Compressed gas diving fatalities in Australian waters 2014 to 2018.

INTRODUCTION: This study aimed to investigate compressed gas diving deaths in Australia from 2014-2018 and make comparison to those from 2001-2013 to identify ongoing problems and assess countermeasures. METHODS: Media reports and the National Coronial Information System were searched to identify scuba diving deaths for 2014-2018, inclusive. Data were extracted from the witness and police reports, medical histories, and autopsies. An Excel® database was created and a chain of events analysis conducted. Comparisons were made with the earlier report. RESULTS: Forty-two fatalities were identified, 38 using scuba and four using surface-supplied breathing apparatus involving 30 males and 12 females. The mean age of victims was 49.7 years, six years higher than the previous cohort. Fifty-four percent were obese. Six victims were unqualified, three were under instruction and at least 28 were experienced divers, significantly more than in the previous cohort. Health-related predisposing factors, predominantly obesity and cardiac-related, were identified as likely contributory to 26 incidents, and planning shortcomings to at least 22 deaths. One-third of the disabling conditions were primary drowning and one-quarter were cardiac. Three divers died subsequent to carbon monoxide poisoning and three likely from immersion pulmonary oedema. CONCLUSIONS: Advancing age, obesity and the associated cardiac disease have become increasingly prevalent in diving fatalities and the need for appropriate assessment of fitness to dive is evident.

A review of diving practices and outcomes following the diagnosis of a persistent (patent) foramen ovale in compressed air divers with a documented episode of decompression sickness.

INTRODUCTION: The presence of a persistent (patent) foramen ovale (PFO) increases the risk of decompression sickness (DCS) whilst diving with pressurised air. After the diagnosis of a PFO, divers will be offered a number of options for risk mitigation. The aim of this study was to review the management choices and modifications to diving practices following PFO diagnosis in the era preceding the 2015 joint position statement (JPS) on PFO and diving. METHODS: A retrospective study was conducted of divers sourced from both the Alfred Hospital, Melbourne and the Divers Alert Network Asia-Pacific during the period 2005-2015. Divers were contacted via a combination of phone, text, mail and email. Data collected included: diving habits (years, style and depths); DCS symptoms, signs and treatment; return to diving and modifications of dive practices; history of migraine and echocardiography (ECHO) pre- and post-intervention; ECHO technique(s) used, and success or failure of PFO closure (PFOC). Analyses were performed to compare the incidence of DCS pre- and post-PFO diagnosis. RESULTS: Seventy-three divers were interviewed. Sixty-eight of these returned to diving following the diagnosis of PFO. Thirty-eight underwent PFOC and chose to adopt conservative diving practices (CDPs); 15 chose PFOC with no modification to practices; 15 adopted CDPs alone; and five have discontinued diving. The incidence of DCS decreased significantly following PFOC and/or adoption of conservative diving practices. Of interest, migraine with aura resolved in almost all those who underwent PFOC. CONCLUSIONS: Many divers had already adopted practices consistent with the 2015 JPS permitting the resumption of scuba diving with a lowering of the incidence of DCS to that of the general diving population. These results support the recommendations of the JPS.

Immersion pulmonary edema: case reports from Oceania.

INTRODUCTION: We aimed to document identified cases of immersion pulmonary edema (IPE) in divers from Oceania (the Indo-Pacific region) from January 2002 to May 2018, inclusive. METHOD: Cases were identified using various sources, including searches of the Divers Alert Network Asia-Pacific (DAN AP) Fatality Database, published case reports, and interviews with survivors who had reported their incident to DAN AP. Where available, investigations, pathology and autopsy results were obtained. Only incidents diagnosed as IPE by diving physicians or pathologists with experience in the investigation of diving accidents were included. Individual case histories and outcomes, together with brief individual summaries of the associations and possible contributing factors were recorded. RESULTS: Thirty-one IPE incidents in divers from Oceania were documented. There were two surface snorkelers, 22 scuba air divers and seven nitrox divers which included three closed-circuit rebreathers (CCR). The mean (SD) age was 53 (12) years, 58% of victims were females, and the average dive profile was to a maximum depth of 19 meters of seawater for 25 minutes. Six victims (19%) had previous episodes of IPE. There were nine recorded fatalities. Cardiac anomalies dominated the associated or possible contributing factors. These included valvular disease in 29%, transient cardiomyopathies in 26% and dysrhythmias in 16%. CONCLUSIONS: Previously reported associations of IPE such as exertion, stress, cold exposure, negative inspiratory pressure, hypertension, overhydration, ascent or surfacing, tight wetsuit, aspiration and certain medications were identified. Cardiac conditions were frequent and included chronic disorders (valvular pathology, coronary artery disease) and transient disorders (dysrhythmias, transient myocardial dysfunction, takotsubo or stress cardiomyopathy). It is likely that the chronic cardiac disorders may have contributed to the IPE, whereas the transient cases could be either sequelae, contributors or coincidental to the IPE.

Provisional report on diving-related fatalities in Australian waters in 2012.

INTRODUCTION: An individual case review of known diving-related deaths that occurred in Australia in 2012 was conducted. METHOD: The case studies were compiled using statements from witnesses and reports of the police and coroners. In each case, the particular circumstances of the accident and details from the post-mortem examination, where available, are provided. RESULTS: There were 26 reported fatalities (four less than the previous year). Only two of the victims were female (one snorkeller and one scuba diver). Fourteen deaths occurred while snorkelling and/or breath-hold diving, 11 while scuba diving and one diver died while using surface supplied breathing apparatus in a commercial pearl diving setting. Two breath-hold divers likely drowned as a result of apnoeic hypoxia. Cardiac-related issues were thought to have contributed to the deaths of at least three and possibly seven snorkellers and four scuba divers. CONCLUSIONS: Pre-existing medical conditions; poor organisation, planning and supervision; equipment-related problems; snorkelling or diving alone or with loose buddy oversight and apnoeic hypoxia were features in several deaths in this series.

Decompressing rescue personnel during Australian submarine rescue operations.

INTRODUCTION: Personnel rescuing survivors from a pressurized, distressed Royal Australian Navy (RAN) submarine may themselves accumulate a decompression obligation, which may exceed the bottom time limits of the Defense and Civil Institute of Environmental Medicine (DCIEM) Air and In-Water Oxygen Decompression tables (DCIEM Table 1 and 2) presently used by the RAN. This study compared DCIEM Table 2 with alternative decompression tables with longer bottom times: United States Navy XVALSS_DISSUB 7, VVAL-18M and Royal Navy 14 Modified tables. METHODS: Estimated probability of decompression sickness (PDCS), the units pulmonary oxygen toxicity dose (UPTD), the volume of oxygen required and the total decompression time were calculated for hypothetical single and repetitive exposures to 253 kPa air pressure for various bottom times and prescribed decompression schedules. RESULTS: Compared to DCIEM Table 2, XVALSS_DISSUB 7 single and repetitive schedules had lower estimated PDCS, which came at the cost of longer oxygen decompressions. For single exposures, DCIEM schedules had PDCS estimates ranging from 1.8% to 6.4% with 0 to 101 UPTD and XVALSS_DISSUB 7 schedules had PDCS of less than 3.1%, with 36 to 350 UPTD. CONCLUSIONS: The XVALSS_DISSUB 7 table was specifically designed for submarine rescue and, unlike DCIEM Table 2, has schedules for the estimated maximum required bottom times at 253 kPa. Adopting these tables may negate the requirement for saturation decompression of rescue personnel exceeding DCIEM limits.

Decompressing recompression chamber attendants during Australian submarine rescue operations.

INTRODUCTION: Inside chamber attendants rescuing survivors from a pressurised, distressed submarine may themselves accumulate a decompression obligation which may exceed the limits of Defense and Civil Institute of Environmental Medicine tables presently used by the Royal Australian Navy. This study assessed the probability of decompression sickness (PDCS) for medical attendants supervising survivors undergoing oxygen-accelerated saturation decompression according to the National Oceanic and Atmospheric Administration (NOAA) 17.11 table. METHODS: Estimated probability of decompression sickness (PDCS), the units pulmonary oxygen toxicity dose (UPTD) and the volume of oxygen required were calculated for attendants breathing air during the NOAA table compared with the introduction of various periods of oxygen breathing. RESULTS: The PDCS in medical attendants breathing air whilst supervising survivors receiving NOAA decompression is up to 4.5%. For the longest predicted profile (830 minutes at 253 kPa) oxygen breathing at 30, 60 and 90 minutes at 132 kPa partial pressure of oxygen reduced the air-breathing-associated PDCS to less than 3.1 %, 2.1% and 1.4% respectively. CONCLUSIONS: The probability of at least one incident of DCS among attendants, with consequent strain on resources, is high if attendants breathe air throughout their exposure. The introduction of 90 minutes of oxygen breathing greatly reduces the probability of this interruption to rescue operations.

A pleural vacuum relief device for pleural drain unit use in the hyperbaric environment.

INTRODUCTION: When a standard water-seal pleural drain unit (PDU) is used under hyperbaric conditions there are scenarios where excessive negative intrapleural pressure (IPP) and/or fluid reflux can be induced, risking significant morbidity. We developed and tested a pleural vacuum relief (PVR) device which automatically manages these risks, whilst allowing more rapid hyperbaric pressure change rates. METHODS: The custom-made PVR device consists of a one-way pressure relief valve connected in line with a sterile micro filter selected for its specific flow capacity. The PVR device is designed for connection to the patient side sampling port of a PDU system, allowing inflow of ambient air whenever negative pressure is present, creating a small, controlled air leak which prevents excessive negative pressure. The hyperbaric performance of a Pleur-Evac A-6000 intercostal drain was assessed with and without this added device by measuring simulated IPP with an electronic pressure monitor connected at the patient end of the PDU. IPP readings were taken at 10, 15, 20 and 30 cmH2O of suction (set on the drain unit) at compression rates of 10, 30, 60, 80, 90 and 180 kPa·min⁻¹ to a pressure of 280 kPa. RESULTS: At any compression rate of > 10 kPa·min⁻¹, the negative IPP generated by the Pleur-Evac A-6000 alone was excessive and resulted in back flow through the PDU water seal. By adding the PVR device, the generated negative IPP remains within a clinically acceptable range, allowing compression rates of at least 30 kPa·min⁻¹ with suction settings up to -20 cmH2O during all phases of hyperbaric treatment. CONCLUSIONS: The PDU PVR device we have developed works well, minimising attendant workload and automatically avoiding the excessive negative IPPs that can otherwise occur. This device should only be used with suction.

Provisional report on diving-related fatalities in Australian waters in 2011.

INTRODUCTION: An individual case review of diving-related deaths reported as occurring in Australia in 2011 was conducted as part of the DAN Asia-Pacific dive fatality reporting project. METHOD: The case studies were compiled using reports from witnesses, the police and coroners. In each case, the particular circumstances of the accident and, where available, details from the post-mortem examination are provided. A chain of events analysis was conducted for each case. RESULTS: In total, there were 30 reported fatalities (10 more than in 2010). These included 15 snorkel/breath-hold divers, 14 scuba divers and one diver using surface-supplied breathing apparatus. Twenty-four victims were males. The mean age of snorkelling victims was 49.6 (range 23-75) years and compressed gas divers 42.2 (range 23-55) years. Cardiac-related issues were thought to have been the disabling injury in the deaths of at least seven snorkel divers and five scuba divers. Immersion pulmonary oedema was implicated in at least one death; and three fatalities resulted from attacks by marine animals. Two novices died while under instruction/supervision after separation from their instructor in poor visibility. CONCLUSIONS: Pre-existing medical conditions, separation and inadequate supervision and seafood collection in areas frequented by marine predators were once again features in several deaths in this series.

A 10-year estimate of the incidence of decompression illness in a discrete group of recreational cave divers in Australia.

INTRODUCTION: The vast majority of freshwater cave diving in Australia occurs within the limestone caves of the Gambier karst in the south-east of South Australia. The incidence of decompression illness (DCI) in cave divers is presumed to be higher than open-water recreational divers because of the greater depths involved, but has not previously been reported. Our aim was to determine the incidence of DCI in cave divers, the patterns of diving and the outcome of hyperbaric treatment. METHODS: This was a retrospective cohort study of cave divers with DCI presenting to the Royal Adelaide Hospital or The Alfred Hospital over a 10-year period between 2002 and 2012. We reviewed case notes of cave divers who were treated for DCI after diving in the Mt Gambier karst. As there are no records of the number of dives performed during the study period we generated a denominator for the incidence of DCI by extrapolating available data and making a number of assumptions about the number of dives per dive permit issued. RESULTS: Sixteen patients were treated for DCI during the study period. The precipitating dive was a single deep decompression dive in seven cases, multiday repetitive dive sequences in eight and a non-decompression dive in one. Three of the 16 cases of DCI involved dives in excess of 90 metres' fresh water (mfw) using trimix. As the total estimated number of dives in the study period was approximately 57,000 the incidence of DCI in Australian cave divers was estimated to be 2.8:10,000 (0.028%). It is possible that the overall incidence of DCI is as high as 0.05%, and even higher when dives to depths greater than 90 mfw are involved. CONCLUSIONS: The estimated incidence of DCS in this series is lower than expected but consistent with other series describing DCI in cold-water recreational diving.

Provisional report on diving-related fatalities in Australian waters 2010.

INTRODUCTION: An individual case review was conducted of known diving-related deaths that occurred in Australia in 2010. METHOD: The case studies were compiled using statements from witnesses and reports of the police and coroners. In each case, the particular circumstances of the accident and details from the post-mortem examination, where available, are provided. A root cause analysis was made for each case. RESULTS: There were 20 reported fatalities, one less than the previous year. Five of the victims were female (four scuba divers) and 15 were males. Twelve deaths occurred while snorkelling and/or breath-hold diving, seven while scuba diving (one of whom was using a rebreather), and one diver died while using surface supplied breathing apparatus. At least two breath-hold divers likely drowned as a result of apnoeic hypoxia. Cardiac-related issues were thought to have contributed to the deaths of at least three and possibly five snorkellers, and of at least one, possibly two compressed gas divers. CONCLUSIONS: Snorkelling or diving alone, poor supervision, apnoeic hypoxia, pre-existing medical conditions, lack of recent experience and unfamiliar and/or poorly-functioning equipment were features in several deaths in this series. Reducing delays to CT-scanning and autopsy and coroners' reports documenting that the victim of a drowning was snorkelling or scuba diving at the time are aspects of the investigation of these fatalities that could be improved.

Oxygen exposure and toxicity in recreational technical divers.

INTRODUCTION: Central nervous system oxygen toxicity is a recognised risk in recreational open-circuit scuba diving with the use of nitrox (oxygen-enriched air mixtures), but other forms of oxygen toxicity in other diving settings are poorly understood. However, divers using constant partial pressure of oxygen closed-circuit rebreathers (CCRs) for multi-day, multi-dive expeditions could potentially experience cumulative oxygen exposures above the current recommended limits. METHODS: We followed a number of technical recreational diving expeditions using CCRs and recorded the cumulative oxygen exposures of the individual divers. Lung function and visual acuity were recorded at intervals during the expeditions. RESULTS: Over several 8- to 12-day expeditions, divers either approached or exceeded the recommended maximum repetition excursion oxygen exposure (REPEX) limits. Lung function did not show any significant decrement. Changes in visual acuity were reported in several divers but were difficult to quantify. Formal testing of one diver's visual acuity on return home demonstrated a myopic change that resolved over the subsequent eight weeks. CONCLUSIONS: Recreational CCR divers conducting multi-dive expeditions of eight days or more may approach or exceed the REPEX oxygen limits. Despite this, there does not appear to be any significant decrement in lung function. Hyperoxic myopia occurs in some individuals. Changes in acuity appear to resolve spontaneously post exposure. Despite the lack of significant changes in respiratory function, divers should be cautious of such exposures as, should they require recompression therapy for decompression illness, this may result in significant pulmonary oxygen toxicity.

Analysis of recreational closed-circuit rebreather deaths 1998-2010.

INTRODUCTION: Since the introduction of recreational closed-circuit rebreathers (CCRs) in 1998, there have been many recorded deaths. Rebreather deaths have been quoted to be as high as 1 in 100 users. METHODS: Rebreather fatalities between 1998 and 2010 were extracted from the Deeplife rebreather mortality database, and inaccuracies were corrected where known. Rebreather absolute numbers were derived from industry discussions and training agency statistics. Relative numbers and brands were extracted from the Rebreather World website database and a Dutch rebreather survey. Mortality was compared with data from other databases. A fault-tree analysis of rebreathers was compared to that of open-circuit scuba of various configurations. Finally, a risk analysis was applied to the mortality database. RESULTS: The 181 recorded recreational rebreather deaths occurred at about 10 times the rate of deaths amongst open-circuit recreational scuba divers. No particular brand or type of rebreather was over-represented. Closed-circuit rebreathers have a 25-fold increased risk of component failure compared to a manifolded twin-cylinder open-circuit system. This risk can be offset by carrying a redundant 'bailout' system. Two-thirds of fatal dives were associated with a high-risk dive or high-risk behaviour. There are multiple points in the human-machine interface (HMI) during the use of rebreathers that can result in errors that may lead to a fatality. CONCLUSIONS: While rebreathers have an intrinsically higher risk of mechanical failure as a result of their complexity, this can be offset by good design incorporating redundancy and by carrying adequate 'bailout' or alternative gas sources for decompression in the event of a failure. Designs that minimize the chances of HMI errors and training that highlights this area may help to minimize fatalities.

Provisional report on diving-related fatalities in Australian waters 2008.

INTRODUCTION: An individual case review of diving-related deaths, reported as occurring in Australia in 2008, was conducted as part of the DAN Asia-Pacific dive fatality reporting project. METHOD: The case studies were compiled using reports from witnesses, the police and coroners. In each case, the particular circumstances of the accident and details from the post-mortem examination, where available, are provided. RESULTS: In total, there were 19 reported fatalities (the same as for 2007), 17 involving males. Twelve deaths occurred while snorkelling and/or breath-hold diving,and six while scuba diving. One diver died while using surface-supply breathing apparatus. Two breath-hold divers appear to have died as a result of apnoeic hypoxia, at least one case likely associated with hyperventilation. Two deaths resulted from trauma: one from impact with a boat and the other from an encounter with a great white shark. Cardiac-related issues were thought to have contributed to the deaths of five snorkellers and at least two, possibly three, scuba divers. CONCLUSIONS: Trauma from a marine creature, snorkelling or diving alone, apnoeic hypoxia and pre-existing medical conditions were once again features in several deaths in this series.

Provisional report on diving-related fatalities in Australian waters 2009.

INTRODUCTION: An individual case review of diving-related deaths reported as occurring in Australia in 2009 was conducted as part of the DAN Asia-Pacific Dive Fatality Reporting Project. METHOD: The case studies were compiled using reports from witnesses, the police and coroners. In each case, the particular circumstances of the accident and details from the post-mortem examination, where available, are provided. RESULTS: In total, there were 21 reported fatalities (two more than in 2008), including 18 males and three females. Twelve deaths occurred while snorkelling and/or breath-hold diving, eight while scuba diving and one while using surface-supply breathing apparatus. Apneic hypoxia continues to be a problem with breath-hold divers and appears to have caused the death of three victims in this series. Cardiac-related issues were thought to have been the disabling injury in the deaths of at least three snorkel divers and at least three scuba divers. One of the victims was a student who became separated from her instructor on an introductory scuba dive in poor visibility. CONCLUSIONS: Apneic hypoxia, pre-existing medical conditions, snorkelling or diving alone, separation and inadequate supervision were once again features in several deaths in this series.

Provisional report on diving-related fatalities in Australian waters 2007.

INTRODUCTION: An individual case review of diving-related deaths reported as occurring in Australia in 2007 was conducted as part of the on-going Divers Alert Network (DAN) Asia-Pacific dive fatality reporting project. METHOD: The case studies were compiled using reports from witnesses, the police and coroners. In each case, the particular circumstances of the accident and details from the post-mortem examination, where available, are provided. RESULTS: In total, there were 19 reported fatalities, comprising three females and 16 males. Nine of the deaths occurred while snorkelling and/or breath-hold diving, eight while open-circuit scuba diving, one while using a closed-circuit rebreather, and one while using surface-supply breathing apparatus. Cardiac-related issues were thought to have contributed to the deaths of at least three but possibly up to six snorkel divers and possibly two scuba divers. One diver is believed to have died as a result of immersion pulmonary oedema of diving. Six of the compressed-gas divers were very inexperienced, three being certified within 14 days prior and dying while under the guidance of an instructor. CONCLUSIONS: Inexperience, pre-existing medical conditions and buoyancy issues were highlighted in several deaths in this series.

Pediatric hyperbaric oxygen therapy in Victoria, 1998-2010.

OBJECTIVES: The aim of this review is to identify clinical conditions currently treated in a pediatric population referred to the Alfred hyperbaric unit, to describe outcomes, and detail any complications occurring during treatment or transfer between units. DESIGN: Retrospective, noncontrolled, clinical study. SETTING: Adult hyperbaric unit in a university hospital. PATIENTS: Children aged <16 yrs referred for hyperbaric oxygen therapy between January 1998 and December 2010. INTERVENTIONS: Hyperbaric oxygen therapy at pressures from 2.0 to 3.0 atmospheres absolute. MEASUREMENTS AND MAIN RESULTS: Fifty-four patients with a median age at presentation of 15 yrs (range, 0.25-16 yrs) received 668 treatment sessions (mean, 12.4; 95% confidence interval, 9.2-15.5). Fourteen patients were identified as having successfully completed treatment while managed in intensive care units. There were 44 events in 668 treatments (6.6%) in the pediatric group and 12 events in 126 treatments (9.6%) in the pediatric intensive care unit group. There were two oxygen toxicity convulsion (0.3%), two episodes of progressive hypoxemia (0.3%), and four episodes of brief hypotension (0.6%). CONCLUSIONS: Provision of hyperbaric oxygen to children with significant illness is feasible and associated with a low risk of complications. The most difficult aspect of managing pediatric hyperbaric oxygen therapy is in the coordination of the treatment with ongoing surgical and intensive care management. The lack of pediatric staff and facilities in major hyperbaric units necessitates multiple transfers for appropriate treatment.

Provisional report on diving-related fatalities in Australian waters 2006.

INTRODUCTION: An individual case review of diving-related deaths reported as occurring in Australia in 2006 was conducted as part of the Divers Alert Network Asia-Pacific (DAN AP) dive fatality reporting project. METHOD: The case studies were compiled using reports from witnesses, the police and coroners. In each case, the particular circumstances of the accident and details from the post-mortem examination, where available, are provided. RESULTS: In total, there were 16 reported fatalities (eight fewer than 2005), all involving males. Ten deaths occurred while snorkelling and/or breath-hold diving and six while scuba diving, one of which involved the use of a closed-circuit rebreather. One death resulted from an encounter with a stingray and two involved scuba divers diving alone after an extended absence from diving. Cardiac-related issues were thought likely to have contributed to the deaths of six snorkel divers and one scuba diver. CONCLUSIONS: Trauma from a marine creature, snorkelling or diving alone and pre-existing medical conditions once again featured in several deaths in this series.

Cerebral arterial gas embolism with delayed treatment and a fatal outcome in a 14-year-old diver.

In today's recreational diving climate, diving fitness examinations are not mandatory, and even divers who go for these examinations may not have routine chest X-rays (CXR) done in the absence of respiratory symptoms or a past history of respiratory problems. We present a case of an ultimately fatal cerebral arterial gas embolism in a 14-year-old boy with an undiagnosed lung cyst, the contribution of which to his death is uncertain. Various factors such as lack of oxygen first aid at the remote dive site; poor communication; lack of diving medicine expertise, poor oxygen administration and management in a local hospital and long delay to recompression therapy contributed to the poor outcome. It is imperative that dive operators and physicians working in close proximity to popular dive sites be educated on how to recognise and treat diving emergencies and be well-acquainted, as should divers, with the contact numbers of diving medical hotlines that offer timely and appropriate advice in case of emergency.

Provisional report on diving-related fatalities in Australian waters 2005.

INTRODUCTION: An individual case review of diving-related deaths reported as occurring in Australia in 2005 was conducted as part of the DAN Asia-Pacific dive fatality reporting project. METHOD: The case studies were compiled using reports from witnesses, the police and coroners. In each case, the particular circumstances of the accident and details from the post-mortem examination, where available, are provided. RESULTS: In total, there were 24 reported fatalities, comprising four females and 20 males. Fourteen deaths occurred while snorkelling and/or breath-hold diving, nine while scuba diving, and one while using surface-supply breathing apparatus. Four deaths from large marine animal attacks were recorded. Once again, cardiac-related issues were thought to have contributed to some deaths: five snorkel divers and at least two but possibly up to four scuba divers. Three of the deaths in breath-hold divers were likely to have been associated with apnoeic hypoxia blackout. CONCLUSIONS: Pre-existing medical conditions, trauma from marine creatures and snorkelling or diving alone were features in several deaths in this series.