Simulation and educationDrones may be used to save lives in out of hospital cardiac arrest due to drowning☆
Introduction
The World Health Organisation (WHO) estimates that 372,000 people more than 90% of whom reside in low and middle income countries (LMIC), lose their lives in accidental drownings worldwide each year.1 Out of n = 137 drowning accidents in Sweden in 2014, a total of 36% occurred in swimming areas and beaches.2
Although rare case reports show that neurologically intact survival is possible after prolonged submersion in extremely cold water for 60–80 min3, 4 the European Resuscitation Council5 states that submersion duration of less than 10 min is normally required for a favorable outcome. The Swedish Registry for Cardiopulmonary Resuscitation (SRCR) shows a median delay of five minutes from collapse to call for help and 15 min from collapse to arrival of emergency medical services (EMS) in out-of-hospital cardiac arrest (OHCA) due to drowning.6
Since standard EMS response time usually extends beyond 10 min nearby bystanders and professional lifeguards play a considerable role in the rescue as well as in early resuscitation.5, 6
In-water resuscitation (IWR) with ventilations can be performed at an early stage by highly trained rescuers with a buoyant rescue aid which increases chances of survival from out-of-hospital cardiac arrest (OHCA) due to drowning.5, 7, 8 When the victim is retrieved from the water, cardiopulmonary resuscitation (CPR) should be initiated and an automated external defibrillator (AED) should be attached promptly, although a shockable rhythm is rare in asphyctic OHCA.5
According to data from the Royal Lifesaving Society in the United Kingdom (UK) from 2002, a majority (79%) of drowning victims in beach conditions are rescued within 50 m of the shore.9 Data from Australia show a rate of 128 rescues/100,000 swimmers with offshore wind and rip currents imposing a serious physical hazard to swimmers.10, 11 Rescuing a drowning victim with swimming as the main rescue method is common in cold, deep water.12 Death of rescuers have been reported due to lack of swimming ability and appropriate flotation devices.13, 14
As part of the search and rescue (SAR) operation, a trained rescuer response preferably using surf-lifeguards (SL) walking the shallow waters in straight lines has traditionally been used to find submerged drowning victims at lifeguard-patrolled Swedish beaches. However forming such a search party is time-consuming often exceeding 10 min before initiation of the search (Peter Karlborg lifeguard captain Tylösand SLSC, personal communication August 8, 2016) visibility is poor and the inclusion of volunteers impose a risk of fatigue, hypothermia and drowning.
The chance of survival decreases for every minute15 and Strömsöe et al. speculate that improvements in logistics could increase survival, if initiation of CPR after collapse occurred earlier – within 0–2 min, as compared to start of CPR in >2 min – many lives could theoretically be saved.16
In addition, accessibility of rescue diving units is limited in Sweden17 and both helicopter emergency services (HEMS)18 and search and rescue (SAR) helicopter services takes time to be recruited.19
Unmanned aerial vehicles (UAV), commonly called drones, have increasingly been used for photography as well as for video surveillance, situational awareness and transport purposes.20, 21 Drones have the capability of instant launch, providing live aerial video-stream to a tablet display. Drones have also reportedly been used to aid swimmers in distress by delivering life-buoys.22
A drone transmitting live video to a tablet can possibly provide a safe and high-quality visual overview of an accident scene for providing earlier location of a submerged possible drowning victim. To our knowledge there is no other study showing the effectiveness of using a drone for this purpose in a recreational beach environment.
The aim of this simulation study was to evaluate the efficiency of a drone for providing earlier location of a submerged possible drowning victim in comparison with standard procedure.
This is a prospective simulation study comparing two methods for the location of a submerged drowning victim in calm conditions at Tylösand beach in the south of Sweden. Live video-stream from a drone to a tablet was compared to traditional rescuer response by a SL search party. Totally n = 10 simulated drownings were presented for each group using a manikin (Laerdal Resusci Junior www.laerdal.com), 112 cm height. The manikin was submerged in the sea between a depth of 0.5–1.5 m within a 100 square m (m2) predefined search area in randomized and GPS-logged (n = 10) locations for each group.
All locations were GPS-logged, identical and blinded to all participants in both control and intervention group before each search. Time from start at the beach (search party positioned at the shoreline in the same location for all n = 10 tests and the UAV-pilot ready with engines shut-off) to contact with the manikin was set as the primary outcome variable. Both groups were briefed in the same way that a child was missing (submerged) within a 100 × 100 m test area.
All tests were performed in daytime hours during 2 weeks in July and August 2016. The sea conditions were calm with <4 m/s wind, no more than 0.5 m wave height at any point, sandy flat bottom conditions which ranged from 20 cm depth at the beach to 2 m at a maximum distance of 100 m out from the beach.
A trained rescuer response consisting of (n = 14) SL from Tylösand Surf Lifesaving Club (SLSC) were recruited to form a search party (control), searching for a submerged manikin, Fig. 1. The search party starting at the beach held a rescue-line (23 m) and then walked outwards vertically from shore, looking and feeling for a submerged victim with their feet. The group proceeded until they had water up to their armpits, at about 1.5 m depth, they then turned and proceeded back to shore. The search party made contact with the manikin and initiated in-water ventilations.
Time from start at the beach to contact of the manikin was documented as well as coverage (m2/min) using a waterproof GPS tracker that recorded time, position and distance. This was placed on the shoulder of one SL at one end of the search party (Finis inc. Hydrotracker, http://www.finisinc.com).
The surf-lifeguards who comprised the search party wore wetsuits, and a maximum time of 30 min to recognition of the manikin was pre-defined in order to protect volunteers from any potential exposure to hypothermia. Delay from entering the water to locating and making contact with the manikin was documented (Fig. 2).
A SL with basic intuitive training for flying a UAV was positioned on the beach and launched the drone to 60 m altitude, an assistant helped in searching for the manikin on the screen. Via live video-stream from the UAV to an Apple iPad air 9.7 inch tablet display (www.apple.com) the pilot and an assistant manually analysed the screen, searching for the submerged manikin. After locating a possible submerged victim a third designated SL was alerted and the drone descended to 5 m altitude, hovering over the scene to mark recognition. The SL followed the position of the drone and made contact with the manikin thus initiating IWR. Time from UAV launch to (a) locating the manikin on the tablet and (b) physical contact with the manikin was documented.
The drone used was a DJI Inspire 1 Pro that communicated using (http://www.dji.com/product/inspire-1/remote-controller#lightbridge) equipped with a DJI Zenmuse x5, Full HD, 4K, f1.7 sensor camera. All video recordings were documented using a 64 GB micro SD card. The search pattern used was a modified creeping line search23. About 50% of the search area was covered with the drone hovering at an altitude of 60 m, giving a maximized footprint, Fig. 3.
In order to safely evaluate this novel method’s effectiveness, tests were performed without swimmers in the water. The drone was programmed to return home (a set GPS-coordinate on the beach) if the battery ran low or if the signal from the pilot was lost. Test flights were all executed within line of sight and granted by the Swedish Aviation Authority.
Data on area coverage (m2) was retrieved through Finis GPS tracker and was exported and computed in Google Earth. Video from each search was recorded and compared to manual timestamps with a focus on delay to locating and making contact with the manikin. For descriptive statistics SPSS version 21 was used. Mann–Whitney’s U-test was used for comparison of the search party and UAV group. A p-value of <0.05 was regarded as significant.
Ethical approval was not applicable as this study did not concern research of humans according to current Swedish legislation.
Section snippets
Results
A total of 20 searches were performed, 10 in each group (control and intervention). The manikin was found within 10 min in all tests. The search party covered a median of 2600 (IQR: 2001–3526) square meters (m2) per search, equaling 570 m2/min (IQR: 459–911). The UAV at 60 m covered approximately 50% of the search area equaling 5000 m2 thus covering a larger median area of 4590 m2/min (IQR: 4319–9282) than the search party, p < 0.001.
The total median time from start to contact with the manikin was 4:34
Discussion
The main finding of this study is that the use of drone transmitting live video to a tablet is feasible and time saving in comparison to traditional search parties for providing earlier location of submerged possible drowning victims. This search method has the potential to possibly contribute to earlier recognition and initiation of CPR in submerged victims, however there are advantages and disadvantages with both methods, see Table 2. To our knowledge this is the first study showing the
Conclusion
A drone transmitting live video to a tablet is feasible, time saving in comparison to traditional search parties and may be used for providing earlier location of submerged victims at a beach. Drone search can possibly contribute to earlier onset of CPR in drowning victims.
Conflict of interest statement
No conflicts of interest to declare.
Acknowledgements
We wish to acknowledge the Swedish Lifesaving Society (SLS) and the surf-lifeguards at Tylösand Surf Lifesaving Club (SLSC) Sweden for their help and participation in performing this study.
References (26)
- et al.
European Resuscitation Council Guidelines for Resuscitation 2015: Section 4. Cardiac arrest in special circumstances
Resuscitation
(2015) - et al.
Characteristics and outcome among patients suffering out-of-hospital cardiac arrest due to drowning
Resuscitation
(2008) - et al.
In-water resuscitation—is it worthwhile?
Resuscitation
(2004) - et al.
Surf lifeguard rescues
Wilderness Environ Med
(2013) - et al.
Rescues conducted by surfers on Australian beaches
Accid Anal Prev
(2015) - et al.
The role of bystanders during rescue and resuscitation of drowning victims
Resuscitation
(2010) - et al.
A study on rescuer drowning and multiple drowning incidents
J Saf Res
(2012) - et al.
Predicting survival from out-of-hospital cardiac arrest: a graphic model
Ann Emerg Med
(1993) - et al.
Characteristics of lifesaving from drowning as reported by the Swedish Fire and Rescue Services 1996–2010
Resuscitation
(2012) - et al.
A 10-year analysis of 214 HEMS backcountry hoist rescues
Air Med J
(2013)
Use of unmanned aerial vehicles for medical product transport
Air Med J
Near-drowning and drowning classification: a proposal to stratify mortality based on the analysis of 1831 cases
Chest
European Resuscitation Council Guidelines for Resuscitation 2015 Section 2: adult basic life support and automated external defibrillation
Resuscitation
Cited by (58)
Blockchain-based privacy-preserving authentication protocol for UAV networks
2023, Computer NetworksDrones reduce the treatment-free interval in search and rescue operations with telemedical support – A randomized controlled trial
2023, American Journal of Emergency MedicineSimulation-Based Drone Assisted Search Operations in a River
2022, Wilderness and Environmental MedicineA fatal drowning filmed in a private pool: Analysis of the sequences of submersion
2021, Forensic Science International: ReportsEffect of supercapacitors on the operation of an air-cooled hydrogen fuel cell
2021, HeliyonCitation Excerpt :Currently, the demand for UAVs is mainly consumed by military applications, however, in recent years the demand in commercial, recreational and public applications has increased tenfold, and it is expected to exponentially shift further in this direction [9, 10]. The use of UAVs in criminal, theft and poaching surveillance is one of the major applications [11], with other applications including scientific monitoring (water sampling, landslides, volcanic activity) and transmission line surveillance [12, 13, 14, 15]. At present, combustion engines remain the most popular power supply for UAVs.
- ☆
A Spanish translated version of the abstract of this article appears as Appendix in the final online version at http://dx.doi.org/10.1016/j.resuscitation.2017.01.003.