Fatal drowning in Indonesia: understanding knowledge gaps through a scoping review

Abstract Little is known about unintentional drowning deaths in Indonesia, the world’s fourth most populous and largest archipelagic country. This study aimed to describe the epidemiology and risk factors of unintentional drowning in Indonesia and explore existing health promotion and drowning prevention approaches in Indonesia within a socio-ecological health promotion framework. A scoping review, guided by PRISMA-ScR, was conducted to locate peer-reviewed studies and government reports/policy documents published until May 2023, in English or Indonesian language, using MEDLINE (Ovid), CINAHL, Informit, PsycINFO (ProQuest), Scopus, SafetyLit, BioMed Central and Google Scholar, Indonesian journal databases (Sinta, Garuda) and government agencies websites around the terms: drown, swim, flood, hurricane, cyclone, disaster, water rescue and maritime/boat safety. This review identified 32 papers. However, a paucity of information on unintentional drowning rates, risk factors and prevention in Indonesia was noted. The unavailability of a coordinated national drowning data collection system in Indonesia, from which national and subnational subcategory data can be collected, underlines the possibility of under-representation of drowning mortality. The association between various exposures and drowning incidents has not been fully investigated. An over-reliance on individual-focused, behaviour-based, preventive measures was observed. These findings highlight the need for improving drowning surveillance to ensure the availability and reliability of drowning data; and strengthening research to understand the risk factors for drowning and delivery of drowning prevention programs. Further policy development and research focusing on health promotion approaches that reflect a socio-ecological approach to drowning prevention in Indonesia is imperative.


BACKGROUND
Drowning is the third leading cause of death by unintentional injury worldwide, after road injury and falls (Ahlm et al., 2013;World Health Organization, 2014, 2016;Mokdad et al., 2016;Franklin et al., 2020).Most drowning deaths worldwide (91%, or 337,240 drowning deaths) occurred in low-and middle-income countries (LMICs), particularly in Southeast Asia (35%, or 130,149 drowning deaths), underlining the importance of providing reliable information on unintentional drowning deaths to inform the development, implementation and evaluation of drowning prevention interventions and policy within countries in this region (World Health Organization, 2014).However, understanding of unintentional drowning deaths in Indonesia, the world's largest archipelagic and fourth most populous country with high numbers of meteorological and hydrological events, is limited (Cenderadewi et al., 2020).To address the global burden of drowning, effective health promotion approaches which address drowning at all levels will be required, and comprehensive information on drowning incidents, including in Indonesia, is vital for the planning and implementation of relevant prevention approaches.
Indonesia is one of the world's most natural disaster-prone areas and is at risk from multiple hazards, including floodings, landslides, earthquakes, tsunamis, volcanic activities and tropical cyclones (Global Facility for Disaster Reduction and Recovery, 2019).Indonesia is also the largest archipelagic state worldwide, consisting of 16 056 islands, extending 5150 kilometres between the Indian and Pacific Oceans (Global Facility for Disaster Reduction and Recovery, 2019; Indonesian National Bureau of Statistics, 2019; United Nations, 2019; Embassy of Indonesia for the United States of America, 2019).Understanding the potential contribution of water transport-related and disaster-related drowning deaths in Indonesia on top of the rate of unintentional drowning will be important, as they are often left out of studies on drowning worldwide (Sindall et al., 2022).
The prevention of drowning encompasses a wide range of measures, ranging from individual-focused approaches, such as swimming training programmes, to community-based actions, such as community participation in controlling access to open water bodies, policy development on water safety regulations and providing equitable access to safe drinking water (Crawford et al., 2014;Leavy et al., 2015Leavy et al., , 2016Leavy et al., , 2023)).This emphasizes the urgent need for cohesive approaches across the spectrum of upstream, midstream, and downstream drowning prevention (Linnan et al., 2012(Linnan et al., , 2013;;World Health Organization, 2014;Guevarra et al., 2015;Leavy et al., 2023).
Drowning prevention is closely linked to health and safety promotion (Plitponkarnpim and Chinapa, 2014).By connecting drowning prevention and health promotion, a broader understanding of drowning prevention can be achieved, which includes the process of empowering individuals and communities in taking control over their own health-related behaviours and practices (Leavy et al., 2016(Leavy et al., , 2023;;Giles et al., 2017; World Health Organization Regional Office for the Eastern Meditteranean, 2017).Talbot and Verrinder (2017) illustrated the concepts of health promotion in the Health Promotion Framework, which comprises of (i) medical approaches, focusing on individual risk assessment and health information; (ii) behavioural approaches, focusing on health education, skills development and social marketing; and (iii) socio-environmental approaches, focusing on increasing cross-sector partnerships and community capacity, including community action, community participation, structural change, policy development and review and economic and regulatory activities.This framework will be used throughout this review to assess gaps in drowning prevention approaches in • By upholding concepts of health promotion, this review may inform the international health promotion community on potential challenges of drowning prevention efforts in resource-limited settings.• The over-reliance on educational interventions aimed at individuals and the limited availability of population-focused drowning prevention measures is observed in Indonesia.• The research area of community participation in creating safe environment and evidence-informed water safety and drowning prevention-related policy development is relatively neglected in Indonesia.• A demand for further research focused on policy formulation, implementation and evaluation to prevent drowning across lowand middle-income countries is apparent.
Scoping Reviews (PRISMA-ScR) (Arksey and O'Malley, 2005;Tricco et al., 2018;Joanna Briggs Institute, 2020;McGowan et al., 2020;Rethlefsen et al., 2021).Scoping review methodology was selected to explore characteristics of unintentional drowning in Indonesia, a little studied area of interest, and to map evidence from both research and non-research sources, providing a comprehensive overview on unintentional drowning in Indonesia and knowledge gaps for subsequent evidence syntheses (Joanna Briggs Institute, 2020).
Search strategy  (Andry and Yuliani, 2014;Hapsari and Zenurianto, 2016;Buchori et al., 2018;Handayani et al., 2019;Isa et al., 2019;Indonesian Ministry of Health, 2020;Mantong et al., 2020).The searches were undertaken in English and Indonesian language, using the most relevant and exhaustive search terms, in accordance with each database utilized (for details on search terms, see Supplementary Appendix 1).water safety and drowning prevention.The exclusion criteria excluded all publications on tsunami-related events, intentional drowning (intentional self-harm by drowning, assault by drowning, drowning by undetermined intent), and drowning in war operations.
Unintentional drowning deaths screened in this review included accidental drowning deaths, disaster-related drowning deaths and water transport-related drowning deaths.This inclusion was based on the ICD 10 coding for non-intentional drowning, which includes accidental drowning (W65-W74), disaster-related drowning (X34-X39), and water transport-related drowning (V90.-,V92.-); while excluding intentional self-harm by drowning (X71), assault by drowning (X92), drowning by undetermined intent (Y21) and drowned in war operations NOS (Y36.4) (World Health Organization, 2019).Studies on the epidemiology of unintentional drowning were included in this review if they reported the epidemiological measures of drowning mortality as a count, proportion or rate.When calculable, mortality rates were inferred from drowning death counts reported in the reviewed studies.Drowning deaths involving a foreign national that occurred in Indonesia were also included.Conversely, instances where Indonesian citizens drowned abroad were excluded.
Despite being disaster-related, tsunami-related drowning deaths were excluded, due to its relatively rare occurrence (only 75 tsunami events, out of 41 470 disaster events, were documented in Indonesia between 1815 and 2023 (Indonesian National Disaster Management Agency., 2023a)), and the low chance of victim survival, even for good swimmers, due to the sheer force of tsunami vortices (Kurisu et al., 2018), underlining the fact that risk factors and prevention of tsunami-related drowning are entirely distinct from those of other causes (Doocy et al., 2007(Doocy et al., , 2009)).Moreover, even though our database searches included papers on boating, shipping and maritime safety and disaster risk management in Indonesia, all studies identified were on regulatory activities and community participation around maritime safety and flood mitigation not specifically linked nor relevant to the efforts of preventing drowning, hence they were excluded.However, it is acknowledged that disaster risk management, including tsunami risk assessment and preparedness, boating, shipping and maritime safety and their link to drowning prevention, are an area of concern urgently needed to be further investigated in Indonesia (Muis et al., 2015;Djalante and Garschagen, 2017;Sari and Prayoga, 2018;Mantong et al., 2020).

Review strategy
Publications were screened for inclusion by title, abstract and full text.Two researchers (M.C. and D.P.S.) independently reviewed search results at the stage of title, abstract and full-text review.The flow of the selection process of all identified records was as follows:

Title and abstract screening
The titles and abstracts of all records identified through the database searching were screened by the inclusion and exclusion criteria, to ensure the relevance of the studies included for the evidence-informed review.

Full-text screening
Full-text versions of identified articles were then appraised to examine the relevance of the finding of the studies to answer the research questions.The PRISMA-ScR flow diagram summarized the review process, as presented in Figure 1.

Data abstraction
The following data were abstracted from the original peer-reviewed publications: authors, year of study and publication, type of publication (original research article, comprehensive scientific review, meta-analysis, statement of clinical standards, case report, opinion piece or grey literature), data source, study aim, study design, study sample and setting, scale of study (national or subnational), categories of drowning investigated, methods, relevant findings on the mortality rate, risk factors, and prevention of drowning, intervention type and comparator (if applicable).For grey literature, the following data were abstracted: authors (government agency), year of publication, scale of study, categories of drowning investigated and relevant findings.Details of data abstracted from reviewed studies are shown in the online Supplementary Appendix 2.

Synthesis/analysis
Mortality numbers, proportions and rates were extracted or inferred from the identified studies as epidemiological measures of unintentional drowning, either at the subnational or national scale.For this review, the Health Promotion Framework (Talbot and Verrinder, 2017), which comprises medical, behavioural and socio-environmental approaches at the individual through to the population level, was used to map the socio-ecological dimension of drowning prevention and health promotion approaches that have been or are currently being used in Indonesia.

Ethics approval
Ethical approval has been obtained from the Human Research Ethics Committee of the University of Mataram-Indonesia (Ethics Approval number 128/ UN18.F8/ETIK/2023).The scoping review did not collect personal, sensitive or confidential information from participants, and only used publicly accessible documents as evidence.

Numbers and rates of unintentional drowning death in Indonesia
Accidental drowning deaths Three descriptive observational studies (Usaputro and Yulianti, 2013;Wulur, 2013;Astreani and Alit, 2015) reported accidental drowning deaths at the subnational level, all outlining the epidemiological measures of drowning mortality as a count.The only data source for drowning deaths identified in these studies was medico-legal/autopsy records.Two of these studies (Usaputro and Yulianti, 2013;Astreani and Alit, 2015) reported drowning mortality at the provincial scale of Bali and included drowning deaths involving Indonesians and foreign nationals.Both studies reported drowning deaths documented by Sanglah Provincial Hospital in the province of Bali, with a total of 209 drowning deaths documented between 2010 and 2014, from which an average annual drowning mortality rate of 1.73/100 000 between 2010 and 2014 can be inferred.Meanwhile, one study (Wulur, 2013) reported 15 drowning deaths documented by the Forensic Department of Prof. Dr. R. D. Kandou Provincial Hospital of North Sulawesi between 2007 and 2011, from which an average annual drowning mortality of 0.18/100 000 can be inferred for the provincial level of North Sulawesi between 2007 and 2011.However, it is important to note that all drowning data reported in these studies were collected from provincial referral hospitals and may miss some incidents.

Disaster-related drowning deaths
Two papers (Hu et al., 2018;Indonesian National Disaster Management Agency, 2023) in this review discussed disaster-related deaths.The Indonesian National Disaster Management Agency (2023) reported the frequency of disasters and disaster-related mortalities and missing cases in Indonesia between 1815 and 2023: (i) 13 927 flooding events, with 22 476 deaths and 8195 missing victims; (ii) 9503 flooding and avalanche events, with 3324 deaths and 379 missing victims; (iii) 499 tidal wave-related disaster events, with 165 deaths and 49 missing victims and (iv) 11 225 cyclone events, with 479 deaths and 49 missing victims.These hydrometeorological disasters, excluding tsunami, contributed to 36.7% (N = 100 434/259 407) of disaster-related deaths and 17% (N = 8672/51 037) of all disaster-related missing victims in Indonesia, from which disaster-related rate of 186 deaths per 100 000 disaster-affected populations can be inferred.
In a study using the Emergency Disasters Database (EM-DAT) and the Dartmouth Flood Observatory data to analyse floods and flood-induced mortality across the globe between 1975 and 2016, Hu et al. (2018) reported Indonesia among countries with the highest flood frequency and flood-induced mortality worldwide, estimating 5000 to 10 000 flood-induced deaths occurred in Indonesia between 1975 and 2016, with tropical cyclones contributing to 0-10% of these deaths, from which flood-induced rates of 1-3 deaths per 100 000 flood-affected populations can be inferred.However, it is important to note that the exact cause of the reported hydrometeorological disaster-related deaths, either by drowning or other causes, were not stated in these two studies, and that data for the period before 1990 may be severely underestimated.There was also no information available on exact data sources and how at-risk populations were estimated, hence the possibility of selection and measurement/ information bias was identified in these studies, resulting in the difficulty of inferring cause-specific or proportionate mortality rates.

Water transport-related drowning deaths
Three studies reported water transport-related deaths in Indonesia (Suwardjo et al., 2010;Saputra, 2021;Indonesian National Disaster Management Agency, 2023).The Indonesian National Disaster Management Agency (2023) reported that a total of 15 shipping/boating accidents occurred across Indonesia between 2014 and 2017, resulting in 121 deaths and 97 missing victims.Saputra (2021) reported on the completed investigations of 120 shipping accidents by the Indonesian National Transportation Safety Committee between 2003 and 2019, and found they contributed to 513 deaths and 701 missing victims across Indonesia.At the subnational level, Suwardjo et al. (2010) reported that 61 fishing vessel accidents occurred around the three subdistrict-level fishing ports of Central Java Province between 2006 and 2008, contributing to a total of 68 deaths, or an annual average of 32 dead/missing fishing crews, and an average fatality accident rate for all three study sites of 115 deaths per 100 000 fishermen, with 26.5% (n = 18/68) of victims falling off the ships into the water during shipping or fishing, and 45.6% (n = 31/68) of victims died due to the ship capsizing.However, it is important to note that the exact cause of the reported water transport-related deaths were not stated in these studies, and that these reports only included cases where the Indonesian Disaster Management Agency was involved in the rescue/evacuation process, or investigated by the Indonesian National Transportation Safety Committee, or reported by the authorities within the study site and period, hence it may miss some incidents and make inferring cause-specific or proportionate mortality rates difficult.
Several factors were identified as potential risk factors of for unintentional drowning in Indonesia: (i) sociodemographic characteristics, including age and sex (males aged 18 and over disproportionately contributed to the majority of drowning death cases in Bali and North Sulawesi (Usaputro and Yulianti, 2013;Wulur, 2013;Astreani and Alit, 2015)), nationality (foreign nationals and Indonesians made up for almost similar proportions of drowning deaths in Bali (Usaputro and Yulianti, 2013;Astreani and Alit, 2015), noting the high population of foreign nationals in the major tourist destination Bali may differ from populations of other provinces), population density (population density had a significantly positive correlation with the number of flood-related deaths (Hu et al., 2018)), and income level (flood-induced mortality increased with the decrease of per capita GDP (Hu et al., 2018)); (ii) environmental factors, including aquatic location of drowning deaths (most accidental and water transport-related drowning cases occurred in open seawater (Usaputro and Yulianti, 2013;Astreani and Alit, 2015; Indonesian National Disaster Management Agency, 2023)), seasonality (highest numbers of fishing vessel accidents were recorded during rainy seasons (Suwardjo et al., 2010)), geographical and environmental conditions (a third of boating/shipping accidents occurred in Indonesia were weather-related (Indonesian National Disaster Management Agency, 2023), and most flooding inundation and flood-induced deaths occurred in low-lying regions with dense river systems (Hu et al., 2018)), and hydrometeorological disasters (the frequency of floods and flood-induced mortality were generally increasing, with a large proportion of flood-induced deaths attributed to tropical cyclone-induced flash floods (Hu et al., 2018)); (iii) risky behaviour (alcohol identified on a fifth of autopsied drowning victims, although there was no information on the blood alcohol content (Usaputro and Yulianti, 2013)); (iv) low knowledge and skills on water safety and water rescue, particularly knowledge on first aid for drowning victims (Prasetyo, 2017;Widyastuti and Rustini, 2017;Elsi and Gusti, 2020;Welembuntu et al., 2021) and (v) low knowledge and compliance for boating, shipping, and maritime safety-related regulations, including poor maintenance of ships, lack of safety equipment on board, poor knowledge and compliance of safety regulations and underqualified seafarers and poor ship crews' capacity in ensuring safe boating/maritime practice (Suwardjo et al., 2010;Saputra, 2021;Indonesian National Disaster Management Agency, 2023), as summarized in Table 1.
However, it is important to note that despite the importance of providing an understanding of risk factors for drowning across Indonesia, no relative risk (RR) or odd ratio (OR) was reported across all studies reviewed, and no studies reported statistical correlations of risk factors of interest in association to fatal unintentional drowning incidents.All information on unintentional drowning risk factors was presented as proportions/counts of drowning deaths between categories of risk factors (Table 1).
Only one grey literature source (Indonesian Ministry of Health, 2020) reported on regulatory activities and social marketing approaches for drowning prevention.Meanwhile, two studies (Rosmi et al., 2020;Fernalia et al., 2022) outlined the potential of training local community groups to perform and assist on local water rescue efforts (Table 2).

DISCUSSION
This scoping review aimed to describe the epidemiology of and risk factors for unintentional drowning in Indonesia and explore existing health promotion and prevention approaches currently in place.Despite a  (Wulur, 2013).
• No measures of association were reported, and a possibility of selection bias was identified in all three studies above (Usaputro and Yulianti, 2013;Wulur, 2013;Astreani and Alit, 2015).
• No measures of association were reported, and a possibility of selection and measurement/information bias was identified in all three studies above (Usaputro and Yulianti, 2013;Wulur, 2013;Astreani and Alit, 2015).
• No measures of association were reported, and a possibility of selection and measurement/information bias was identified in the two studies above (Usaputro and Yulianti, 2013;Astreani and Alit, 2015).
Population density • Significant positive correlation between population density and the number of flood-related deaths.A possibility of selection bias was identified (Hu et al., 2018).
Income level • The flood-affected population and flood-induced mortality increased with the decrease of per capita GDP.A possibility of selection bias was identified (Hu et al., 2018).
• No measures of association were reported, and a possibility of selection and measurement/information bias was identified in all three studies (Usaputro and Yulianti, 2013;Astreani and Alit, 2015;Indonesian National Disaster Management Agency, 2023).

Seasonality
• Of 61 fishing vessel accidents reported to occur around three fishing ports of Central Java Province between 2006 and 2008, the highest numbers of fishing vessel accidents were recorded during rainy seasons (November to February).No measures of association were reported.
A possibility of selection bias was identified (Suwardjo et al., 2010).
Geographical and environmental conditions • Weather-related: 33.33% (n = 5/15) of ship accidents in Indonesia between 2014 and 2017.No measures of association were reported.A possibility of selection bias and measurement/information bias was identified (Indonesian National Disaster Management Agency, 2023).
• Floodings and flood-induced deaths most often occurred in low-lying regions with dense river systems.Plain areas with slope of <0.5°: 40% of flood events and 50% of flood-induced deaths worldwide; plain areas with slopes 0.5°-15°: 60% of flood events and 48% of floodinduced deaths worldwide.No measures of association were reported.A possibility of selection bias was identified (Hu et al., 2018).

Study findings
Hydrometeorological disasters • Between 1975 and 2016, the frequency of floods and flood-induced mortality were generally increasing globally, including in Indonesia.
• The annual variation of mortality per flood event was highly related to floods with higher intensity, with the highest flood frequency and flood-induced mortality worldwide reported to occur in Asia, particularly in China, India, Indonesia and the Philippines.
• A large proportion of flood-induced deaths and the highest flood-induced mortality rate worldwide can be attributed to tropical cycloneinduced flash floods.
• No measures of association were reported.A possibility of selection bias was identified in this study (Hu et al., 2018).
Risky behaviours Alcohol consumption • Blood alcohol was identified in 20% (n = 4/20) of autopsied drowning victims at Sanglah Provincial Hospital of Bali between 2010 and 2012.No information on the blood alcohol content.No measures of association were reported (Usaputro and Yulianti, 2013).

Biological factors
Underlying medical conditions • Comorbid conditions were recorded on 15% (n = 3/20) of drowning victims underwent autopsy at Sanglah Provincial Hospital of Bali between 2010 and 2012.No information on comorbid conditions.No measures of association were reported (Usaputro and Yulianti, 2013).

Knowledge and skills on water safety and water rescue
Knowledge on first aid for drowning victims • Significant correlation between fishermen's knowledge on Basic Life Support (BLS) and attitude to BLS being given for drowning victims (p < 0.05) (Elsi and Gusti, 2020).
• No measures of association were reported, and a possibility of selection and measurement/information bias was identified in two studies above (Indonesian National Disaster Management Agency, 2023; Saputra, 2021).
• Contributing factors to fishing vessel accidents in Central Java Province between 2006 and 2008: (i) underqualified shipping crews; (ii) poor knowledge, awareness, and compliance of safety regulations; (iii) unfulfillment of safety requirements.Approximately 84.3% of skippers and ship crews did not go beyond primary level education, hence did not qualify to undertake Basic Safety Training for shipping crews.Seventy per cent (n = 45/64) of all fishing vessels registered did not fulfil safety requirements due to insufficient number of life jackets and rescue buoys, unequipped with fire extinguishers and life rafts, and lacking in other safety equipment (Suwardjo et al., 2010).
• No measures of association were reported, and a possibility of selection and measurement/information bias was identified in all three studies above in all three studies above (Suwardjo et al., 2010;Saputra, 2021;Indonesian National Disaster Management Agency, 2023).• No information on long-term knowledge and self-efficacy retention.
• A possibility of selection bias and measurement bias was identified.
Ose • No information on long-term knowledge retention.
• A possibility of selection bias and measurement bias was identified.
Patimah et al. • No measures of association were reported.
• A possibility of selection bias and measurement bias was identified.• No evaluation on participants' level of knowledge and skills on flooding-related water rescue.
• No measures of association were reported.
• A possibility of selection bias and measurement bias was identified.Drowning mortality data across Indonesia: the disparity on data availability Comprehensive data are imperative for agenda setting (World Health Organization, 2017).However, this review outlined the limited availability of drowning data across Indonesia.The data sources for drowning deaths identified in this review were derived from medico-legal autopsy records, records of investigated shipping accidents and rescue reports, with no drowning data derived from health and demographic surveillance data, national death registry, integrated coronial information system, police department records, national health reports or news reports.This is due to the unavailability of a coordinated national drowning data collection system and national death registry in Indonesia, from which national and subnational data for all categories of unintentional drowning can be recorded (World Health Organization, 2021).This unavailability underlines the possibility of under-representation of drowning data and a public health issue regarding the health system capacity on collecting, reporting and utilizing health data.
A similar reliance on medico-legal reports has been reported in other LMICs, such as in South Africa, where the dependence on hospital-based reports means that the surveillance system may miss drowning-related injuries which occur in and around homes, particularly in rural areas with inadequate access to healthcare facilities (Matzopoulos, 2002;Donson and Van Niekerk, 2013;Morris et al., 2016).The underreporting of drowning incidents can be linked to the nature of drowning deaths, where victims often suffer a quick death on location and never reach medical facilities or are reported to authorities, which may impede accurate data collection in countries with dependence on facility-based reporting systems, such as in Indonesia (Linnan et al., 2012(Linnan et al., , 2013;;World Health Organization, 2014).These limitations underline the urgent need to strengthen public health system capacity; develop standardized, national health, death and coronial data reporting frameworks; enhance multi-sectoral collaboration and advocate for political and financial investment, to develop a robust drowning data collection system in Indonesia (Schuurman et al., 2011;Reynolds et al., 2013;Scarr and Jagnoor, 2022).

Drowning risk factors in Indonesia: a knowledge gap
This review identified a lack of exploration on predictors for unintentional drowning fatalities, with no studies reporting RR/OR or statistical correlations of predictors of interest and fatal unintentional drowning incidents.This may hinder the development of prevention interventions needed across different provinces of Indonesia, which may be different to prevention interventions which have been shown to be effective in high income countries (Bennett and Linnan, 2014;Franklin and Scarr, 2014;Hyder et al., 2014;Linnan et al., 2014).
Increasing supervision of children, raising community awareness and skills on water safety, and encouraging community participation have been identified as important protective factors for drowning in LMICs (Rahman et al., 2010(Rahman et al., , 2012;;Cenderadewi et al., 2020).As advocated by the WHO, the effectiveness of community driven crèches in preventing child drowning for children aged 12-59 months, and swimming lessons in reducing the risk of drowning for children aged 6 and above, along with their cost effectiveness, have been widely reported in rural settings in Bangladesh (Rahman et al., 2012;Talab et al., 2016;Alonge et al., 2020;Alfonso et al., 2021;Rahman and Rahman, 2022).Further scaling up of these interventions to other LMICs, including to Indonesia, should be considered, as part of the global effort to reduce the global burden of drowning, particularly among children (Rahman et al., 2012).Further research is required to evaluate the process, impact and outcome of the interventions against key indicators in Indonesia; improving local capacity and understanding in preventing drowning in the country (Guevarra et al., 2021).
Further research on contributing factors to disaster-related and water transport-related drowning deaths is also imperative to inform the development of drowning prevention approaches suitable for Indonesian context, given the high frequency of hydrometeorological disasters and water transport-related incidents and the nation's high vulnerability to the impacts of disasters and climate change (Suwardjo et al., 2010;Muis et al., 2015;Djalante and Garschagen, 2017;Hu et al., 2018;Sari and Prayoga, 2018;Mantong et al., 2020;Saputra, 2021;Indonesian National Disaster Management Agency, 2023a, 2023b).

Understanding the linkage between drowning prevention and health promotion
Using the Health Promotion Framework (Talbot and Verrinder, 2017) to analyse the socio-ecological approaches utilized in the prevention of drowning in Indonesia, this review identified the under-exploration of the concept of the socio-ecological approach of health promotion related to drowning prevention in Indonesia, leaving the research area of community participation and development of evidence-informed policy around water safety relatively neglected.The limited availability of population-focused midstream and upstream drowning prevention interventions in Indonesia does not align with the underlying premise of the Health Promotion Framework, which supports the need for multisectoral, multi-strategic approaches including educational, behavioural, socio-environmental and regulatory approaches to ensure effective individual and community-level injury prevention (Stempski et al., 2015;Denehy et al., 2016;Leavy et al., 2016;Giles et al., 2017; World Health Organization Regional Office for the Eastern Meditteranean, 2017).A comprehensive approach to drowning prevention in Indonesia requires approaches that go across the downstream, midstream, upstream continuum and include strengthening individual and broader community level knowledge and skills, fostering coalitions and networks, changing organizational practices and policy and legislation setting (Franklin and Scarr, 2014).

Policy development and practice
A national, evidence-informed regulatory framework for drowning prevention, guided by the WHO implementation guide on preventing drowning (World Health Organization, 2017) and affirmed by the 2021 United Nations General Assembly Resolution on global drowning prevention (United Nations, 2021), is required to reduce drowning fatalities in Indonesia.To inform this, a situational assessment must be performed in Indonesia, to: (i) review available data on drowning; (ii) assess current efforts regarding drowning prevention including existing policy and regulation; (iii) identify key stakeholders who play a role in drowning prevention and (iv) assess required human and financial resources (World Health Organization, 2017).This approach is supported by the WHO, who also recommends advancing drowning prevention through robust data collection, to inform burden and risk factor identification, and agenda setting, as well as development and evaluation of regulatory frameworks and prevention interventions (World Health Organization, 2017).

Contributing to situational assessment on drowning prevention in Indonesia
Several findings of this review may inform the situational assessment on drowning prevention in Indonesia.Furthermore, although not included in this study, several studies identified during the database searches may provide baseline information around disaster risk management and enforcement of maritime, shipping and boating safety in Indonesia, informing the urgently needed multisectoral collaboration to develop coordinated national drowning prevention efforts:  (Muis et al., 2015;Hapsari and Zenurianto, 2016;Buchori et al., 2018;Handayani et al., 2019;Isa et al., 2019).4 Main actors on maritime and boating safety enforcement include the Ministry of Transportation as the lead agency, harbourmasters and Sea and Coast Guard.Low human resource capacity, low compliance to regulations, overlapping regulatory frameworks with unclear responsibilities and coordination mechanisms between agencies and the geographic nature of Indonesia as a vast archipelagic nation were identified as challenges in enforcing maritime and boating safety nationwide (Andry and Yuliani, 2014;Mantong et al., 2020).5 A continuous effort to systemically integrate drowning prevention framework across regulatory activities enforcing water safety, adequate supply of clean water, maritime/boating safety, occupational health and safety, disaster risk management and sustainable spatial and non-spatial development planning, emphasizing the importance of disaster risk reduction and climate change adaptation to the efforts of reducing fatalities, with clear obligations, responsibilities and coordination mechanisms between agencies, is required in Indonesia.

Future research
More research on drowning is needed in Indonesia across all domains, however several priorities for future research were noted, with a particular focus on developing and improving the performance of drowning surveillance to ensure the availability and reliability of drowning mortality and morbidity data across Indonesia.Studies investigating the measurement of association between accidental, disaster-and water transport-related drowning risk factors, and the incidence of fatal and non-fatal unintentional drowning incidents are essential to inform the appropriate prevention measures for Indonesia.Further research on drowning cases related to disasters, disabilities, genders, older populations and children is also vital to provide a better understanding of the at-risk populations in the country (Pearn and Franklin, 2013;Franklin et al., 2014;Peden and Franklin, 2019;Clemens et al., 2021;Roberts et al., 2021).
In addition, further investigation on broader health promotion approaches that reflect a socio-ecological approach to drowning prevention is imperative for Indonesian context.It is particularly important to explore the potential vital link between unintentional drowning prevention and safe water provision, boating and maritime safety regulations and enforcement, occupational safety and health, rural development, disaster risk management and efforts to bridge economic inequities and disparities in accessing healthcare across different populations, particularly impacting the socioeconomically disadvantaged populations of rural and remote areas of archipelagic Indonesia (Suwardjo et al., 2010;Muis et al., 2015;Djalante and Garschagen, 2017;Hu et al., 2018;Sari and Prayoga, 2018;Mantong et al., 2020;Saputra, 2021;Indonesian National Disaster Management Agency, 2023a, 2023b).

Strengths and limitations
Several factors contributed to the strength of this review, including the inclusion of studies published in Indonesian language and the use of broad search terms to capture the extensive causes of drowning and the wide areas where drowning may occur.However, several limitations were also identified.Firstly, it is possible that not all studies exploring unintentional drowning in Indonesia were located.In addition, the lack of information on accidental, disaster-related and water transport-related drowning mortality numbers, cause of deaths, data sources and at-risk populations, along with the inconsistency in data collection and reporting between articles, may compromise the accuracy and quality of drowning data identified, limiting the generalizability and mortality rate inference from the findings, potentially resulting in the underestimation of the magnitude of unintentional drowning mortality across Indonesia.

CONCLUSION
Limited publications on drowning rates, risk factors and prevention were observed within Indonesia.The unavailability of a coordinated national drowning data collection system in Indonesia underlines the possibility of under-representation of drowning mortality.The under-investigated measurement of association between various exposures and fatal drowning incidents was identified, potentially hindering the development of prevention interventions.The over-reliance on individual-focused prevention measures was observed in Indonesia, with an apparent under-development of socio-ecological health promotion approaches to drowning prevention.Several highlights for future research were noted, with a particular focus on improving the performance of drowning surveillance, to ensure the availability and reliability of drowning mortality and morbidity data across Indonesia.Broader health promotion approaches that reflect a socio-ecological approach to drowning prevention in Indonesia is also imperative in reducing drowning incidents and fatalities in Indonesia.

FUNDING
Open access publishing facilitated by James Cook University, as part of the Oxford University Press -James Cook University Read and Publish Agreement negotiated by the Council of Australian University Librarians.

Table 2 :
Analysis of prevention interventions investigated in Indonesia using the Health Promotion Framework Author

Table 2 .
Continueddire need to provide better understanding on the magnitude of drowning as a health problem in Indonesia, limited literature reporting on drowning epidemiology, risk factors or prevention and health promotion approaches in Indonesia were identified.