Accident Analysis and Prevention 40 (2008) 496–501

Alcohol consumption among recreational boaters:

Factors for intervention
J.R. Miller, T.J. Pikora ∗
School of Population Health, M431 University of Western Australia, 35 Stirling Highway, Crawley, WA 6009, Australia
Received 12 March 2007; received in revised form 23 May 2007; accepted 7 August 2007

Abstract
Recreational boating is a popular leisure time activity in many countries. It is estimated that, in Australia, boating incidents cause more harm
than rail and air crashes combined and, in terms of transport, are second only to motor vehicle crashes as a cause of serious injury. The consumption
of alcohol among recreational boaters is considered an important risk factor for fatalities and injuries among both operators and passengers. Using
a database of all recreational vessels registered in Western Australia (WA), a sample of 500 adult boaters was recruited to participate in a telephone
survey. The effects of demographic variables and boating characteristics upon the use of alcohol among recreational boaters on their last trip
were explored using logistic regression. The odds of not having a drink were associated, after adjusting for age, with having completed a boating
education course and with carrying children less than 12 years on board. The use of alcohol was not found to be prevalent among WA recreational
boat owners. Based on these findings, it is recommended that efforts to decrease boating-related incidents, such as through education and legislation
measures, be monitored over time to determine the effects of these strategies upon safety behaviours.
© 2007 Elsevier Ltd. All rights reserved.

Keywords: Marine safety; Alcohol; Safety education

1. Introduction
Recreational boating is a popular leisure time activity in many
countries, including Australia. At the national level, it has been
suggested that boating incidents cause more harm than rail and
air crashes combined and, in terms of transport, are second only
to motor vehicle crashes as a cause of serious injury (O’Connor,
2002). Australia-wide there are an average of approximately
1000 people admitted to hospital each year as a result of a boating
incident (O’Connor, 2002). While many more receive minor
injuries that do not require hospital admission, very few non-
fatal injuries or incidents have been investigated for the role that
alcohol may contribute. Measuring blood or breath alcohol levels
are not routine practice among either hospitalised injuries or
for those presenting to emergency departments (EDs), although
legislatively based collection has been required for some other
types of injury, for example motor vehicle drivers (Driscoll et
al., 2003). However, there are several important issues related to
measurement among boaters, including time delays between an
∗ Corresponding author. Tel.: +61 8 6488 7057; fax: +61 8 6488 1188.
E-mail address: terri.pikora@uwa.edu.au (T.J. Pikora).
injury and when blood alcohol levels are determined, that may
restrict the use of these measurements.
While alcohol consumption among recreational boaters is
considered an important risk factor for boating injuries and
deaths, there is limited empirical evidence of the prevalence
of alcohol drinking while boating and the vast majority of
research in this area has been conducted in the United States
(US). This research suggests that between 30 and 40% of recre-
ational boaters consume alcohol while on the water (Howland
et al., 1990, 1993, 1996; Logan et al., 1999). While the relation-
ship between alcohol consumption and boating incidents has
received less scrutiny than that between alcohol and motor vehi-
cle crashes, there is some evidence that alcohol consumption
plays a role in boating injuries and fatalities (O’Connor and
O’Connor, 2005). It has been shown that consuming alcohol
while boating may increase the likelihood of entering the water
(e.g., falling overboard) as well as the ability to avoid drown-
ing once in the water (Browne et al., 2003; Smith et al., 2001).
When this is coupled with evidence that the initial events that
most often result in a boating fatality are capsize and a person
falling overboard (Browne et al., 2003; Bugeja, 2003; O’Connor,
2004), consuming alcohol while boating may exacerbate these
dangers.

0001-4575/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.aap.2007.08.004
 J.R. Miller, T.J. Pikora / Accident Analysis and Prevention 40 (2008) 496–501 497

While considered to be a risk factor for boating injuries and Table 1

deaths, there is limited evidence of the prevalence of alcohol use Characteristics of recreational boaters and whether they reported having had an
alcoholic drink on their most recent boating trip
among recreational boaters in Australia. The aim of this study
was to explore the relationship between consuming alcohol on Characteristics % Alcoholic drink on
the most recent boating trip as well as demographic and boating (n = 500) most recent trip
characteristics among recreational boaters. This information is Yes No
important when considering the development of educational and Gender
legislative strategies to prevent boating incidents, loss of life and Male 79.4 15.4 64.0
injuries. Female 20.6 3.2 17.4
Agea
2. Methods 18–24 years 1.9 0.2 1.6
25–34 years 5.6 1.6 4.0
35–44 years 20.9 4.2 16.7
The Marine Safety Directorate at the department for planning 45–54 years 25.1 5.1 20.0
and infrastructure (DPI) maintains a database of all recreational 55–64 years 23.5 3.0 20.5
vessels registered in WA. Recreational vessels must be regis- 65+ years 23.0 3.0 20.0
tered with DPI if they have a motor and are used on navigable Household structure
waters (Pikora and Cercarelli, 2004). There are more than 72 Single living at or out of home 8.8 2.0 6.8
000 recreational vessels registered in WA (DPI, 2006). Using Young couple no children 4.8 1.8 3.0
this database, 1500 owners of recreational vessels were ran- Family with children aged 12 or under 18.2 3.8 14.4
Family with older children 25.2 5.0 20.2
domly selected and this sampling frame was then divided into
Older couple with no children at home 43.0 6.0 37.0
segments based on type of vessel and location (i.e., metropoli-
tan or regional) so that more owners of common vessel types Been boating in the previous 12 months
Yes 88.4 16.2 72.2
were included to reflect recreational boat ownership across No 11.6 2.4 9.2
the State. A quota of 500 adult boat owners was required
Member of a boating association
from this sampling frame to participate in a telephone sur-
Yes 25.9 6.0 19.8
vey using a computer-assisted telephone interviewing (CATI) No 74.1 12.4 61.6
system. Ethics approval was provided for the survey by the
Completed a boating education course
University of Western Australia Human Ethics Committee. Yes 27.4 7.2 20.2
In 2006, boaters were invited to participate in the survey; if No 72.6 11.4 61.2
they did not wish to be in the study, their telephone number was
Carry any children <12 years on board
withdrawn from the sample and another boat owner was ran- Yes 40.4 11.2 29.2
domly chosen until the quota of 500 boaters was achieved. The No 59.6 7.4 52.2
response rate for the survey was 46.6% (i.e., 500 surveys/1071 Activitiesb
calls made including 380 refusals, 20 language barriers, 113 Fishing 80.4 13.2 67.2
invalid at screening, 58 not available). Information related to Cruising/motoring 33.6 8.8 24.8
age was mistakenly not recorded at the time of the initial tele- Water sports (e.g., skiing and wakeboarding) 10.4 3.0 7.4
phone call. The respondents were re-contacted to provide this Diving 8.2 2.4 5.8
Racing 2.0 0.2 1.8
information however 70 could not be re-contacted that resulted Other 0.6 0 0.6
in missing information.
Location go boating most often
The questionnaire contained a total of 35 separate questions
Protected waters 42.6 9.0 33.6
and included four items directly related to alcohol consump- Within five nautical miles of mainland 30.8 4.0 26.8
tion that were adapted from previous surveys (e.g., Howland et More than five nautical miles from mainland 26.6 5.6 21.0
al., 1996; Smith et al., 2000). Each respondent was asked: on a Missing (n = 70).
your most recent boating trip did you have anything alcoholic to b Multiple responses allowed.
drink; do you agree or disagree that it is safe for passengers on
recreational boats to drink alcohol as long as the skipper does
not drink any alcohol; do you agree or disagree that if someone regression analyses were conducted to examine the association
fell overboard when they had been drinking, alcohol would help between alcohol use and demographic variables, boating charac-
them maintain their body temperature; do you think that alcohol teristics and alcohol knowledge and beliefs. The first regression
affects people less when they are boating because they are in the analysis was a single factor model in which each factor was con-
fresh air; and demographic information including the boat own- sidered in isolation. The final model included only those factors
ers’ age, gender and household structure was also collected. The that were significant (p < 0.05) and age.
remaining items were related to safe boater behaviour and atti-
tudes toward boating safety. The alcohol questions were grouped 3. Results
together approximately midway through the questionnaire.
Statistical Package for the Social Sciences (SPSS) Version 14 As shown in Table 1, the majority of recreational boaters in
for Windows was used to analyse the data. A series of logistic the sample were male (79.4%), aged 45 years or older (71.6%)
498 J.R. Miller, T.J. Pikora / Accident Analysis and Prevention 40 (2008) 496–501

Table 2 Table 4
Boating experience Multivariate analysis of those who did not have an alcoholic drink on their most
recent boating trip compared with those who did
Years Overall boating Driving boata Ownership of
experiencea boata Factors Single Final
factor model
≤2 years 7.2 11.2 14.4
30–5 years 7.0 9.2 15.9 OR OR p-Value 95% CI
6–10 years 12.9 14.3 16.1 Demographic characteristics
11+ years 72.9 65.3 53.6 Age
a n = 498. 18–34 years 1.00 1.00
35–49 years 0.84 0.81 0.662 (0.32–2.05)
50+ years 0.48 0.56 0.563 (0.23–1.40)
and described themselves as part of families with older children
or as older couples with no children at home (68.2%). The major- Household structure
ity of the sample had been boating in the previous 12 months Single/young couple no 1.00
(88.4%), were not a member of any boating association (74.1%) children
Family with children 0.68
and had not completed any formal boating education course aged 12 or under
(72.8%). Less than one-half (40.4%) of the boaters had carried Family with older 0.64
any children aged less than 12 years on board in the previous children
12 months (see Table 1). The majority (72.9%) of the boaters Older couple with no 0.42
had more than ten years overall boating experience, two-thirds children at home
Boating characteristics
(65.3%) had more than 10 years experience driving a boat, while Completed a boating education course
one-half (53.6%) had owned a boat for more than 10 years (see Yes 1.00 1.00
Table 2). No 0.52 0.55 0.031 (0.32–0.95)
The boaters were provided with a list of marine safety issues Carry any children <12 years on board
and were asked to indicate which they considered required more Yes 1.00 1.00
effort or focus. Boating safety education was selected by more No 0.37 0.37 ≤0.000 (0.22–0.63)
than one-third (39.0%) of the boaters as requiring more focus Activities
or effort, while fewer suggested the control of boaters under the Fishing
influence of alcohol or drugs (31.6%) or the control of reck- Yes 1.00
less operation (30.2%). Less than one-fifth of boaters (18.6%) No 1.94
Cruising/motoring
reported having consumed alcohol on their most recent boating Yes 1.00
trip. As shown in Table 3, slightly more boaters (43.8%) agreed No 0.49
than disagreed (40.4%) that it is safe for passengers to drink
Location go boating most often
Protected waters 1.00
Table 3 Within five nautical 0.56
Alcohol practices, knowledge and beliefs miles of mainland
More than five nautical 0.10
Alcohol practices, knowledge and beliefs (n = 500) %
miles from mainland
Had alcoholic drink on most
OR: odds ratio; CI: confidence interval.
recent boating trip
Yes 18.6
No 81.4 as long as the skipper does not drink any alcohol. Only 2.6%
It is safe for passengers to drink alcohol as
of boaters agreed that if someone fell overboard when they had
long as the skipper does not drink any been drinking, alcohol would help them maintain their body tem-
alcohol perature, while 6.2% thought that alcohol would affect people
Agree strongly/agree 43.8 less when boating because they were out in the fresh air.
Neither agree nor disagree 15.8 As shown in Table 4, two variables were significantly and
Disagree strongly/disagree 40.4
independently associated with alcohol consumption on the
If someone fell overboard when they had been drinking, most recent boating trip. Although not statistically significant
alcohol would help them maintain their body
(p > 0.05), one additional variable (age) was included in the final
temperature
Agree strongly/agree 2.6 model as age has been found to influence alcohol consump-
Agree 1.8 tion in previous studies (e.g., Coggan et al., 2001; Glover and
Neither agree nor disagree 18.4 Lane, 1994). After adjustment for age, the odds of not consum-
Disagree strongly/disagree 79.0 ing alcohol was significantly associated (p < 0.05) with having
Do you think that alcohol affects people less when they completed a boating education course and carrying children aged
are boating because they are in fresh air? less than 12 years on board. Compared with those who had
Yes 6.2 completed a boating education course, those with no boating
No 85.2
Do not know/unsure 8.6
education were 45% more likely to report consuming alcohol
on their most recent trip (OR 0.55; 95% CI 0.32–0.95). Simi-
 J.R. Miller, T.J. Pikora / Accident Analysis and Prevention 40 (2008) 496–501
larly, compared with those boaters who carry children aged less
than 12 years on board, those who did not have children onboard
were 63% more likely to report consuming alcohol (OR 0.37;
95% CI 0.22–0.63).
4. Discussion
Previous studies suggest that between 30 and 40% of recre-
ational boaters consume alcohol when on the water (Howland et
al., 1990, 1993, 1996; Logan et al., 1999), although these were
US-based studies and may not be representative of all boaters
across different countries. The prevalence of alcohol consump-
tion was found to be much lower in this study. While 31.6%
of the respondents reported that boaters under the influence of
alcohol or drugs were a safety concern, only 18.6% of those sur-
veyed reported that they had an alcoholic drink while on their
most recent trip. This may not reflect the true measure of drink-
ing behaviour among recreational boaters based on two issues:
many included in the survey may have felt uncomfortable in
reporting drinking due to concerns related to driving and drink-
ing alcohol; and that previous studies have suggested that males
tend to underreport their drinking behaviour (Sommers et al.,
2000). In addition, the sample comprised of an older age group
and only included those boaters with a registered vessel, perhaps
reflecting a group that is more safety conscious.
Perhaps reflecting the small proportion of participants report-
ing that they had consumed alcohol on their last trip, there were
few significant results in the logistic regression analysis. How-
ever, some encouraging results were obtained. After adjustment
for age, the odds of consuming alcohol was significantly associ-
ated (p < 0.05) with having completed a boating education course
and carrying children aged less than 12 years on board.
It is not clear the effect of participating in an education
course has upon alcohol consumption while boating and the
current results may have been influenced by the relatively small
proportion of boaters reporting both alcohol consumption and
completion of a formal boating education course. Nevertheless,
these levels are similar to those reported by Bell et al. (2000),
although this study found no significant differences in the like-
lihood of drinking while boating between those who had been
formally trained and those who had not. Conversely, Glover et al.
(1995) reported that 61% of boaters had received boating educa-
tion and that those with safety education reported higher alcohol
use than boaters without safety education (53% compared to
38%). Similarly, Coggan et al. (2001) found that boaters who
had completed a formal boating education course (64%) were
significantly more likely to have consumed alcohol on their last
boating trip (61%) compared to those with no formal boating
education (42%). This lead to suggestions that the association
between formal training and unsafe boating practices, includ-
ing alcohol consumption, may be due to an increased sense of
confidence associated with both training and greater experience,
and an underestimation of the risks posed by alcohol when con-
sumed on or near the water (Bell et al., 2000; Coggan et al.,
2001). Additionally, it has been suggested that formal training
programs may be lacking in terms of conveying information
about the risks of consuming alcohol while boating (Bell et al.,
499
2000). Thus, there has been a call for these courses to have an
emphasis on safe alcohol consumption while boating (Coggan
et al., 2001). The results from these studies highlight that the
effects of boater training upon alcohol consumption while boat-
ing are mixed and it is suggested that further investigation is
needed to determine the direction of this influence before any
recommendations are proposed.
There are no other published studies assessing the consump-
tion of alcohol among recreational boaters and whether children
have been on board. However, our finding that boaters who carry
children less than 12 years on board were less likely to consume
alcohol is reasonable given that these boaters may be more safety
conscious. It is suggested that this relationship is investigated
further.
The present study found similar results to previous studies
in relation to alcohol knowledge and beliefs. The findings indi-
cate that, in general, recreational boaters are aware of the effects
of alcohol. Only 2.6% of boaters in the present study agreed
that if someone fell overboard when they had been drinking,
that alcohol would help them maintain their body temperature
and 6.2% thought that alcohol would affect people less when
boating because they were out in the fresh air. These results are
consistent with others including Coggan et al. (2001), Howland
et al. (1996) and Smith et al. (2000). Interestingly in this study
there was almost equal number of respondents who agreed and
disagreed with the statement that it is safe for passengers to con-
sume alcohol as long as the skipper does not. While this finding
is similar with that found in a US study (Howland et al., 1996),
more participants agreed that it is safe in a New Zealand study
where participants were asked this question following a cam-
paign of alcohol safety among recreational boaters (Coggan et
al., 2001; Smith et al., 2000). The current findings are important
as it is suggested in the literature that passengers are at as much
risk of injury as boat operators (Coggan et al., 2001; Howland
et al., 1993; McCarthy and Talley, 2001; Smith et al., 2001).
For example, Smith et al. (2001) estimated the relative risk of
dying while boating under the influence of alcohol. In addition
to the finding of a strong positive association of blood alco-
hol concentration (BAC) with risk of death among recreational
boaters, they established that operator and passenger drinking
is associated with the same increased risk of death (Smith et
al., 2001). Therefore, it is suggested that future studies include
alcohol consumption among both operators and passengers to
collect more comprehensive information on the prevalence of
alcohol consumption while boating.
There are several limitations to this study. The first was the
reliance on self-reported data that may have resulted in boaters
under reporting their drinking while boating due to social desir-
ability factors, although the vast majority of previous studies in
this area have also collected self-reported alcohol consumption
data. One exception is a study that, in addition to administering a
self-complete questionnaire, breathalysed participants to obtain
a blood alcohol reading (Smith et al., 2000). This New Zealand
study found a strong correlation between self-reported alcohol
consumption and BAC measurements providing some evidence
that self-reported information is an appropriate method of col-
lecting this data. It is recommended that the issue of the best
500 J.R. Miller, T.J. Pikora / Accident Analysis and Prevention 40 (2008) 496–501
method of collecting alcohol consumption be explored further
and a set of questions be developed that will allow comparative
data to be collected.
Another limitation was that the sample included more older
and experienced boaters. The high proportion of males in the
sample may simply reflect that more males participate in recre-
ational boating than females, although it is not clear whether
the sample is representative of the majority of boaters in WA as
demographic factors are not recorded in the vessel registration
database. As no information was collected from those who did
not complete the survey, it is not possible to determine differ-
ences between those who did and did not complete the survey.
The large proportion of males included in the sample is consis-
tent with other surveys (Coggan et al., 2001; Glover et al., 1995;
Howland et al., 1996; Marine and Safety Tasmania, 2000; Smith
et al., 2000), suggesting that this reflects the population of inter-
est. The inclusion of passengers in future studies may provide a
larger proportion of female participants to be surveyed as well
as a higher number of younger boaters and those who are less
experienced.
Another limitation is that the sampling frame only included
owners of registered vessels and information related to passen-
gers and other people who use the vessel was not collected.
This was because the register of recreational vessels used as
the sampling frame only provided contact information for boat
owners. Databases of registered boat owners have been used
previously to survey recreational boaters (for example Howland
et al., 1996; Mello and Nirenburg, 2004). Alternative methods
for collecting this type of information include conducting sur-
veys at boat ramps and marinas. While this would allow for
the inclusion of passengers, it is resource intensive and would
restrict the sample to boaters at certain locations on a partic-
ular day. Another method is the use of random-digit-dialling
telephone surveys. Using these methods a higher proportion
of females may be obtained in the sample (Bell et al., 2000;
Howland et al., 1990), although these methods have their own
limitations.
There are a number of issues relating to alcohol consump-
tion and recreational boating that warrant further investigation.
Research is needed to explore the effect of boating education
courses on alcohol use among recreational boaters and to mon-
itor the effects of alcohol-related boating laws and education
campaigns upon alcohol use. It may be beneficial to include
passengers in future boating surveys. One challenge will be to
ascertain whether the sample in the present study is represen-
tative of WA recreational boaters and, if not, to gather a more
representative sample.
In Australia relatively few vessel operators involved in fatal
incidents are tested for alcohol and it is suggested by O’Connor
(2004) that this situation would not be tolerated in other areas
of transport. In the case of road crashes, the autopsy process
applied to fatally injured persons requires a blood test and coro-
ners’ records provide BAC information for about 80% of fatally
injured drivers and pedestrians (Chikritzhs et al., 2000). This is
not the case for boating-related fatalities, suggesting that there is
a need to strengthen legislation surrounding testing for alcohol
among boaters.
It has been suggested by O’Connor (2004) that a combination
of legislation and education is needed to reduce the number of
boating-related incidents associated with alcohol. The contribu-
tion of alcohol to boating deaths in Australia (28% in excess of
0.05 BAC) is similar to the contribution to road deaths (26% in
excess of 0.05 BAC) although there is greater BAC enforcement,
policing and legislation among motor vehicle drivers (O’Connor,
2004). This suggests that there is the need for strengthened leg-
islation and enforcement in conjunction with high profile media
and public education activities among boaters (O’Connor, 2004).
It has been suggested by Smith et al. (2001) that countermea-
sures to reduce drinking by all boat occupants are more likely
to be effective in reducing boating fatalities than those that
focus on operators alone. These issues highlight the need for
future research to be focused upon all boat occupants in order
to determine the risks for all boaters.
Although the effects of education upon alcohol consumption
among recreational boaters are not clear and warrants further
research, it is recommended that information regarding the dan-
gers of consuming alcohol while boating should be included
in all boating education courses. While increasing boat oper-
ator training has been suggested to improve boating safety,
current training programs have not been well evaluated (Bell
et al., 2000) and may provide boaters with a false sense of
security. In particular, little is known about the effects of improv-
ing boater knowledge of risks related to drinking and boating
(Bell et al., 2000). The recent introduction of a Recreational
Skippers Ticket, whereby all recreational operators in WA are
required to undergo formal boating training and complete a prac-
tical and written evaluation of boating skills, may provide an
opportunity to explore the relationship between alcohol con-
sumption and education further and to monitor changes over
time.
5. Conclusion
While there was not a high level of reported alcohol con-
sumption among this group of WA recreational boat owners,
this may reflect the sample included in the survey. The use of
alcohol among recreational boaters is an issue that warrants
further research. It is recommended that efforts to decrease
boating-related incidents, such as through education and legis-
lation measures, be monitored over time to determine the effects
of these strategies upon safety behaviours.
Acknowledgements
This research was undertaken as part of the marine safety pro-
gram of research conducted at the School of Population Health
at the University of Western Australia for the Marine Safety
Directorate at the Department for Planning and Infrastructure,
Western Australia. The authors acknowledge the contributions
of staff at the Marine Safety Directorate at DPI for providing
advice and support for the study as well as the participants who
completed the survey. The information contained in this paper
are those of the authors and do not necessarily reflect the views
of the funding organization.
 J.R. Miller, T.J. Pikora / Accident Analysis and Prevention 40 (2008) 496–501
References
Bell, N., Howland, J., Mangione, T., Senier, L., 2000. Boater training, drink-
ing and boating, and other unsafe boating practices. J. Drug Educ. 30 (4),
467–482.
Browne, M., Lewis-Michl, E., Stark, A., 2003. Watercraft-related drownings
among New York State residents, 1988–1994. Publ. Health Rep. 118 (5),
459–463.
Bugeja, L., 2003. Recreational vessel fatalities in Victoria 1999–2002. State
Coroner’s Office, Melbourne, Victoria.
Chikritzhs, T., Stockwell, T., Heale, P., Dietze, P., Webb, M., 2000. Trends in
alcohol-related road injury in Australia 1990–1997. National Drug Research
Institute, Perth, Western Australia.
Coggan, C., Hooper, R., Patterson, P., 2001. Alcohol and Boating Safety
Survey—Stage 2, Centre Report Series No. 60, Injury Prevention Research
Centre University of Auckland, New Zealand.
Department for Planning and Infrastructure, 2006. Recreational boat-
ing [Online], Available from:http://www.dpi.wa.gov.au/imarine/1152.asp
(accessed July 14, 2006).
Driscoll, T., Steenkamp, M., Harrison, J., 2003. Alcohol and Water Safety.
Commonwealth Department of Health and Ageing. Canberra, Australia.
Glover, E., Lane, S., 1994. The incidence of alcohol use and recreational boating
in Beaufort County, North Carolina. Wellness Perspect. 10 (3), 37–43.
Glover, E., Lane, S., Wang, M., 1995. Relationship of alcohol consumption and
recreational boating in Beaufort County, North Carolina. J. Drug Educ. 25
(2), 149–157.
Howland, J., Mangione, T., Hingson, R., Levenson, S., Winter, M., Altwicker, A.,
1990. A pilot survey of aquatic activities and related consumption of alcohol,
with implications for drowning. Publ. Health Rep. 105 (4), 415–419.
Howland, J., Smith, G., Mangione, T., Hingson, R., DeJong, W., Bell, N., 1993.
Missing the boat on drinking and boating. J. Am. Med. Assoc. 270 (1),
91–92.
501
Howland, J., Mangione, T., Minsky, S., 1996. Perceptions of risks of drinking
and boating among Massachusetts boaters. Publ. Health Rep. 111 (4), 372–
377.
Logan, P., Sacks, J., Branche, C., Ryan, G., Bender, P., 1999. Alcohol-influenced
recreational boat operation in the United States, 1994. Am. J. Prev. Med. 16
(4), 278–282.
Marine and Safety Tasmania, 2000. Recreational boating safety review
[Online]. Available from: http://www.mast.tas.gov.au/ (accessed July 17,
2006).
McCarthy, P., Talley, W., 2001. Safety investments, behaviours and injury sever-
ity. Appl. Econ. 33 (6), 701–710.
Mello, M., Nirenburg, T., 2004. Alcohol use by Rhode Island recreational
boaters. Med. Health R. I. 87 (1), 13–14.
O’Connor, P., 2002. Assessment of Fatal and Non-Fatal Injury Due to Boating
in Australia. National Marine Safety Committee, Rozelle, NSW.
O’Connor, P., 2004. National Assessment of Boating Fatalities in Australia
1992–1998. National Marine Safety Committee, Rozelle, NSW.
O’Connor, P., O’Connor, N., 2005. Causes and prevention of boating fatalities.
Accid. Anal. Prev. 37 (4), 689–698.
Pikora, T., Cercarelli, R., 2004. Quantifying Boating-Related Fatalities Injuries
and Incidents in Western Australia. Injury Research Centre, Perth,
WA.
Smith, G., Coggan, C., Hooper, R., Patterson, P., Fairnie, V., 2000. Boat-
ing Safety Survey—Stage 1 [Online]. Available from http://www.health.
auckland.ac.nz/ipc/pdf/cr46.pdf (accessed July 12, 2006).
Smith, G., Keyl, P., Hadley, J., Bartley, C., Foss, R., Tolbert, W., McKnight,
J., 2001. Drinking and recreational boating fatalities: a population-based
case–control study. J. Am. Med. Assoc. 286 (23), 2974–2980.
Sommers, M.S., Dyehouse, J.M., Howe, S.R., Lemmink, J., Volz, T., Manharth,
M., 2000. Validity of self-reported alcohol consumption in nondepen-
dent drinkers with unintentional injuries. Alcohol. Clin. Exp. Res. 24 (9),
1406–1413.