
ucat
ion
IllC
O~
~
Lif
esty
le
fact
ors
Som
etim
es
or o
ften
ha
ve
trou
ble
getti
ng
to s
leep
or
stay
ing
awak
e O
ften
w
orn
out
at t
he e
nd
of t
he d
ay
Not
to
o ha
ppy
Lif
e fa
irly
or
ver
y st
ress
ful
Exe
rcis
e at
lea
st
once
pe
r w
eek
Dhag
ree
to.
1 ge
t en
ough
sl
eep
Agw
r to
: I
rare
ly
wor
ry
abou
t th
ings
I
usua
lly
have
no
mon
ey
in t
he b
ank
befo
re
payd
ay
I am
con
cern
ed
mar
riag
e is
in
trou
ble
I w
ould
lik
e to
hav
e co
unse
lling
Sub
stan
ce
use
vari
able
s Ci
gare
ttes
Smok
e on
ave
rage
of
at
leas
t on
e ci
gare
tte
per
day
Alco
hol
Had
an
alc
ohol
ic
drin
k in
pas
t ye
ar
Eve
r qu
it dr
inki
ng
Alc
ohol
pr
oble
m
by t
he S
hort
ened
M
ichi
gan
Alc
ohol
Sc
reen
ing
Tes
t (M
AST
) 29
or
mor
e dr
inks
la
st
wee
k O
ther
dr
ugs
Eve
r qu
it dr
ags
Dru
g pr
oble
ms
by t
he D
rug
Abu
se
Scre
enin
g T
est
(DA
ST)
U,w
in
the
past
yea
r of
i Sl
eepi
ng
pills
P
ep p
ills,
st
imul
ants
T
ranq
uiliz
ers,
su
ch
as V
aliu
m
Coc
aine
/cra
ck
Mar
ijuan
a or
has
hish
L
SD/a
cid/
mus
hroo
ms
Any
lic
it su
bsta
nce
Any
ill
icit
subs
tanc
e
0.03
42
0.00
18
0.34
46
0.00
49
0.00
551
33%
ver
sus
20%
3 I %
ver
sus
63%
12
% v
ersu
s 47
% 61
% v
ersu
s 20
%
32%
ver
sus
63%
5%
ver
sus
47%
0.02
27
54%
ver
sus
73%
54
% v
ersu
s 84
%
54%
ver
sus
84%
0.16
39
0.41
15
0.72
98
0.01
80
72%
ver
sus
92%
70
% v
ersu
s 92
%
0.16
27
0. I
797
0.00
06
19%
ver
sus
39%
19
% v
ersu
s 56
%
0.71
20
0.02
10
5% v
ersu
s 24
%
0.02
13
22%
ver
sus
4 I %
22
% v
ersu
s 44
%
0.89
01
0.43
09
0.00
06
16%
ver
sus
35%
16
% v
ersu
s 47
%
0.04
55
3% v
ersu
s 16
%
0.01
65
18%
ver
sus
47%
0.
0027
28
% v
ersu
s 68
%
31%
ver
sus
68%
0.34
12
0.00
22
0.00
62
0.00
18
0.00
10
0.00
13
0.01
00
0.00
17
I % v
ersu
s 8%
2% v
ersu
s 16
%
0% v
ersu
s 16
%
3% v
ersu
s 21
%
7% v
ersu
s 32
%
0% v
ersu
s 32
%
1% v
ersu
s 8%
1%
ver
sus
11%
9%
ver
sus
32%
7%
ver
sus
32%
5%
ver
sus
32%
Zer
o,
zero
in
juri
es;
wor
k,
wor
k in
juri
es;
hom
e,
hom
e in
juri
es;
rec.
, re
crea
tiona
l in
juri
es;
mul
tiple
, in
juri
es
in m
ultip
le
setti
ngs.
W
hen
the
expe
cted
va
lues
w
ere
< 5
in
any
give
n ce
ll,
Fish
er’s
E
xact
T
est
was
us
ed.
26 S. MACDONALD et al.
respondents were involved was not collected. In this study, exploratory analyses were employed, involving numerous x2 tests. A limitation of this approach is that the likelihood that a null hypothesis will be falsely rejected (Type I error) is increased consider- ably. Therefore, caution should be taken when inter- preting the results.
The response rate was low, which is not unusual for mail surveys; however, it reduces the generalizabil- ity of the findings. The sample was not perfectly representative of the population, with fewer females and fewer people in younger age groups. Since the purpose of this paper is to better understand variables related to accidental injuries, the failure to achieve
representativeness is not viewed as a critical limita- tion. A final limitation is that the number of respon- dents with different injury types was low. Therefore, the number of cases in each cell was below preferred levels and the reliabilities of estimates were less than ideal. Due to the low response rate and small sample size, these results should be treated as preliminary and suggestive.
DISCUSSION
Little research has been conducted on factors that might be differentially related to different injury groups. Most existing studies focus on one type of injury which makes comparisons across studies prob- lematic due to varying methodologies. As well, studies have been conducted in diverse countries around the world with varying methods and measures, making generalizations problematic. A strength of the present study is that the same variables were examined across different injury groups.
For comparisons between those with zero injuries
and each of the injury groups, some lifestyle factors
were important for certain types of injuries, but not for others. For example, those with work injuries and those who had injuries in multiple locations were significantly more likely than those with zero injuries to have sleep problems. No significant difference was found for home and recreational injury groups com- pared with the zero injury group. A possible explana- tion for this finding is that people feeling tired at home are more likely to rest. Similarly, people might be less likely to engage in recreational activities when they are tired. At work, however, most employees are unable to rest until the end of the day and their accident risk may be higher due to a lack of attentive- ness on the job. Cigarette smoking was also signifi- cantly different for those with work injuries compared with those with zero injuries, but not for home or recreational injuries. A possible explanation is that in workplaces where smoking is prohibited, employ-
ees might experience withdrawal effects, which may increase accident risk. If smoking is permitted at work, perhaps the act of smoking itself increases one’s accident risk. Another interesting finding was that those with home injuries and those with multiple injuries were significantly more likely to have alcohol problems compared to those with zero injuries. The recreational and work injury groups were not signifi- cantly different from the zero injury group in terms of alcohol problems. Perhaps this is because people are generally more likely to drink at home rather
than at work, or at recreational events, thus elevating their risk of injury. Finally, those with multiple injuries appear to be significantly more likely to use and abuse a wide range of licit and illicit substances compared with those without injuries.
Further analyses were conducted that directly compared work, home and recreational injury groups. These findings were useful for assessing whether risk factors vary for these three main injury categories. Overall, respondents with work, home or recreational injuries did not differ significantly from each other in terms of demographic, lifestyle and substance use variables. Two exceptions were those with recrea-
tional injuries exercised significantly more than those with work injuries, and those with home injuries had significantly more education than those with work injuries. These findings do not provide support for the hypothesis that different risk factors exist for different types of accidents (excluding the zero and multiple injury categories), which has been suggested in some studies (Cherpitel, 1994; Girre et al., 1988).
Analyses were conducted comparing the multiple
injury episode group with work, home and recreational
injury categories. Those with multiple injuries were significantly different from those with home injuries in terms of illicit drug use. The elevated drug use among those with multiple injuries is likely partially
related to the younger age of this group. The multiple injury group also differed significantly from those with recreational injuries for marital status, sleep problems and drug problems. The multiple injury category is unique because individuals in this group not only had more than one injury, but their injuries took place in more than one location. Therefore, individual factors rather than the environment may
predispose these people to injuries. The findings from this study indicate that risk
factors vary when comparing the zero injury group with the work, home and recreational injury groups. However, when work, home and recreational injury groups were compared directly with each other few significant differences were found. Finally, the find- ings suggest that risk factors vary when the multiple injury group is compared with other injury groups.
Comparison of lifestyle and substance use factors related to accidental injuries 27
More extensive research needs to be conducted in order to adequately understand how risk factors vary for different injury groups. Studies employing large samples and examining a broad array of relevant variables are needed to provide definitive conclusions. As well, more research is needed that focuses on people who have injuries at multiple locations.
Disclaimer The views expressed in this document are those
of the authors and do not necessarily reflect those of the Addiction Research Foundation.
Acknowledgements-The authors are indebted to Robert Mann for useful comments on earlier versions of this paper.
REFERENCES
Borges, G., Garcia, G., Gil, A. and Vandale, S. (1994) Casu- alties in Acapulco: results of a study on alcohol use and emergency room care. Drug and Alcohol Dependence 36, l-7.
Bross, M. H., Pace, S. K. and Cronin, I. H. (1992) Chemical dependence: analysis of work absenteeism and associated medical illnesses. Journal of Occupational Medicine 34, 16-19.
Cherpitel, C. J. ( 1994) Alcohol and casualties: a comparison of emergency room and coroner data. Alcohol and Alco- holism 29, 21 l-218.
Colquhoun, W.P. (1976) Accidents, injuries and shiftwork. In Shiftwork and Health, eds P.G. Rentos and R.D. Shepard, pp. 160-175. Department of Health Education Welfare, Washington, D.C.
Crouch, D., Webb, D., Peterson, L., Buller, P. and Rollins, D. (1989) A critical evaluation of the Utah power and light company’s substance abuse management program: absenteeism, accidents, and costs. In Drugs in the Work- place: Research Evaluation Data, eds SW. Gust and J.M. Walsh. National Institute on Drug Abuse, Rockville, MD.
Folkard, S. and Monk, T. H. (1979) Shiftwork and perfor- mance. Human Factors 21, 483-492.
Giesbrecht, N., Gliksman, L., Douglas, R.R. et al. (1992) Questions and Answers: Alcohol and Other Drug Use in Ontario. Addiction Research Foundation, Toronto.
Girre, C., Facy, F., Lagier, G. and Dally, S. (1988) Detec- tion of blood benzodiazepines in injured people: relation- ship with alcoholism. Drug and Alcohol Dependence 21, 61-65.
Health and Welfare Canada (1988a) Canada’s Health Pro- motion Survey: Technical Report. Minister of Supply and Services, Ottawa.
Health and Welfare Canada (1988b) Licit and illicit drugs in Canada. Minister of Supply and Services, Ottawa.
Hingson, R., Lederman, R. and Walsh, D. (1985) Employee drinking patterns and accidental injury: a study of four New England states. Journal of Studies on Alcohol 46, 298-303.
Honkanen, R. (1977) Records based on clinical examina- tion as an indicator of alcohol involvement in injuries at emergency stations. Scandinavian Journal of Social Medi- cine 5, 91-95.
Honkanen, R. (1993) Alcohol in home and leisure injuries. Addiction 88, 939-944.
Macdonald, S. (1995) The role of drugs in workplace inju- ries: is drug testing appropriate?. The Journal of Drug Issues 25, 703-722.
Mackie, I. (1978) Alcohol and aquatic disasters. Medical Journal of Australia 1, 652-653.
Normand, J., Salyards, S. D. and Mahoney, J. J. (1990) An evaluation of pre-employment drug testing. Journal of Applied Psychology 75, 629-639.
Redfern, T. R., Rees, D. and Owen, R. (1988) The impact of alcohol ingestion on the orthopaedic and accident service. Alcohol Alcoholism 23, 415-419.
Richardson, G.S., Miner, J.D. and Czeisler, CA. (1989) Impaired driving performance in shift workers: the role of the circadian system in a multifactorial model. Alco- hol, Drugs and Driving 5(4)/6(l), 265-273.
Rush, B., Tyas, S., Lee, K. and Giesbrecht, N. (1993) Regional variations in the prevalence of various indices of alcohol consumption, heavy drinking and “alcohol- ism”: a comparison of alternative methods and implica- tions for policy and program development in Ontario. Addiction Research Foundation, Toronto, Ontario.
Savery, L. K. and Wooden, M. (1994) The relative influence of life events and hassles on work-related injuries: some Australian. Human Relations 47, 283-305.
Selzer, M. L., vinokur, A. and VanRooijen, L. (1975) A self-administered Short Michigan Alcoholism Screening Test (SMAST). Journal of Studies on Alcohol 36, 117-126.
Skinner, H. A. (1982) The drug abuse screening test. Addic- tive Behaviors 7, 363-371.
Stallones, L. and Kraus, J. (1993) The occurrence and epi- demiologic features of alcohol-related occupational inju- ries. Addiction 88, 945-951.
Walsh, M. E. and McLeod, D. A. D. (1983) Breath alcohol analysis in the accident and emergency department. Injury 15, 62-66.
Webb, G.R., Redman, S., Hennrikus, D.J. et al. ( 1994) The relationships between high risk and problem drinking and the occurrence of work injuries and related absences. Journal of Studies on Alcohol 55, 434-446.
Zwerling, C., Ryan, J. and Orav, E. J. (1990) The efficacy of pre-employment drug screening for marijuana and cocaine in predicting employment outcome. Journal of the American Medical Association 264, 2639-2643.