Recreational Drowning Prevention Interventions for Adults, 1990–2012: A Review

Abstract

Drowning is a frequently occurring and preventable public health issue. Internationally, drowning literature has focussed on children under 5 years, however, evidence based interventions to prevent adult drowning are needed to reduce deaths on a global scale. The aim of this paper is to systematically identify and analyse the evidence for drowning interventions with an adult focus. A systematic search was undertaken for peer-reviewed articles which were published in English between 1990 and 2012, focused on adults and described a drowning intervention. After quality appraisal by expert reviewers using a purposively tailored checklist, a final total of six studies were included for review. The six studies were all conducted in high income countries. Four were drowning interventions, two were retrospective analyses. The drowning interventions duration ranged from 10 days to 5 years, the analysis studies from 6 to 21 years. Two of the studies reviewed used behaviour change theory to inform development, and two reported formative evaluation. Prevention strategies included education (n = 3), technology (n = 1) and environmental (n = 1). Positive short term effects and significant behaviour change in life jacket use was reported (n = 2). A mixed effect was observed in the six studies. The complexity of the issues surrounding drowning requires the collection of robust data and evaluation of preventative measures to support the development of targeted and tailored prevention interventions. This review reinforces the need for a genuine and sustained global approach to addressing adult drowning prevention. Drowning is a serious public health issue and should receive the same attention as other public health priorities
.

Full text

ORIGINAL PAPER
Recreational Drowning Prevention Interventions for Adults,
1990–2012: A Review
Justine E. Leavy •Gemma Crawford •Linda Portsmouth •
Jonine Jancey •Francene Leaversuch •Lauren Nimmo •
Kristen Hunt
ÓSpringer Science+Business Media New York 2015
Abstract Drowning is a frequently occurring and pre-
ventable public health issue. Internationally, drowning lit-
erature has focussed on children under 5 years, however,
evidence based interventions to prevent adult drowning are
needed to reduce deaths on a global scale. The aim of this
paper is to systematically identify and analyse the evidence
for drowning interventions with an adult focus. A sys-
tematic search was undertaken for peer-reviewed articles
which were published in English between 1990 and 2012,
focused on adults and described a drowning intervention.
After quality appraisal by expert reviewers using a pur-
posively tailored checklist, a final total of six studies were
included for review. The six studies were all conducted in
high income countries. Four were drowning interventions,
two were retrospective analyses. The drowning interven-
tions duration ranged from 10 days to 5 years, the analysis
studies from 6 to 21 years. Two of the studies reviewed
used behaviour change theory to inform development, and
two reported formative evaluation. Prevention strategies
included education (n =3), technology (n =1) and envi-
ronmental (n =1). Positive short term effects and signifi-
cant behaviour change in life jacket use was reported
(n =2). A mixed effect was observed in the six studies.
The complexity of the issues surrounding drowning
requires the collection of robust data and evaluation of
preventative measures to support the development of tar-
geted and tailored prevention interventions. This review
reinforces the need for a genuine and sustained global
approach to addressing adult drowning prevention.
Drowning is a serious public health issue and should
receive the same attention as other public health priorities.
Keywords Drowning Drowning prevention Adults 
Review Evaluation design
Introduction
Drowning is a frequently occurring and preventable public
health issue. In 2011, drowning was the third leading cause
of unintentional injury internationally, resulting in an
estimated 359,449 deaths and accounting for 7 % of all
injury associated deaths (World Health Organisation
(WHO) [16]). The focus of much of the drowning literature
is on children, particularly those aged under 5 [11]. Not-
withstanding this, two-thirds of reported drowning deaths
in 2011 were for people aged 15 and over, with almost
50 % of all drowning deaths occurring in males aged over
J. E. Leavy (&)G. Crawford L. Portsmouth J. Jancey 
F. Leaversuch K. Hunt
Western Australian Centre for Health Promotion Research,
School of Public Health, Faculty of Health Sciences, Curtin
University, GPO Box U1987, Perth 6845, WA, Australia
e-mail: j.leavy@curtin.edu.au
G. Crawford
e-mail: g.crawford@curtin.edu.au
L. Portsmouth
e-mail: l.portsmouth@curtin.edu.au
J. Jancey
e-mail: j.jancey@curtin.edu.au
F. Leaversuch
e-mail: francene.leaversuch@curtin.edu.au
K. Hunt
e-mail: Kristen.hunt@curtin.edu.au
F. Leaversuch L. Nimmo
Health Promotion and Research, The Royal Life Saving Society
(Western Australia Inc.), PO Box 28, Floreat Forum, WA 6014,
Australia
e-mail: lnimmo@rlsswa.com.au
123
J Community Health
DOI 10.1007/s10900-015-9991-6

15 [16]. Under-reporting of global drowning data may
result in significant underestimates in the number of
drowning deaths in adults [13,27].
The risk factors for adult drowning can be categorised
into behavioural and environmental factors [18]. The
higher risk for adult males in high income countries (HICs)
is reflected in their greater level of participation in recre-
ational swimming, boating and fishing [7,21]. This expo-
sure, is exacerbated by higher risk behaviours such as
swimming alone, not wearing a personal floatation device
when boating or rock fishing, and alcohol consumption [7,
8,18]. In contrast, in low and middle income countries
(LMICs), the risk of drowning stems from open access to
waterways as part of daily life [7,18]. This exposure to the
risk of drowning is compounded by a number of factors
including lack of swimming skills, illiteracy, and unsafe
watercraft [7,8]. Accordingly, any adult drowning pre-
vention effort must take into consideration multiple risks,
social and environmental factors [12,19].
Evidence-based interventions to prevent adult drowning
are needed on a global scale. The literature reinforces the
notion that public health interventions that are well
designed and delivered, underpinned by theory and sup-
ported by robust evaluation are more likely to result in
sustained, long term behavioural change [1,4,19,26]
Unfortunately, very few drowning prevention programs are
informed by behavioural theory and/or formally evaluated.
The overall aim of this review is to identify and analyse
the evidence base for adult focused, public health inter-
ventions to address drowning and non-fatal drowning. This
review describes the key characteristics of the studies
identified with the intent of making recommendations that
will ensure future interventions align with evidence based
drowning prevention, and public health practice.
Methods
Search Strategy
To identify published articles on public health interven-
tions to prevent or reduce drowning deaths or non-fatal
drowning incidents with an adult focus, a literature search
was conducted between February and August 2012 using
the following six databases: ProQuest; ScienceDirect; CI-
NAHL; Ovid, PubMed; and the Web of Knowledge (see
Fig. 1). No grey literature was searched.
The search strategy consisted of a basic ‘in topic’ or ‘in
title’ field search or a ‘keyword/abstract’ topic search using
‘boolean phrase and apply related terms’ depending on the
database chosen to search. Titles were identified using
firstly the search term (including all variations of the word)
drowning; and repeated for drowning AND prevention;
drowning AND evaluation; drowning AND best practice;
drowning AND alcohol, drowning AND program; drown-
ing AND intervention; drowning AND water safety. The
search strategy was subsequently repeated using near
drowning; submersion; then submersion AND injury;
submersion AND mortality; and water safety AND
drowning in the title field search.
Criteria for Inclusion
Articles were included that met the following criteria:
published in English between 1990 and 2012; peer
reviewed; a focus on adults; and a drowning intervention
described. To identify studies for inclusion, all titles and
abstracts were screened by two reviewers against the
inclusion criteria. The exclusion criteria are shown below
in Table 1. A list of excluded studies is available on
request.
Selection of Articles, Screening and Quality Appraisal
Article titles and abstracts were examined for appropri-
ateness for full text review by two reviewers (GC and KH).
A hand search of relevant studies was also undertaken by
locating studies that were recommended through experts in
the field of drowning prevention and by reviewing the
reference lists and bibliographies of the final studies
selected for review. A total of six studies were included for
review (see Table 2). The six studies were then quality
appraised by expert reviewers (FL and LN). At the time of
conducting the review the expert reviewer credentials
included Chief Executive Officer of the peak body of an
international drowning prevention organisation, and a
senior public health practitioner in state-based life-saving
organisation. Their combined expertise equates to more
than 25 years of experience in design, delivery and eval-
uation of public health interventions to prevent drowning
and non-fatal drowning. Using a purposively tailored
quality appraisal checklist for interventions adapted from
the National Institute for Health and Clinical Excellence
(NICE) Quality appraisal checklist [15] and the Joanna
Briggs series of assessment and review instruments [9] The
quality appraisal checklist covered five areas which were:
study population; study methods; outcomes; analyses; and
summary (including validity and limitations).
Data Analysis and Synthesis
The data analysis consisted of de-identified, publicly
available data and therefore is IRB exempt. A meta-ana-
lysis was not conducted due to the large heterogeneity
across the small number of studies that met the inclusion
criteria. The following outcomes were sought across the six
J Community Health
123

drowning studies: setting and sample characteristics;
recruitment and response rate; duration, and drowning
prevention strategies, behavioural theory and formative
research, evaluation design, measures and intervention
effects (impact/outcome/results) and limitations. Of the six
studies, four of the studies were drowning prevention
interventions, while the remaining two studies represent
retrospective analysis of the effectiveness of either
drowning prevention policy or practice interventions,
which were deemed to have met the criteria for inclusion in
this review (Table 2).
Results
Setting, Duration and Evaluation Design
All studies were delivered in high income countries; four of
the studies in the USA [2,6,22,24]. The remaining two
studies were delivered in Australia [5] and New Zealand
[14]. The drowning interventions ranged in duration from
10 days [5] to 5 years [14], with sample sizes ranging from
14 [22] to 2,991 participants [24]. In contrast, the analysis
studies ranged from 6 [2] to 21 years [6] and the number of
participants ranged from 1,008 [6] to 1,597 [2]. Of the six
studies, one was a quasi-experimental design [5], one a
matched cohort analysis [2], one a cross sectional time
series analysis [6] and one was an annual cross sectional
survey [14]. Two were observational studies [22,24] (see
Table 2).
Drowning Prevention Strategies
Three of the four drowning prevention interventions studies
described educational interventions, using print educational
materials [5,14,24] including posters, postcards, fact
sheets [5] static on-site displays [14] and celebrity sports
cards [24] all of which focussed on drowning prevention.
One study employed local media to promote safety
Inial studies idenfied
through database searching
(n=167592)
Titles and abstracts screened
(n=310)
Excluded based on tle or
abstract (n=278)
Full text studies screened and
hand search completed
(n=32)
Final studies included
(n=6)
Excluded based on limitaons
and duplicaon (n=167282)
Studies excluded aer
consensus (n=3)
Studies reviewed before data
extracon (n=9)
Excluded at quality appraisal
(n=23)
Fig. 1 Search strategy
schematic
Table 1 Exclusion criteria
Children as the primary target group
Clinical trials
Therapy trials
Medical interventions
Studies focusing on risk factors
Studies focusing on rescuer behaviour
J Community Health
123

Table 2 Review of the drowning prevention interventions 1990–2012
Author
(year)
Setting &
sample (n)
Recruitment/
response rate
Duration; & Drowning prevention
strategies
Behavioural theory
(BT)/formative
research (FR)
Evaluation
design &
measures
Intervention effects: impact/outcome/
results
Limitations of the
intervention
Drowning interventions (n=4)
Hatfield
et al.
(2012)
[5]
New South
Wales,
Australia
Recruitment:
visitors were
selected at
random from the
selected
intervention/and
control sites
Length: unclear. ‘‘January/February’’ stated.
Post intervention interviews took place
5–12 days after campaign launch ‘‘while
the campaign was running.’’
(BT) Cognitive theories of
behaviour
Quasi-
experimental,
pre/post/follow-
up –intercept
interviews,
4–6 months post
campaign
In the intervention area 28.8 % of post-
intervention & 57.2 % of follow-up
respondents had seen the campaign
Contamination between
the control and
intervention sites
Sample:
Beachgoers.
Response rate: pre-
intervention:
Intervention
83.7 %
Strategies: poster, postcard, fact sheet and
brochure aimed at improving recognition
of rips, the danger of rips, the importance
of swimming between the flags, what to do
if caught in a rip and how to avoid panic
(FR)Structured interviews
with beachgoers
assessing knowledge,
beliefs and behaviours
relating to beach flags
and rip currents
Penetration and
recall of
campaign.
Intention,
confidence,
knowledge
Post-intervention respondents who had seen
intervention (compared to not seen) were
more likely to report correct swimming
intentions for rip (p\.0001), identify
(p\.0001) and describe a rip (p\.0001).
Pre- and post-
intervention samples
treated as independent,
although some may
have been interviewed
at both stages.
Pre-intervention:
Intervention
n=180;
Post-intervention respondents in the
intervention area compared to the control area
were more likely (relative to baseline data
from beachgoers visiting the two selected
areas 9 months earlier) were more likely to
report correct swimming intentions for rip
(p=.019), identify (p\.0001) and describe
a rip (p=.0002) and report an intention
never to swim at unpatrolled beaches
(p=\0.001) and signal a guard (p\.0001)
Use of static vs video
footage of rips
Control n =196 There were no significant differences in either
comparison for intention to swim between the
flags
Post-
intervention:
intervention
n=552
Control 56.0 % 6 month follow-up, comparing respondents in
the intervention area with the control area,
demonstrated continued retention of reporting
correct swimming intentions for rip
(p\.0001) and identification of a rip
(p\.0001). Moderately confident in
identifying the rip
Control n =408 Post-intervention:
Follow-up:
intervention
n=222
Intervention
79.3 %
Control n =161 Control 85.3 %
Age range:
14–50 ?years
Follow-up (mail
out to those who
gave contact
details):
intervention
57.1 %
J Community Health
123

Table 2 continued
Author
(year)
Setting & sample
(n)
Recruitment/
response rate
Duration; & Drowning prevention
strategies
Behavioural theory
(BT)/formative
research (FR)
Evaluation
design &
measures
Intervention effects: impact/outcome/
results
Limitations of
the intervention
Sex: 47.4–59.9 %
female across the 6
sample groups
Control 54.9 %
Schwebel
et al.
(2007)
[22]
United States Recruitment:
observed
lifeguards and
patrons at an
outdoor public
swimming pool
Length: weather permitting, 4 days/week over
one summer with 20 observation days before
the education session and 14 observation days
afterwards
(BT) Health Belief
Model
Observational pre
and post data of
lifeguards and
pool patrons
The lifeguards were already looking at the
pool at an adequate level. After the
education session, their scanning behaviour
significantly increased (p\.01) and
distraction significantly decreased (p\.01)
Observational data
only
Sample: Lifeguards
n=14(8 female and
6 male)
Response rate:
N/A
Strategy: education session (1 h) aimed at
increasing lifeguards perception of
susceptibility of drowning, perceived potential
severity of drowning and introduced
techniques to overcome barriers to effective
pool surveillance. Lifeguards aware of
observation.
FR–NR Lifeguard
behaviours:
looking at pool,
scanning and
distractions
Three out of five risky patron behaviour
measures significantly reduced post
intervention: running (p\.01); pushing
under (pB.05); and jumping near others
(p\.01)
Poor statistical
power for
analyses
Age range of
lifeguards16–30 years
Patron behaviours:
running,
pushing under,
diving into
shallow water,
and aggression.
Pool patrons: included
adults (although the
primary group of
concern to the authors
was children)
Inter-coder
reliability was
established
Treser
et al.
(1997)
[24]
King County,
Washington, United
States
Recruitment:
observed
boaters at 12
sites-different
week days and
times
Length: 2 years (1992–1994) BT & FR–NR Observational PFD use significantly increased in adult males
and females (14.2 % in 1992 to 24.7 % in
1994, p\.001).
Changing attitudes
about general
safety could
have played a
role in elevating
usage rates
Sample: recreational
boaters of small
vessels (16 feet or less
in length)
1992: 3 times/
week over
2 months
Strategies: education campaign with boating
safety video for use in elementary schools, a
life jacket loaner program for local beaches,
community events, fliers, brochures and
celebrity sports cards
Number of
recreational
boaters wearing
PFD
Not replicated in
other states
1992: n =1097 (89 %
adult; 73 % male)
1994: 1–3 times/
week over
2 months
Details of community events and the
development, pre-testing and distribution of
media materials not reported
Observers trained
to develop inter-
rater reliability
1994: n =2991 (85 %
adult; 64 % male)
Response rate:
N/A
J Community Health
123

Table 2 continued
Author
(year)
Setting &
sample (n)
Recruitment/response
rate
Duration; & Drowning
prevention strategies
Behavioural theory
(BT)/formative
research (FR)
Evaluation design
& measures
Intervention effects: impact/outcome/
results
Limitations of the
intervention
Moran,
(2011)
[14]
Setting:
Auckland,
New
Zealand
Recruitment: potential
participants approached
if fishing from 4 high-risk
sites or were in transit to
and from the sites, on
summer weekend during
daylight hours
Length: 5 years (2006–2010) BT–NR Annual on-site cross
sectional survey:
awareness, safety
knowledge, attitudes,
preventative
behaviours in
English, Mandarin
and Korean.
48 % of were aware of the project in 2010 Convenience based
sampling: participants
were recent residents,
had ESL or infrequent
visitor to the site.
Direct comparisons
cannot be estimated
Sample: Rock
fishing
participants.
Response rate: not stated Strategies: on-site static on-site
displays of fishing safety, written
material and verbal advice from
trained field officers fluent in
Chinese. Safety messages were also
communicated via television, radio,
newspapers and magazines
(FR) Initial on-site
structured survey to
enhance understanding
of fisher’s socio-
demographic
information, awareness
of fishing safety
project, safety
knowledge, attitudes,
and behaviours.
Responses integrated
into prevention
program strategies.
4-year follow-up almost 2/3rds surveyed each
year were not aware of the Fisher Safety
project (M=64.8 %).
Cross sectional—no
causality can be
determined.
2006 n =250 Fishers’ awareness of severity of risk ass’d
with drowning increased 50 % in 2006 to
66 % in 2010
2007 n =112 Self-efficacy of preventive actions: avoiding
fishing in bad weather, no change 86 %
2006 & 2010
2008 n =241 Knowledge of sea would keep safe, increased
2006, 20 % versus 2010, 33 %
2009 n =128 A comparison of 2006–2010 revealed a
significant decrease in fishers reporting they
never wear a lifejacket (2006: 72 %, 95 %
CI 0.66–0.77; 2010: 34 %, 95 % CI
0.25–0.44); and never wear gumboots or
waders (2006: 64 %, 95 % CI 0.63–0.76;
2010: 45 % (95 % CI 0.36–0.54). There
was a significant negative change in fishers
reporting never drinking alcohol when
fishing (2006: 80 %, 95 % CI 0.75–0.85;
2010: 54 %, 95 % CI 0.49–0.67)
2010 n =107 No other significant reported changes in
behaviours such as turning your back to the
sea, taking a cell phone and checking the
weather beforehand.
J Community Health
123

Table 2 continued
Author
(year)
Setting &
sample (n)
Recruitment/
response rate
Duration; & Drowning
prevention strategies
Behavioural theory
(BT)/formative
research (FR)
Evaluation design &
measures
Intervention effects: impact/outcome/
results
Limitations
of the
intervention
Total n =838
Ethnicity:
45–70 %
Asian over
sample
groups
Age range:
16?
Sex: 83–92 %
male over
the sample
groups
Retrospective analysis of drowning prevention policy or practice (n =2)
Cummings
et al.
(2011)
[2]
United States Records for 104,683
people in 46,234
vessels were
assessed. 1,597
boaters in 625 vessels
met the inclusion
criteria.
Length: 6 years (2000–2006) BT & FR–NR Matched cohort Only 21 % (336/1,597) of the boaters wore a
PFD. Wearers were more often female and
less often aged 20–39 years compared with
persons who did not wear a PFD
Missing
information
for age and
sex
Sample:
Recreational
boaters
n=1,597
Strategies: reviewed US Coast Guards
electronic Boating Accident Report
Database and US compressed
mortality files. Estimated risk ratios
by comparing outcomes for people in
each vessel.
Drowning deaths of boaters
from the same vessel
comparing outcomes for
those who entered the water
wearing a PFD with those
who did not wear a PFD
Unadjusted RR for drowning among those
wearing a PFD, compared with those who
did not 0.78 (95 % CI 0.68–0.91).
Possible
selection and
confounding
bias noted,
plus bias due
to missing
values
Age range:
82 % over
20 years of
age
Adjusted RR was 0.49 (95 % CI 0.31–0.78)
Sex: 90 %
male
Howland
et al.
(1998)
[6]
United States Recruitment: N/A Length: 21 years (1970–1990) BT & FR–NR Cross sectional time series No significant association between drowning
and MLDA was found for any of the age
groups studied.
Failed to show
any
association
between
drowning and
changes in
public policy
J Community Health
123

messages [14], one used both video messages coupled with
educational sessions and community events [24], one used
face to face education sessions [22] and one new tech-
nology in the form of i-Pad presentations shown to
beachgoers [5]. The development and pre-testing of the
media materials was not described by any of the studies.
One US study implemented an environmental strategy in
the form of a loan program of life jackets for local beach
goers [24] (see Table 2).
The two analysis studies assessed the effects of legis-
lative changes in the minimum legal drinking age on
drowning [6], and the association between death by
drowning in recreational boaters and wearing a personal
floatation device (see Table 2).
Behavioural Theory and Formative Research
Two of the studies reviewed stated using a behaviour
change theory to underpin strategy development [5,22].
The Health Belief Model [20] was used to develop an
intervention to ‘‘increase the lifeguards’ perception of
susceptibility of drowning’’ [22], ‘‘educate the lifeguards
about the potential severity of drowning incidents’’, and
‘‘help lifeguards overcome scanning barriers by reviewing
American Red Cross recommendations about strategies for
pool surveillance and introducing alternative scanning
techniques’’ [22]. The other study was less clear about the
specific use of behaviour change theory in the intervention
and instead identified the use of cognitive theories of
behaviour as the framework to develop the education
campaign [5].
Two of the four interventions reported formative eval-
uation [5,14]. In these studies structured interviews with
beach goers and recreational fishers to assess awareness,
knowledge, beliefs and behaviours towards personal beach
safety and fishing safety were completed prior to strategy
design and implementation [5,14] (see Table 2).
Intervention Effects
Of the four drowning prevention interventions reviewed all
reported positive changes in short term effects. These
included awareness of the intervention messages for two of
the interventions, Don’t get sucked in by the rip (Australia)
[5], and Fisher Safety Project (NZ) [14]. Together with
confidence in identifying a rip, intentions to swim outside
a rip, and improved ability to identify a rip [5]. In addition,
perceptions of the severity of the risk of drowning and
vulnerability to risk were reported as ‘appearing’ to have
improved, however perceptions of the efficacy of pre-
ventive action such as avoiding fishing in bad weather did
not change over a 5 years intervention prevention period
[14].
Table 2 continued
Author
(year)
Setting &
sample (n)
Recruitment/
response rate
Duration; & Drowning
prevention strategies
Behavioural theory
(BT)/formative
research (FR)
Evaluation design & measures Intervention effects: impact/outcome/
results
Limitations
of the
intervention
Sample:
n=1,008
Response rate:
N/A
Strategy: used pooled cross sectional
time series analysis to test effects of
changes in MLDA on drowning rates.
Information derived from the US
National Health Statistics Mortality
files
Impact of changes in MLDA on drowning
Ages:
15–23 years
Also considered: urbanicity, unemployment,
poverty, education, divorce, police per
capita, all age liquor law arrests, state beer
tax, and residence within 30 miles of a
border state with a lower drinking age.
Sex:
proportions
not reported
Ass’d associated, CI confidence intervals, MLDA minimum legal drinking age, N/Anot applicable, NR not reported for one or either outcome, RR rate ratio
J Community Health
123

In the Fisher Safety Project (NZ) there was a significant
increase in the use of life jackets by fishers; 72 % (95 % CI
0.66–0.77) in 2010 compared to 34 % (95 % CI 0.25–0.44)
of fishers in 2006. Changes in behaviour were reported for
turning your back to the sea when fishing and checking
weather/water conditions before setting out however, these
were not significant [14]. Personal floatation device usage
increased in adults from 19.9–31.3 % (p[0.001) post
intervention in King County, Washington, US however,
there were no differences between males and females [24].
In the Lifeguard (US) observational study, there was a
change in two of the three lifeguard behaviours including
‘decreased distraction’ and ‘increased scanning’, both
changes in behaviour were significant [22]. Alcohol use
had a negative change in behaviour in the Fisher Safety
Project (NZ), with 54 % (95 % CI 0.49–0.67) of fishers
never drinking alcohol in 2010 compared to 80 % (95 %
CI 0.75–0.85) in 2006 [14].
Of the two analysis studies, one reported that wearing
a personal floatation device reduced the risk of drowning
by almost half [Adjusted RR 0.49 (95 % CI 0.31–0.78)]
[2], and no significant association between drowning and
minimum legal drinking age was found for any of the
adult age groups analysed in the second study [6](see
Table 2).
Identified Limitation of the Study Design
Four of the six studies reported less than optimal par-
ticipant recruitment and sampling [2,5,14,23]. In
addition, observational data collection and missing data
resulted in poor power calculations [2,22], and limited
generalisability.
Discussion
Adult drowning is a multifaceted and complex public
health issue, highlighted by a paucity of published peer-
reviewed research outcomes. This review of six adult
drowning prevention studies delivered between 1990 and
2012 found mixed results for the overall design, delivery
and effectiveness of interventions. All of the interventions
reviewed were delivered in high income countries. Inter-
ventions varied by duration, strategies employed and
diversity of evaluation measures. Five different evaluation
designs were utilised across the six studies. There was little
use of conceptual theories or frameworks and formative
evaluation as part of the intervention design. The four
drowning prevention interventions varied from just several
weeks to 5 years, whilst the analysis studies ranged from
six to 21 years and provided an opportunity to measure
sustained behaviour change. Evaluation focussed on short
term effects such as awareness, intermediate effects such as
knowledge and intention, and changes in drowning pre-
vention behaviour; measurement varied however for
almost all the studies.
Behavioural Theory and Formative Research
Drowning prevention behaviour namely supervision,
attention, and concentration is inherently psychological in
nature, however behavioural theory has been used spar-
ingly in adult prevention program design [22]. Only two
interventions described the use of behavioural theory to
guide the development of their intervention, Don’t get
sucked in by the rip [5] and, the US brief lifeguard inter-
vention [22]. The brief US lifeguard intervention was
framed around three Health Belief Model constructs, and
targeted the lifeguards’ perception of susceptibility for
drowning incidents, the potential severity of drowning and
near-drowning incidents in public swimming pools, and
perceived barriers about conducting surveillance over
longer time periods [22]. Hatfield et al. [5] highlighted their
study as being one of few to thoroughly address near
drowning intervention development and evaluation meth-
odology. However, the authors did not describe in any
detail the theory or constructs used to guide the design and
development of the campaign materials, simply noting the
campaign was developed using relevant psychological
theory [5]. Of interest, only two of the interventions used
formative research to inform the design of educational
materials [5,14] and to subsequently update materials [14],
which is an important component of effective health pro-
motion practice [15].
Overall, the review shows the limited use of behavioural
and evaluation theory, and reinforces the need to employ a
theory-based approach to injury prevention programs spe-
cifically drowning, to enhance the design, implementation
and evaluation [25]. Accordingly, there is a requirement for
a multi-disciplinary team approach when designing future
drowning prevention interventions, to ensure that inter-
ventions are built on robust health promotion foundations
[10].
Previously interventions have focussed and reported on
the proximal or short term effects in the prevention of adult
drowning including: raising general awareness; education
of high risk groups; and promoting the reduction of alcohol
consumption while boating and the use of personal floa-
tation devices [19]. Over 5 years ago, the International Life
Saving Federation (ILS) developed a framework of best
practice to assist nations and organisations to reduce
drowning [8]. This framework recommended continuous
evaluation of interventions with results being well docu-
mented and shared so that results that demonstrate some
effectiveness can inform practice and prioritisation of
J Community Health
123

interventions [8]. In addition, the framework highlighted
that communication and behaviour change theory as an
integral part of the evidence based intervention design was
essential. To date this approach to the design and delivery
of interventions to prevent drowning events has yet to be
implemented widely, resulting in a scarcity of clear, the-
oretically based relevant information and interventions.
Measures
The six interventions reviewed used different measures
including self-report [5,14] and observation [22,24]to
assess changes in drowning prevention behaviour. These
measures have well known inherent limitations, such as
measurement bias and social desirability [3,15]. Two of
the studies retrospectively analysed data to determine
changes in annual drowning deaths [2,5]. In a recent
review of drowning prevention in children and adolescents
they also reported a lack of consistent measures of expo-
sure data [19]. The establishment of valid, reliable and
standardised measures that are comparable across drown-
ing prevention interventions are recommended for future
research. This will enhance capacity for assessing effec-
tiveness across campaigns both nationally and globally.
Drowning Prevention Effectiveness
After 5 years just over one-third of fishers reported that
they were aware of the ‘‘Fisher Safety Project’’ (NZ)
campaign in the previous year. However, they reported
statistically significant decreases in never wearing a life
jacket, gumboots or waders [14]. It is concerning that so
few fishers were aware of the Fisher Safety Project how-
ever, this may in part be explained by the respondents who
included infrequent visitors to the sites where they were
interviewed, recent residents and non-English speaking
fishers [14]. These factors made the measurement of impact
problematic. The Don’t get sucked in by the rip (Australia)
reported a significant impact on correct identification, and
swimming intention, in relation to a rip [5]. Lifeguard
distraction reduced and scanning behaviour significantly
increased in the shorter lifeguard intervention, however, it
was noted that the lifeguards behaviour change ‘dwindled’
over time [22]. As with other public health issues,
drowning prevention requires ‘‘booster’’ campaigns over
the course of time to maintain and sustain behaviour
change outcomes, and ideally would comprise multilevel
interventions including a combination of educational,
environmental and policy components [9,19] Overall, the
intervention effects in this review were short-term and not
replicated, and could be summarised as insufficient evi-
dence to determine intervention effectiveness.
Implications for Policy and Research Needs
Drowning prevention designs and the measurements used
to evaluate the interventions need to be more robust if the
level of observed evidence is to be influential with policy
makers [17]. There is an acute need for adult drowning
interventions to use appropriate study designs, objective,
valid and reliable measures, larger representative samples,
and quality evaluations of a sufficient time period. Fur-
thermore, it is essential that results are reviewed for rigour
and published in the peer reviewed literature [1].
Many drowning prevention interventions are conducted
by community based agencies and, these agencies may not
have the capacity or research expertise to effectively
evaluate programs. In addition, drowning prevention
campaigns may be subject to evaluation that are not doc-
umented and disseminated appropriately. On review, this
seems to be a frequent feature of drowning prevention
interventions that are funded and/or delivered by not for
profit agencies and community based organisations, per-
haps due to a lack of dedicated funding to invest in quality
evaluation, being funded to provide specific outputs or
competing priorities from funding bodies [4].
Better documentation and dissemination could be
addressed through greater collaboration between research
academics and community based organisations. Potentially
this could enhance the distribution of research findings in
both peer, and non-peer reviewed forums, build capacity
and increase knowledge transfer between stakeholders, and
accordingly is recommended as future priority between
practitioners and researchers.
Methodological Limitations
This review has a number of limitations. Although six
databases were included in the search strategy, time limits
were imposed and studies were restricted to publications in
the English language. In addition, evaluations reported
outside peer-reviewed journals were not captured and no
grey (fugitive) literature was reviewed. All final interven-
tions included in the review were delivered in high income
countries and many were from the US, limiting the gen-
eralisability of these findings to other countries or regions,
especially LMICs. It is recognised that they present a
unique and different set of priorities to high income
countries. Accordingly, there is a need for greater, ongoing
research to be undertaken.
Conclusion
The complexity of the issues surrounding drowning
requires the collection of robust drowning data and
J Community Health
123

evaluation of preventative measures to support the devel-
opment of targeted and tailored prevention interventions.
This review reinforces the need for a genuine and sustained
global approach to addressing adult drowning prevention,
with consideration of potential pitfalls to be avoided such
as, quality evaluations of a sufficient time period, and the
need for collaboration around a prioritised research agenda
[13]. Adult fatal and non-fatal drowning are a priority for
prevention and evidence informed health promotion
approaches to intervention implementation and evaluation
should be adopted for maximum gains. Accordingly,
drowning must be recognised as a serious public health
issue and receive the same attention as other public health
priorities.
References
1. Crawford, G., Leavy, J., Portsmouth, L., Jancey, J., Leaversuch,
F., Nimmo, L., & Hills, E. (2014). Development of a systematic
review of public health interventions to prevent children
drowning. Open Journal of Preventive Medicine, 4(3), 100–106.
doi:10.4236/ojpm.2014.43014.
2. Cummings, P., Mueller, B. A., & Quan, L. (2011). Association
between wearing a personal floatation device and death by
drowning among recreational boaters: A matched cohort analysis
of United States Coast Guard data. Injury Prevention, 17(3),
156–159. doi:10.1136/ip.2010.028688.
3. Fisher, R. J. (1993). Social desirability bias and the validity of
indirect questioning. Journal of Consumer Research, 20(2),
303–315. doi:10.2307/2489277.
4. Franks, H., Hardiker, N. R., McGrath, M., & McQuarrie, C.
(2012). Public health interventions and behaviour change:
Reviewing the grey literature. Public Health, 126(1), 12–17.
doi:10.1016/j.puhe.2011.09.023.
5. Hatfield, J., Williamson, A., Sherker, S., Brander, R., & Hayen,
A. (2012). Development and evaluation of an intervention to
reduce rip current related beach drowning. Accident Analysis and
Prevention, 46, 45–51. doi:10.1016/j.aap.2011.10.003.
6. Howland, J., Birckmayer, J., Hemenway, D., & Cote, J. (1998).
Did changes in minimum age drinking laws affect adolescent
drowning (1970–90)? Injury Prevention, 4(4), 288–291. doi:10.
1136/ip.4.4.288.
7. International Life Saving. (2007). World drowning report. Leu-
ven, Belgium. Retrieved from http://www.ilsf.org/sites/ilsf.org/
files/filefield/world-drowning-report-final-sept-27-2007.pdf
8. International Life Saving. (2008). Drowning prevention strate-
gies: A framework to reduce drowning deaths in the aquatic
environment for nations/regions engaged in lifesaving. Leuven,
Belgium: Retrieved from http://www.ilsf.org/sites/ilsf.org/files/
filefield/world-drowning-report-final-sept-27-2007.pdf
9. Joanna Briggs Institute. (2008). Joanna Briggs Institute Review-
ers’ Manual: 2008. Adelaide, SA: Joanna Briggs Institute.
10. Lincoln, J. M. (2006). Occupational drownings. In J. J. L.
M. Bierens (Ed.), Handbook on drowning: Prevention, rescue,
treatment (pp. 73–75). Heidelberg, Germany: Springer.
11. Mackie, I. (2006). Availability and quality of data to assess the
global burden of drowning. In J. J. L. M. Bierens (Ed.), Handbook
on drowning: Prevention, rescue, treatment (pp. 54–56). Hei-
delberg, Germany: Springer.
12. McClure, R. J., Davis, E., Yorkston, E., Nilsen, P., Schluter, P., &
Bugeja, L. (2010). Special issues in injury prevention research:
Developing the science of program implementation. Injury,
41(Supplement 1(0)), S16–S19. doi:10.1016/j.injury.2010.05.
001.
13. Meddings, D. (2011). Drowning prevention:Challenges and
prospects for global progress. Paper presented at the World
Conference on Drowning Prevention, Danang, Vietnam.
Retrieved from Drowning prevention: Challenges and prospects
for global progress.
14. Moran, K. (2011). Rock-based fisher safety promotion: Five years
on. International Journal of Aquatic Research and Education,
5(2), 164–173.
15. National Institute for Health and Clinical Excellence. (2009).
Methods for the development of NICE public health guidance.
London: National Institute for Health and Clinical Excellence.
16. World Health Organisation. (2013). Disease and injury regional
mortality estimates for 2000–2011: Global summary estimates
Retrieved from http://www.who.int/healthinfo/global_burden_
disease/estimates_regional/en/index.html. Retrieved 2 September
2013, from WHO http://www.who.int/healthinfo/global_burden_
disease/estimates_regional/en/index.html
17. Pearn, J. H., Nixon, J. W., Franklin, R. C., & Wallis, B. (2008).
Safety legislation, public health policy and drowning prevention.
International Journal of Injury Control and Safety Promotion,
15(2), 122–123. doi:10.1080/17457300802150587.
18. Petridou, E., & Klimentopoulou, A. (2006). Risk factors for
drowning. In J. J. L. M. Bierens (Ed.), Handbook on drowning:
Prevention, rescue, treatment (pp. 63–69). Heidelberg, Germany:
Springer.
19. Rogmans, W., & Wilson, J. (2006). The challenges of prevention.
In J. J. L. M. Bierens (Ed.), Handbook on drowning: Prevention,
rescue, treatment (pp. 84–87). Heidelberg, Germany: Springer.
20. Rosenstock, I. M. (1974). The health belief model and preventive
health behavior. Health Education & Behavior, 2(4), 354–386.
doi:10.1177/109019817400200405.
21. Royal Life Saving Society – Australia. (2014). National
Drowning Report 2014..http://www.royallifesaving.com.au/__
data/assets/pdf_file/0007/11995/RLS_NDR2014_LR.pdf.
22. Schwebel, D. C., Lindsay, S., & Simpson, J. (2007). Brief report:
A brief intervention to improve lifeguard surveillance at a public
swimming pool. Journal of Pediatric Psychology, 32(7),
862–868. doi:10.1093/jpepsy/jsm019.
23. Schwebel, D. C., Simpson, J., & Lindsay, S. (2007). Ecology of
drowning risk at a public swimming pool. Journal of Safety
Research, 38(3), 367–372. doi:10.1016/j.jsr.2007.01.010.
24. Treser, C. D., Trusty, M. N., & Yang, P. P. (1997). Personal
floatation device usage: Do educational efforts have an impact?
Journal of Public Health Policy, 18(3), 346–356. Retrieved from
http://www.jstor.org/stable/3343315
25. Trifiletti, L. B., Gielen, A. C., Sleet, D. A., & Hopkins, K. (2005).
Behavioral and social sciences theories and models: Are they
used in unintentional injury prevention research? Health Educa-
tion Research, 20(3), 298–307. doi:10.1093/her/cyg126.
26. Wakefield, M., Loken, B., & Hornik, R. (2010). Use of mass
media campaigns to change health behaviour. The Lancet,
376(9748), 1261–1271.
27. World Health Organisation. (2012). Drowning. Retrieved from
http://www.who.int/mediacentre/factsheets/fs347/en/index.html
J Community Health
123