A Review of Interventions for Drowning Prevention Among Adults

Adult drowning is a complex and multifactorial public health challenge requiring community, national and global efforts to mitigate impacts. This study updates the evidence base for public health interventions that address adult fatal and non-fatal drowning. A systematic review was undertaken of the peer-reviewed literature for English-language primary studies published between 2011 and 2021describing a drowning intervention with adults. Twenty-two studies were included. Most studies (n = 16) were conducted in high-income countries. Yearly trends in drowning prevention intervention publications were analysed with 2015 (n = 6) the peak publishing year. Over half of the study designs were pre-post (n = 15). Intervention duration ranged from 4 hours to 11 years. Ten studies described either behaviour change theory or formative evaluation to inform design. Thirteen studies targeted interventions at a population level, seven at a group level and two at individual level. Studies identified a range of prevention strategies, categorised as behavioural (n = 9) (e.g., swimming lessons), socio-ecological (n = 8) (e.g., mandatory personal flotation devices) and mixed (n = 5) (e.g., awareness campaign and barriers to prevent access to water). A range of outcomes were described including changes in awareness, water safety knowledge, attitudes, water safety behaviours and skills, environmental, policy and regulation changes and drowning rates. Findings indicate a small but important increase in the evaluation and publication of effective interventions to prevent adult drowning. The complexity of the issues surrounding drowning requires multi-strategy and context -specific adult focused prevention interventions. Contemporary evidence that identifies effective interventions that contribute to prevention efforts is an essential first step in addressing the challenge.


Introduction
Drowning is a global public health challenge. In 2019, more than 230 000 people drowned, with the majority (90%) occurring in low-income and middle-income countries (LMICs) [1,2]. Close to two-thirds of all reported drowning deaths involved those aged 15 or older [1]. Both fatal and non-fatal drowning have significant impacts, including loss of household income and support, family breakdown and increased burden of care for survivors [3][4][5][6].
Drowning prevention in adults is a complex, multifaceted public health issue [7,8]. Some risk factors (age, male 20,21]. In contrast, in LMICs, males are exposed daily to natural waterbodies as a source of livelihood, using water transport to and from work, including in vessels that are often overcrowded and poorly maintained [12,22,23]. Age is also a significant factor in drowning risk. For example, young people (particularly males aged over 16 years) in HICs are at greater risk of drowning, attributed to risk-taking behaviours, peer pressure, poor aquatic competency and alcohol use [8,21,24,25]. In contrast, for older adults (65 years and over) in HICs, risks relate to bathing, falls into water, poor health, alcohol consumption, medication use and lack of safety preparation around natural water bodies or during recreational activities [26,27].
In 2016, the authors published a systematic review of adult drowning prevention interventions [8]. Only six studies met the inclusion criteria, briefly, three studies reported on interventions using educational strategies, one reported an environmental change intervention, and two studies reported legislative change interventions -all studies were located in HICs [8]. Studies were of mixed design and quality, with limited reference to theory or formative research. The 2016 review called for more robust studies underpinned by evidence-informed health promotion approaches [8]. More recently, the literature has seen an increase in the publication of research focusing on drowning prevention [28] supported by the impetus from the 2021 United Nations (UN) Global Charter for Drowning Prevention [29] which called for strengthened multi-strategy, multisectoral actions in drowning prevention around the globe. This updated systematic review aimed to assess the contemporary evidence base for adult -focused, public health interventions that address fatal and non-fatal drowning.

Methods
The review replicated the steps for the previous review outlined by Leavy, Crawford [8] and Crawford, Leavy [30], in line with the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines [31].

Study Setting and Participants
The study setting and participants matched the description provided in the previous adult review [8]. Briefly, we analysed public health, primary studies, focusing on drowning interventions designed to prevent adult fatal or non-fatal drowning. For the purposes of the review, adults were defined as individuals over the age of 18. Eighteen years of age has been designated as the age at which childhood ends by the UNICEF Convention on the Rights of the Child [28]. Interventions aimed at both children/young people and adults were included where there was clear reporting on the impact of the intervention on adults.

Criteria for Inclusion
Included articles met the following criteria: peer-reviewed, published in English between 2011 and 2021 that described and evaluated adult drowning primary prevention interventions targeting an individual, and/or group and/or population level. We included experimental and observational studies, randomised and non-randomised controlled trials, cohort studies, case-control studies, cross-sectional studies, retrospective analysis, and qualitative studies that evaluated primary adult drowning prevention interventions/strategies [8,30] See Table 1. As established by the previous review, primary prevention strategies were those that eliminate or reduce causes of poor health, and/or promote protective factors. Secondary and tertiary intervention/s (those that include early detection, delaying complications, management, and rehabilitation, which occur after the possibility of prevention e.g., resuscitation) and reviews were excluded [8,30].

Search Strategy
The research team searched 15 databases: PubMed; JSTOR; CINAHL; EMBASE; ERIC; ProQuest; PsycINFO; Scien-ceDirect; Scopus; Global Health; Web of Science; Current Contents; Wiley Online Library; Medline; and Sport Discus. Keywords and MESH terms were: 1. drown* adjacent to (prevent* or safety) within 3 words [MESH: Drowning/ pc [Prevention & Control]. 2. interven* OR evaluat* OR "best practice" OR "good practice" OR "best practise" OR "good practise" OR "health promot*" OR "public health" OR polic* OR program* OR research OR prevent* OR education OR

Outcome Measures
Outcome measures included: drowning rates; water safety behaviour changes, or changes in behavioural intention or drowning awareness, knowledge, attitudes, water safety policy and legislation; changes to environment; and water safety skills.

Selection of Articles, Screening and Quality Appraisal
Endnote X9 [32] and Rayyan.AI [33] supported the selection and screening process [34]. Duplicates were removed, and two research assistants (AV) and (TG) screened article titles and abstracts for relevance against the inclusion criteria using-Rayyan.AI [33], a tool for screening and recordkeeping to facilitate researcher collaboration. Excluded publications were scanned, and one research assistant (TG) randomly cross-checked 10% to identify any selection anomalies. The full-text articles from relevant studies that met the inclusion criteria were retrieved and reviewed by three reviewers (AV, JL & GC) to confirm eligibility for inclusion (see Fig. 1). Reviewer discussion resolved inconsistencies. The reference lists of all included articles were hand-searched for any relevant studies that had not been previously retrieved.
Twenty-six studies were quality appraised by two reviewers (JL and ND), using a purposively tailored quality appraisal checklist adapted from the MetaQAT framework [35]. The tool consists of four domains: relevancy; reliability; validity; and applicability [35]. Appraisers allocated a score to each of the nine criteria in the domains, whereby: Met the criteria = 2, Not sure/Unclear = 1; Did not meet the criteria = 0. A maximum score of 18 was allocated. Studies were then categorised based on the overall score, ≤ 9 = low quality, 10-14 = medium quality; and 15-18 = high quality. After the quality appraisal, four (n = 4) studies were reviewed independently by GC and deemed unsuitable for inclusion. A final sample of n = 22 studies were included in the revie.

Data Extraction, Analysis and Synthesis
Data extraction replicated the procedure from the previous review. Briefly, descriptive and outcome data for all included articles were identified and recorded using a Fig. 1 Systematics Review Flow Chart following PRISMA 2020 flow diagram for new systematic reviews template [31] The study by Schwebel et al. (2011) used a regulatory audit to assess lifeguard behaviour.
Four studies (18%) described using some form of formative research to inform the intervention design and delivery and profile the target audience. Methods included a cross-sectional survey [44] and focus group discussions [51] with community members. Sansiritaweesook et al.

Intervention Outcomes
Outcomes described included: changes in awareness, water safety knowledge, attitudes, water safety behaviours and skills, together with environmental, policy and regulation changes and drowning rates. All but two studies reported some intervention effect [43,50]. Scurati et al. (2019) assessed two different forms of feedback (traditional or through mobile devices) in a breaststroke swim program but found no significant differences. A cross-sectional survey standardised data extraction form by two reviewers (AV, ND) and checked by a third reviewer (CG). Full-text articles were read and annotated to extract information under each of the following headings: author (year) & aims and objective; location & sample (n); recruitment & response rate; intervention level (individual, group, population), strategy type (behavioural defined as actions that individuals take concerning drowning prevention [36]; or socio-ecological defined as part of the social, physical and policy environment [37]), activity type (e.g. education, environmental, regulatory); duration; behaviour theory (BT) and formative research (FR); evaluation design; measures; ethical approval; and impacts/outcomes. Results were reviewed by the authors (JEL, GC, CG) for consistency.
Basic reporting of "pooled" data on age, gender and key findings are included. No meta-analyses were conducted due to the considerable heterogeneity across the studies.

Author
Year Aim Location

Mandatory LJ regulations:
Adult LJ wear increased from 13.7% pre-intervention to 68.1% post-intervention.      Yes Water safety skills Participants described sense of achievement in developing swimming skills, increased confidence in water, and awareness of water safety.

Discussion
This study updates a previous systematic review undertaken in 2016 [8]. This review noted a three-fold increase in studies (from six to twenty-two articles), published between 2011 and 2021, that reported on public health interventions by Matthews et al. (2014) asked beachgoers about hazard signage but found better knowledge of hazards on beaches without signs. Savage and Franklin (2015) found culturally and linguistically diverse (CaLD) participants more likely to complete a swimming and water safety course when using specific Indigenous/ CaLD resources than with the standard course, but less likely to complete the Korean course compared to the standard course. Six studies measured changes across multiple outcome categories [38, 44-46, 57, 58].
Eight studies (36%) measured water safety behavioural changes, with five reporting increases in the use of PFDs or lifejackets [38-40, 42, 44]. Other reported behavioural changes included reduced water entry [41], reduced incidence of risky swimming behaviour [58], and improved lifeguard-looking behaviour [49]. Eight studies (36%) measured changes in water safety skills, with six describing improvements in swimming ability [48,50,53,54,57,59]. The study by Petrass et al. (2018) measured simulated rescue competency, and the two studies by Sansiritaweesook   [68]. The findings of the review highlight the availability, and regulation of lifejacket use as a critical and cost-effective approach to drowning prevention. Those in public health have long advocated for the use of theory to plan, deliver, and evaluate interventions [69]. In this review, the use of behavioural theory and formative research was limited, similar to the findings of our previous reviews with adults [8] and children [7]. Studies used a range of theory with varying levels of application. In general theories were mentioned, but description of their application -to design and evaluation -was limited. Of note, one recent Australian qualitative study [53] described and mapped constructs from two theories, Health Belief Model [70] and the Theory of Planned Behaviour [71]to guide the focus group and individual interview question guide. This study provides an exemplar for future drowning prevention evaluation research. Similarly, Stempski et al. (2015) used the Socio-Ecological Model [72] to outline how policy and systems at a community level may help to improve the factors that influence swimming and water recreation. The use of formative research was limited in included studies, despite being a critical component of effective public health practice [73]. The use of theory has been highlighted as instructive to improve our understanding of the complexities of health behaviour and the environment which influences them [74].
Studies mostly reported positive outcomes from their interventions. Our previous review called for an increase in appropriate study designs, objective, valid and reliable measures and quality evaluations over a sufficient period [8]. The current review found the most frequently reported outcomes included water safety behaviour (mainly the use of PFDs) and water safety skills (predominately swimming skills). These outcomes were mainly evaluated through a pre-post study design, using observational data. While issues of observational bias and misclassification of data were reported, they may be more reliable than self-reporting of behaviour and skills [74]. However, several of these studies had relatively short timeframes of less than a few months [41,45,46,49,50,54,59] and/or were conducted with small samples [49,50,53,54,59], meaning it is unknown as to whether behaviour and skills were sustained over time [74], and likely to impact on both fatal and non-fatal drowning for adults. Of note, a recent bibliometric analysis found 39% of drowning research publications were from Australia and the US [28]. We found similar results with 37.5% of intervention studies from Australia and the US, and most (n = 16) from HICs. Interventions varied by intervention level, duration and design across the 22 studies. Nearly half of studies described the use of behavioural theory or formative research, and all but two interventions reported some intervention effect. Overall, the review found some positive shifts in the design and evaluation of adult focused drowning prevention interventions highlighting the vital contribution of research and evidence for interventions delivered across countries and communities globally.
Promisingly, and in line with contemporary health promotion and prevention approaches [62], more than half (n = 13) of the included studies in this review focused on populationlevel interventions using a socio-ecological or mixed strategy approach. In comparison, our previous review found a predominance of behavioural-only strategies emphasising education [8]. While behavioural interventions have merit, changes are often short-term. They may have limited impact on longerterm, population-level outcomes [63], such as fatal drowning events. In comparison, socio-ecological strategies often involve sustained interventions, such as regulatory or environmental changes, that are more likely to produce populationlevel outcomes [64]. Regulatory or environmental changes are more likely to be cost-effective long term [65] and translate to sustained behaviour change [66]. In this review, six studies assessed changes in drowning rates following either the introduction of regulatory PFD usage or environmental changes (e.g., increased barriers to water or availability of rescue equipment), with all demonstrating a considerable reduction in the number of drownings in the population sustained over time. The novel study by Girasek (2019) used signage and a considerable fine to achieve change. The study is an important reminder that multi-strategy prevention efforts targeting both the person at risk and structural considerations such as using regulatory options are more likely to be successful than programs relying on a single strategy. This finding is consistent with the broader literature on health promotion and prevention [62,67]. By comparison, studies in this review focusing on behavioural-only strategies generally reported on shortterm changes, such as knowledge and swimming skills, often amongst a small sample size and within a short time frame.
Socio-ecological strategies may also be more effective than behavioural-only campaigns for some behavioural measures [67]. For example, Mangione and Chow (2014) observed differences in lifejacket use between an educational lifejacket campaign and mandatory lifejacket regulations. Mandatory regulations increased lifejacket use by more than 50%, whilst the educational campaign increased lifejacket use by only 2% and was unsustained. While the authors could not accurately estimate the costs of implementing campaigns, they noted that prevention interventions. Community-based agencies are likely to deliver most interventions but may be less likely to publish in a peer-reviewed forum; therefore, their contributions may have been excluded. Despite these limitations, this review provides a contemporary overview of drowning prevention interventions for adults in low-, middle-and high-income countries.

Conclusion
Drowning prevention in adults is a complex, multifaceted issue of public health significance but evidence for intervention development has been lacking This review updates the evidence base for adult-focused, public health interventions that address fatal and non-fatal drowning. Since our last review there has been a modest increase in the number of interventions published, together with some positive gains in design, delivery and evaluation over the past decade. Findings reinforce global calls for multi-level, multi-strategy approaches to intervention design, implementation and evaluation for maximum gains aligned with contemporary health promotion and prevention approaches.
Acknowledgements Aleena Varghese Curtin University Master of Public Health student and Tanikka Gravatt, Bachelor of Science (Health Promotion) student for their contribution to data extraction and data synthesis.
Authors' Contribution JEL and GC conceived the study design. Material preparation, data collection and analysis were performed by JEL, GC, CG, MDB and ND. Additional data analysis was performed by CG. The first draft of the manuscript was written by JEL and CG and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Funding Open Access funding enabled and organized by CAUL and its Member Institutions.

Conflict of Interest
The authors have no competing interests to declare that are relevant to the content of this article.

Ethics Approval Not required as this is publicly available data.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. outcomes. Despite being published post the previous review, these studies do not incorporate previous recommendations. Promisingly, six studies measured changes in drowning rates, using a pre-post analysis to compare data across several years. All studies demonstrated a positive impact of interventions on drowning rates, though causation cannot be assumed. Only four studies used a cross-sectional design, measuring a mixture of knowledge, attitudes, skills and behaviour. Previous reviews have cautioned against reliance on self-report measures [7,8,24]. Overall, findings suggest incremental improvements in use of robust and standardised measures by researchers and practitioners, although improvements are warranted.
While almost 90% of drownings occur in LMICs [1,2], the majority (n = 16) of articles included in this review were from HIC settings. Interventions in LMICs are delivered by a range of government and community-based agencies, that may not always have the capacity or resources for large-scale evaluation and knowledge translation [28,75,76]. Alternatively, knowledge translation may not be a priority or evaluation results may not be disseminated widely [77][78][79] or undertaken through other means, such as the grey literature [79,80]. Of interest, a recent scoping review of the grey literature identified intersections for drowning prevention within global reports for occupational, environmental and urban health, refugee and migrant safety and disaster risk reduction, identifying potential partners and sectors to increase drowning prevention efforts [79] across a variety of settings. Additionally, while population approaches to drowning prevention are needed, there remains a role for targeted programs for potential at-risk groups, such as migrants or people from lower socioeconomic backgrounds [10,53]. Three studies in this review focused on migrants [48,53] or female non-swimmers [59]; however, the quality of design varied, sample sizes were small, and one study did not report an improvement [48]. Further evidence of approaches that work with different groups across different settings to address inequities in drownings is warranted.

Strengths and Limitations
This study updates a previous review of drowning prevention interventions amongst adults, totalling 31 years of peer-reviewed literature. Strengths include searching fifteen databases, providing a broad scope, and following procedures from previously published systematic reviews [7,8]. Additionally, use of the MetaQAT tool for quality appraisal is viewed as a strength as it is a public health specific tool. A teamsapproach was adopted to finalise article selection, minimising potential errors. Several limitations are noted. Included articles were restricted to the English language, and grey literature was excluded. We acknowledge that non-English articles and the grey literature may yield valuable information on drowning