Quantifying rip current-related drowning deaths and exposure on Australian beaches

Abstract

Background

Rip currents are the leading coastal hazard on Australian beaches and a significant contributor to coastal drowning. This comprehensive study explored the long-term epidemiology and exposure-based mortality rates of Australian rip current drowning deaths, with the purpose of informing rip-related education efforts to reduce coastal drowning risk.

Methods

A retrospective descriptive analysis of rip current drowning deaths between 1 July 2004 and 30 June 2023 was conducted, and rip current involvement was determined using data triangulation methods (incident narrative, beach imagery and expert opinion). Temporal trends were examined using Joinpoint regression, while Australian population and coastal participation data were used to calculate mortality rates.

Results

407 rip current drowning deaths were identified (21 deaths/year); an exposure-based mortality rate of 0.11/1 million coastal visits. Incidents predominantly involved males (85%), 20–34 years (38%), occurred in regional/remote areas (59%) and in the presence of others (80%). For every rip current drowning death, an estimated 2449 people were rescued by someone else and 8171 individuals self-rescued.

Discussion

Despite small reductions in the rate of rip current drowning among males, little overall progress has been seen in attempts to decrease the mortality burden. Future efforts will require an expanded approach, beyond traditional means.

Conclusions

Rip currents remain a significant public health threat and the number one coastal hazard contributing to preventable mortality along Australia’s coastline.

Implications for public health: Improved rip current education approaches, development and evaluation are needed to direct effective messaging to at-risk groups.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Rip currents represent Australia’s number one coastal hazard, on average being a causal factor in 20% of the country’s coastal drowning deaths each year.

WHAT THIS STUDY ADDS

• This study addresses previously identified research gaps around exposure for rip current drowning deaths in Australia by calculating exposure-adjusted person-time data to accurately calculate drowning rates and identify those most at risk of rip drowning.

• This study reaffirms rips as a major coastal hazard in Australia, with rates in rip current drowning remaining unchanged over the 20-year study period.

• Young males represent a high-risk demographic in rip current drowning deaths, with unpatrolled locations and times being an area of high concern.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Lifesaving personnel represent a key asset in coastal safety infrastructure; however, the evaluation and implementation of a broader range of preventative and response measures, including improved evidence-based education and intervention methods, are needed to aid in reducing rip current drowning at unpatrolled locations.

Background

Rip currents are the leading coastal hazard on beaches globally¹ and are a significant contributor to coastal drowning.^2–7 Rip currents, herein referred to as ‘rips’, are narrow, often channelised flows of water that travel offshore from close to the shoreline, through the surf zone and varying distances beyond.¹ They represent a significant risk to beachgoers, with substantial efforts throughout the world to increase community awareness and understanding of the rip hazard to reduce rip-related drowning and other fatalities.⁸

In Australia, more than 120 people die from unintentional coastal drowning each year, and thousands more are rescued by lifesaving personnel and bystanders.^{9 10} Australia has an extensive coastline totalling more than 12 000 km, with an estimated 17 000 rips present on any given day.^{11 12} On average, 20% of coastal drowning deaths in Australia each year are due to rips,¹³ making them Australia’s dominant coastal hazard, typically accounting for more deaths annually than sharks, floods and cyclones combined.¹⁴

Given the prevalence of rip drowning, a growing body of beach safety research has explored the epidemiology of rip mortality^2–7 and experiences of those who get caught in rips.^{15 16} Studies have also examined the attitudes, behaviours and risk perceptions of rips within the general population¹³ and how these may differ for survivors of rip current incidents.^{15 16} For many, getting unintentionally caught in a rip can be traumatic and can significantly impact subsequent behaviour.^{15 16} Relationships exist between rips, unpatrolled beaches (without lifesaving personnel), swimming and bystander rescue drowning,^{4 10 13 17} which would benefit from further research.

To date, Brighton et al is the most detailed epidemiological study of rip drowning mortality and rescue burden in Australia, with authors highlighting the need for exposure-adjusted person-time data to accurately calculate drowning rates and identify those most at risk of rip drowning.⁴ As such, this study represents an updated epidemiological analysis of rip drowning, including addressing previously identified research gaps around exposure, to help inform rip-related education and reduce coastal drowning risk in Australia.

Methods

We performed a retrospective descriptive epidemiological analysis of rip drowning deaths that occurred along the Australian coastline between 1 July 2004 and 30 June 2023. Specifically, we (1) described person, time and event characteristics; (2) analysed trends in time series data; (3) calculated cumulative population and exposure-based mortality rates of rip drowning deaths and (4) estimated the ratio of rip drowning deaths to survivals by method of escape.

Data sources

Mortality data were obtained from Surf Life Saving Australia’s (SLSA) National Coastal Fatality Database—a repository recording all fatal incidents that have occurred along the Australian coastline since 2004 with data obtained from multiple sources, including the National Coronial Information System (an electronic database of all deaths notified to coroners within Australia and New Zealand), with supplementary incident details obtained via media and internal incident reports. The database includes variables previously detailed.^{7 18 19}

Data relating to estimates of the Australian population by age, sex, remoteness classification and all-cause mortality were obtained from the Australian Bureau of Statistics.^{20 21} Estimates for exposed populations, that is, the number of coastal visitors, activity participants and those who have previously been caught in a rip were derived from SLSA’s National Coastal Safety Survey (NCSS), the method for which has previously been described.²² This study used NCSS survey responses collected in the years 2014–2023 (n=19 787).

Data variables and inclusion

All unintentional rip drowning deaths that occurred in Australian coastal waters (ie, including up to three nautical miles offshore) between 1 July 2004 and 30 June 2023, were collated for analysis. Both open (ie, under coronial investigation) and closed coronial cases were included, although at the time of analyses, 85% were closed (n=344). Open case data are correct as of 10 October 2023.

Rip presence and rip attribution to drowning incidents were identified via narratives reported within coronial, police, autopsy, incident and media reports. Rip involvement was further determined on a case-by-case approach, using incident proximate details and narratives, reference to Google Maps and images of beaches for rip presence, and expert geomorphological analysis by author RB. Fatalities caused by boundary rips that form alongside structures such as headlands, rock platforms, groynes and jetties¹ were included. Incidents for which rip involvement was unclear (n=3) or were suspected to be due to a different type of current, such as tidal or longshore, were excluded (n=49).

To simplify interpretation of results, rip deaths from causes other than drowning (ie, medical episodes) were excluded from the dataset (n=14), as were incidents that occurred on offshore Australian territories (n=3).

Statistical analyses

Analyses of rip drowning deaths by person and event characteristics were performed in R (V.4.2.3). We calculated four risk measures, the unintentional fatal rip-related incident rate for (1) the total population, per 100 000 persons; (2) the total exposed population, per 100 000 coastal visitors; (3) exposed person-time, per 1 million coastal visits and (4) those who have ever been unintentionally caught in a rip over their lifetime, per 100 000 persons. Rate calculations were performed using the epiR package, which employs Ulm’s exact method for confidence intervals. Australian residents, students and those on working holiday visas were included in numerator values in rate calculations (n=339). Short-term visitors, asylum seekers and those whose residency status could not be determined were excluded as these populations are unable to be accounted for in national population estimates (n=68).

Joinpoint regression was performed to analyse trends in crude annual rip drowning rates by total population, sex, age and remoteness classification using the National Cancer Institute Joinpoint Trend Analysis Software.²³ A log-linear model was used to determine the average annual percentage change (AAPC), with an alpha level of 0.05 used for model selection permutation testing. Drowning and population data were aggregated across 2-year time periods to overcome the challenge of zeros in the numerator values that were present in the data when stratified by different population cohorts. Numerators were determined as the number of rip drowning deaths occurring within a 2-year period from the beginning of the year one financial year to the end of the year two financial year. Denominators were determined as the sum of the estimated annual population at the beginning of each financial year for the 2-year period relative to the numerator value. This method of aggregating data across years has been determined as a valid approach in Joinpoint regression analysis when seeking to overcome the presence of zeros in the dependent variable.²⁴ As such, the calculated AAPC is reflective of 2-year periods and is interpreted accordingly in the results.

To measure the impact of coastal exposure on rip drowning deaths, we calculated and compared population, exposed-person and person-time-based rates and rate ratios occurring between 1 April 2013 and 31 March 2023. Denominators for population-based estimates were derived from 2016 and 2021 census data,²⁰ with cases occurring between 2013/2014 and 2018/2019 financial years being matched to 2016 census data, with those occurring between 2019/2020 and 2022/2023 financial years matched to 2021 data. The rate denominator for exposed population and person-time-based estimates was derived from the NCSS, the method for which has previously been described in detail.⁷ The time period to be included in these analyses was restricted to align with available exposure data derived from the NCSS.²² Rates were not calculated for those aged 0–15 as exposure estimates for this age group were unable to be obtained from NCSS data.

To explore the ratio of rip drowning deaths to survivals (ie, people successfully escaping from rips), we used responses from the NCSS2023 to estimate the number of Australian residents caught in a rip and their method of escape for the 10-year period of 1 April 2013 to 31 March 2023. We used responses to the questions ‘Have you ever been caught in a rip current unintentionally?’ (Responses: ‘Yes’, ‘No’) and ‘Thinking about the last time you were caught in a rip current, when was it’ (Responses: ‘In 2023’, ‘last year (2022)’, ‘2–5 years ago’, ‘6–10 years ago’, ‘more than 10 years ago’, ‘Can’t say/Can’t remember’), which provided an estimate of the number of Australians who had been caught in a rip within the previous 10 years. We then used responses to ‘How did you get out of the rip current’ to estimate the proportion of each rip escape method used. An escape ratio in relation to drowning deaths was calculated using the number of rip drowning deaths within this date range.

Results

Person and event characteristics

Summary statistics for person and event (rip drowning death) characteristics are summarised in table 1. Of note, 407 rip drowning deaths were recorded during the 19-year study period (mean 21.4 (SD=5.8); figure 1a). Males accounted for 84.8% (n=345) of rip drowning deaths and the greatest proportion were recorded among 20–34 years (37.8%, n=154; figure 1b; table 1). Over half occurred in regional and remote areas (58.8%, n=239). Three in five rip drowning deaths occurred greater than 1 km from a surf lifesaving service (58.7%, n=239) and four in five decedents fatally drowned in the presence of others (80.2%, n=326).

Table 1. Person and event characteristics of rip deaths, Australia, 1 July 2004 to 30 June 2023 (n=407).

Person characteristics
Variable		n	Per cent
Age	0–15	24	5.9
	16–24	83	20.4
	25–34	99	24.3
	35–49	93	22.9
	50–69	85	20.9
	70+	23	5.7
Sex	Female	62	15.2
	Male	345	84.8
Index of Relative Socio-Economic Advantage and Disadvantage (IRSAD)	1–2	74	18.2
	3–4	42	10.3
	5–6	58	14.3
	7–8	79	19.4
	9–10	101	24.8
	Unknown	53	13.0
Remoteness classification of incident location	Major cities of Australia	168	41.3
	Inner regional	164	40.3
	Outer regional	52	12.8
	Remote	21	5.2
	Very remote	<5	0.5
Residents distance to drowning location	< 10 km	95	23.3
	10–50 km	65	16.0
	> 50 km (intrastate)	146	35.9
	> 50 km (interstate)	46	11.3
	International	47	11.5
	Unknown	8	2.0
Event characteristics
Distance to surf life saving service/clubhouse	<1 km	168	41.3
	1–5 km	132	32.4
	> 5 km	107	26.3
State	New South Wales	189	46.4
	Queensland	89	21.9
	Victoria	62	15.2
	Western Australia	31	7.6
	South Australia	20	4.9
	Tasmania	11	2.7
	Northern Territory	5	1.2
Activity	Swimming/wading	293	72.0
	Attempting a rescue	63	15.5
	Watercraft	28	6.9
	Snorkelling	6	1.5
	Fishing	6	1.5
	Fall	<5	0.7
	PWC	<5	0.7
	Rock fishing	<5	0.5
	Jump	<5	0.2
	Other	<5	0.2
	Unknown	<5	0.2
Victim alone	Yes	36	8.8
	No, with friends/family	286	70.3
	No, with organised activity	6	1.5
	No, with strangers	34	8.4
	Unknown	45	11.1
Alcohol contributory	Alcohol	27	6.6
	No alcohol	380	93.4
Drugs contributory	Drugs	40	9.8
	No drugs	367	90.2
Season (Australia)	Winter (June–August)	16	3.9
	Spring (September–November)	56	13.8
	Summer (December–February)	227	55.8
	Autumn (March–May)	108	26.5
Day of the week	Sunday	87	21.4
	Monday	53	13.0
	Tuesday	41	10.1
	Wednesday	44	10.8
	Thursday	47	11.5
	Friday	47	11.5
	Saturday	85	20.9
	Unknown	<5	0.7
Incident time of day	Night (00:00–06:00)	18	4.4
	Morning (6:00–12:00)	64	15.7
	Afternoon (12:00–18:00)	261	64.1
	Evening (18:00–00:00)	55	13.5
	Unknown	9	2.2

Figure 1. Annual trends of rip current drowning deaths (number and rate per 100 000 pop.) 1 July 2004–30 June 2023 (a) and number of rip drowning deaths by age and sex, 1 July 2004–30 June 2023 (b).

Temporal trends

Time series analysis showed no significant difference in overall rip drowning rates over the study period (2-year period average percentage change=−1.9% (95% CI: −4.1% to 0.4%), p=0.089; figure 1; table 2). However, the rate of male rip drowning deaths significantly decreased by an average percentage change of 2.0% per 2-year period (95% CI: −3.9% to −0.1%, p=0.044).

Table 2. Joinpoint trends in rip drowning deaths among the Australian population (1 July 2004–30 June 2023) by age, sex and remoteness.

Variable		2004/2005–2022/2023
		AAPC (%)*	95% CI	P value
Total population		−1.9	(−4.1 to 0.4)	0.089
Sex	Female	−0.8	(−6.3 to 5.1)	0.762
	Male	−2.0	(−3.9 to −0.1)†	0.044
Age	16–24	−2.5	(−5.3 to 0.5)	0.090
	25–34	−3.1	(−7 to 1)	0.115
	35–49	3.0	(−0.9 to 7.1)	0.115
	50+	−3.1	(−6.2 to 0.1)	0.055
Remoteness	Major cities of Australia	−3.6	(−7.3 to 0.2)	0.058
	Regional and remote Australia	0.0	(−2.2 to 2.4)	0.967

Rates are not calculated due to low sample size (n<16).

^*AAPC is representative of 2 fiscal years (1 July–30 June).

^†AAPC significantly different from zero at the alpha=0.05 level.

AAPC, average annual percentage change.

Comparing rip drowning rate calculations

Rate comparisons are described in table 3. Of note, the cumulative population-based rate of rip drowning deaths was 0.09 deaths per 100 000 residents (aged 16+ years; 95% CI: 0.07 to 0.10), which increased to a rate of 0.11 per 100 000 exposed residents (95% CI: 0.09 to 0.13). Exposure-based estimates showed male residents drowned at a rate 5.51 times that of females (0.04 per 1 million coastal visits (95% CI: 0.03 to 0.04) vs 0.01 per 1 million coastal visits (95% CI: 0.00 to 0.01)). By age group, 16–24 yeas recorded the highest rip drowning rates (0.033 per million coastal visits (95% CI: 0.02 to 0.05)). The exposure-based rip drowning rate among residents of regional and remote Australia was 3.05 times greater than that of residents of major cities (0.03 per 1 million coastal visits (95% CI: 0.03 to 0.04) vs 0.01 per 1 million coastal visits (95% CI: 0.01 to 0.01)).

Table 3. Population and exposure-based rates, with 95% CIs, of rip current drowning deaths in Australia (16+) by person and event characteristics, 1 April 2013–31 March 2023.

			Rate per 100 000 pop			Rate per 100 000 exposed pop			Rate per 1 million visits
variable		n	Rate	95% CI	RR	Rate	95% CI	RR	Rate†	95% CI	RR
Total	Total 16+	170	0.09	(0.07 to 0.10)	Ref	0.11	(0.09 to 0.13)	Ref	0.022	(0.02 to 0.03)	Ref
Sex	Female	24	0.02	(0.02 to 0.04)	Ref	0.03	(0.02 to 0.05)	Ref	0.006	(0.00 to 0.01)	Ref
	Male	146	0.15	(0.13 to 0.18)	6.37	0.19	(0.16 to 0.22)	6.12	0.036	(0.03 to 0.04)	5.51
Age	16–24	35	0.13	(0.09 to 0.18)	Ref	0.16	(0.11 to 0.22)	Ref	0.033	(0.02 to 0.05)	Ref
	25–34	43	0.12	(0.09 to 0.17)	0.96	0.14	(0.10 to 0.19)	0.92	0.029	(0.02 to 0.04)	0.86
	35–49	51	0.10	(0.08 to 0.14)	0.81	0.13	(0.10 to 0.17)	0.82	0.027	(0.02 to 0.04)	0.81
	50+	41	0.05	(0.04 to 0.07)	0.38	0.07	(0.05 to 0.09)	0.42	0.012	(0.01 to 0.02)	0.36
Remoteness	Major cities of Australia	64	0.04	(0.03 to 0.05)	Ref	0.05	(0.04 to 0.07)	Ref	0.011	(0.01 to 0.01)	Ref
	Regional and remote Australia	106	0.17	(0.14 to 0.21)	4.15	0.23	(0.18 to 0.27)	4.33	0.035	(0.03 to 0.04)	3.05
Activity*	Swimming/wading	112	0.06	(0.05 to 0.07)	Ref	0.11	(0.09 to 0.13)	Ref	0.043	(0.04 to 0.05)	Ref
	Attempting a rescue	28	0.03	(0.02 to 0.04)	0.56	1.21	(0.93 to 1.54)	11.11			0.00

^*Rate per 1 million activity sessions was unable to be calculated for attempting a rescue due to no available denominator.

^†Rates per 1 million visits shown to three decimal places due to small numbers.

RR, Rate Ratio.

Rip current survival and escape methods

Between 1 April 2013 to 31 March 2023, 2.04 million Australian adults aged 16+ years were estimated to have been unintentionally caught in a rip, and 170 rip drowning deaths were recorded among Australian residents (aged 16+ years) table 4. This equated to an estimated 1 rip current drowning death for every 11 977 people who have been caught in a rip. When calculated by the method of escape, for each rip drowning death, 2449 people were assisted to safety by someone else (705 rescued by surfers, 762 were rescued by a lifeguard/surf lifesaver) and 8171 individuals rescued themselves (self-rescue).

Table 4. Comparative 10-year analysis of rip current drowning deaths to survivals by rip escape methods used by respondents collated within a nationally representative coastal safety survey dataset 2013–2023.

Ratio of deaths to escape
Rip escapes		Rip drowning deaths	Escape:drowning death	95% CI
2 036 043		170	11 977:1	(11 960 to 11 993)
Ratio of deaths to escape by escape method
Escape methods	%	n	Escape:drowning death	95% CI
Self-rescued	68%	1 389 194	8171:1	(8157.46 to 8184.65)
Surf lifesaver/lifeguard	6%	129 560	762:1	(757.91 to 766.21)
Surfer	6%	119 843	705:1	(700.91 to 708.9)
Other member of the public	15%	296 515	1744:1	(1737.79 to 1750.35)
Other	3%	58 426	344:1	(340.88 to 346.46)
Can not say	2%	42 505	250:1	(247.63 to 252.39)
Total	100%	2 036 043

Discussion

This comprehensive epidemiological update of rip drowning in Australia reaffirms rip currents as the major coastal hazard. Rip current drowning deaths are most prevalent among males aged 16–69 years. The incidence of rip current drowning deaths was found to be lowest among children aged 0–15 and those aged 70 years and over, likely in part due to lower exposure to coastal waters.

Since Brighton et al,⁴ one area of optimism has been the significant reduction in the rate of male rip drowning deaths, a traditionally high-risk group,¹⁸ declining by 2% every 2 years. If the rate of male drowning deaths had remained stable at the high point of the study period (2004–2005: a rate of 0.16/100 000 population), a total of 369 drowning deaths would have been seen to have occurred among the Australian male population, a total of 24 deaths in excess of the actual number of deaths (n=345). This reduction could relate to the significant collective efforts of water safety stakeholders in Australia and may indicate preventative impacts of targeted education campaigns within the Australian community (ie, swim between the flags and the think line²⁵). Similar rates of decline were seen among 16–24, 25–34 and 50+ years, as well as among those living in major cities—however, these effect sizes were not found to be statistically significant, in part as a result of small sample sizes (ie, lower prevalence of rip current drowning among these population groups).

Despite the significant decline in male rip current drowning, no meaningful changes have been seen in the overall rip current drowning rate in Australia (0.09 per 100 000 population in the present study compared with 0.10 in Brighton et al). As such, continued efforts into finding ways to address the rip current hazard must be explored.

There is a clear need to invest in a range of preventative and response measures to address reducing drowning risk, particularly in unpatrolled locations. One such approach is improved and enhanced evidence-based education on the rip hazard, such as learning how to spot a rip in situ to prevent unintentionally entering one. The growing popularity of technologies such as smart devices, geo-targeting of safety messages and artificial intelligence should be explored as avenues for rip education.

The majority of rip drowning deaths occurred in the presence of others (80%, n=326), often family and friends (70.3%, n=286). Our study estimated that members of the public assisted in 296 515 rescues over the 10-year study period, the majority of which are not recorded in official rescue statistics. Linked to this is the finding that bystanders performing rescues, often of family members,¹⁷ account for the second highest proportion of rip drowning deaths (15.5%), commonly at unpatrolled beaches, suggesting that safety messages and advice focused on bystander rescue scenarios should be developed, tested and implemented. Equally important is the need to improve rescue, CPR and swimming skill levels more broadly within the community (ie, high-risk populations such as older Australians, regional and multicultural communities) and to build safer recreation in dynamic, less controlled aquatic environments (ie, at the beach, estuaries or other natural waterways) through targeted programmes.

There is a need to evaluate the effectiveness and usability of traditional public rescue equipment, such as rescue tubes, angel rings and emergency alert systems. Public rescue equipment has previously shown promise as a drowning prevention intervention;²⁶ however, there is a need to identify public awareness and use of such devices in the Australian context. Additionally, given the significant risk associated with performing a bystander rescue,²⁷ evaluating measures which help bystanders to rapidly alert emergency services, such as Emergency Response Beacons or other strategies to better perform non-contact rescues should also be a priority.

Lifesaving personnel are an essential component of the coastal safety infrastructure, adding essential knowledge and resourcing to the broader emergency response system.²⁸ This study estimates that over the 10-year study period, lifesaving personnel performed 129 549 rip-related rescues. The vast majority of rip drowning deaths occur away from lifesaving services, with 59% of deaths occurring more than 1 km from a surf lifesaving service. Further, one in two rip current drowning deaths (44.2%) occurred outside of the summer season, while 56.9% occurred on weekdays, periods which are typically unpatrolled by lifesaving services in Australia. A geospatial analysis of rip drowning deaths to determine where the addition of lifesaving services at selected high-risk, high visitation locations may be warranted, as well as extending lifesaving services delivery during the high-risk periods identified in the data examined in this study.

In addition to lifesaving personnel, surfers represented a valuable safety resource at unpatrolled locations, performing an estimated almost 120 000 rescues over the 10-year study period; 5.9% of rip current rescues. This supports previous research highlighting the important first responder role of surfers.^{10 29 30} Evaluation of surfer rescue training programmes has found such programmes to be effective both in training surfers on how to perform rescues and in maximising the personal safety of the surfer themselves.³¹ Any holistic approach to rip drowning prevention should incorporate such programmes.

Finally, the impact of rip drowning extends beyond the deaths reported in this study. There is an uncalculated injury toll that needs further investigation to fully understand the burden, with an estimated ~2 million people unintentionally caught in a rip current over the study period. For every one rip drowning death that occurred, it was estimated that an additional 11 976 individuals survived being caught in a rip, an experience which can leave the survivor with significant, and at this time unquantifiable, psychological trauma.¹⁶ This hidden burden further supports the need for additional investment in rip drowning prevention measures.

Strengths and limitations

Strengths of this study include the total population sample, extended time period, and integration of additional data sources to calculate novel population and exposure-based rates. However, limitations still exist. The years and age group available in the NCSS data limit the extension of these analyses to the entire rip drowning death dataset. Similarly, while the method of triangulation between multiple sources and review by experts to determine if a case was related to a rip is likely the most accurate way possible to determine such a factor, it is vulnerable to error, in part, as it relies on narrative accounts from witnesses who themselves may not be experts in the field (ie, members of the public, as opposed to lifesaving personnel). It is possible that some of the cases included in this analysis were not related to rips, and equally, that other cases not included in this analysis were rip related. This will always be a limitation in any analysis related to rip incidents, but we believe that our multiple source triangulation and expert review have minimised error to the greatest extent possible. Finally, in relation to activity at the time of the incident, due to the nature of the data collected, a distinction was not able to be made between swimming in deeper waters and wading in shallower waters. Distinguishing between the risk profile of these two activities is something future research should seek to address.

Opportunities and recommendations

Future research should seek to use geospatial modelling techniques to determine where services are most needed and which areas are underserved. There remains a need to consider the potential effect of local knowledge of an area serving as a protective factor in drowning prevention, comparing the incidence of rip current drowning deaths among decedents who live locally to the drowning location compared with those not local to the region.

Conclusions

Rip currents are the dominant hazard on surf beaches. While some reductions were identified in Australian rip current drowning rates, especially among males, other demographic groups remain at significant drowning risk. Lifesaving personnel represent a vital part of the coastal safety infrastructure, and their presence is generally understood as the most effective means to keep people safe while recreating in aquatic locations. However, they are only one component of drowning prevention, with broader efforts to engage the community with rip current education and awareness programmes required to truly decrease the incidence of rip current drowning deaths, particularly at unpatrolled locations.

Data availability statement

Data are available on reasonable request.