What accident analyses have found

Cave diving is unusual among sports in how directly its safety culture was built from the study of deaths. Course standards, equipment habits and the rules of thumb that cave divers repeat to each other can mostly be traced to people who looked hard at what went wrong and wrote it down. This article summarises the main published analyses in date order, with a note on their limits. It reports what the authors found; it does not turn their findings into procedures.

Before the analyses

The NSS Cave Diving Section was founded in 1973 by cave-diving members of the National Speleological Society, with Sheck Exley as its first chairman ^[2]. In the foreword to his Blueprint for Survival, Exley wrote that no fewer than 26 divers died in Florida's underwater caves in 1974, and that NSS recovery divers had noticed something troubling: friends and relatives of many victims said the dead divers had known about recommended cave diving practices but had chosen not to follow them ^[1].

Exley's Blueprint for Survival (1979)

Exley's answer, inspired, he wrote, by roadside accident photographs that had made him slow down on a highway, was to teach safety through real accidents ^[1]. Basic Cave Diving: A Blueprint for Survival, published by the NSS-CDS in 1979, opens each chapter with an accident report, names changed, followed by an analysis and a principle ^[1][2]. The chapters cover the guideline, air supply planning, going too deep, panic, lights, scuba equipment, silt, emergency procedures and technological and physiological emergencies, and the book closes with ten recommendations ^[1]. In summary, they are: always run a single, continuous guideline from the entrance; plan gas with the 'third rule', keeping two thirds of the starting supply for the exit and emergencies; avoid deep diving in caves; avoid panic by building experience slowly and preparing for emergencies; carry at least three lights each; carry the safest scuba equipment available; avoid stirring up silt; practise emergency procedures with your partner and review them often; carry the equipment needed for emergencies and know how to use it; and never let overconfidence talk you into breaking safety practices ^[1].

The book also reports what the NSS had found in its own case reviews: that failing to keep two thirds of the gas supply for the exit and emergencies came a close second, behind not using a continuous guideline, among the prime causes of cave diving deaths ^[1]. Exley noted that his procedures were written for Florida caves and that conditions elsewhere could call for different approaches ^[1]. The NSS-CDS credits this rule set, and the education built around it, with reducing cave diving deaths from 26 in 1974 to fewer than five a year ^[2]. The book remains in circulation as a free download from the NSS-CDS ^[3].

Florida forensic review (1997)

Writing in the Journal of Forensic Sciences, Byrd and Hamilton reviewed Florida cave diving deaths from 1962 to 1991 and from 1995 to mid-1996, noting that data for 1992-1994 were not available and that no formal accident reporting system existed ^[4]. They concluded that lack of experience and failure to follow safety precautions were the leading causes, and proposed a standard one-page form to make the initial reporting of a fatality more accurate and consistent ^[4]. The paper also suggested that fatality numbers rose and fell in an eight-year cycle; the abstract presents this as a basis for predicting a rise in deaths over the following five to seven years.

DAN: 368 deaths, 1969-2007 (2009)

In 2009, Buzzacott and colleagues at Divers Alert Network published an analysis of 368 American cave diving deaths from 1969 to 2007 ^[5]. Deaths per year peaked in the mid-1970s and declined afterwards. Drowning was the most frequent cause of death, most often after the diver ran out of gas, usually after getting lost or after starting the dive with too little gas. Trained divers who died tended to be older, died deeper and further into caves, carried more cylinders and were more often alone; untrained divers who died were more likely to have dived without a guideline, without enough lights or without enough gas ^[5]. The authors recommended that gas management rules receive the greatest emphasis in cave diving courses ^[5].

DAN: thirty years of American cave diving deaths (2016)

Potts, Buzzacott and Denoble's 2016 paper in Diving and Hyperbaric Medicine reviewed cave diving deaths reported to DAN between 1 July 1985 and 30 June 2015, recording each diver's training status, which safety rules were broken, how relevant each violation was, and the root cause of death ^[6]. They identified 161 divers who died: 67 trained cave divers and 87 untrained divers (the abstract does not account for the remaining seven). Annual deaths fell over the period from about eight to fewer than three, but the proportion of trained divers among them doubled. Comparing two 15-year halves, trained divers who died in the later period were older, but little else differed. The most common cause of death was drowning, preceded by running out of gas, usually after getting lost when suspended silt destroyed visibility, and the great majority of deaths occurred in Florida ^[6]. The authors concluded that, despite improvements in technology, the greatest hazards in cave diving are unchanged, and that work on preventive measures should continue ^[6].

Recovery and reporting

A 2017 paper in Forensic Science International, co-written by a member of IUCRR, put the average number of divers who die inside flooded caves in the United States at about three a year, and described the standard recovery procedure IUCRR has used since 1999, from call-out to handover to the authorities and record-keeping, with five illustrative cases ^[7]. Recovery practice matters for analysis too: the scene evidence a recovery team records, such as the position of the diver and the state of valves, gauges and lines, is part of what later reports rely on ^[7][11].

DAN's own data rest on voluntary reports. It has kept a diving incident database since 1989, invites divers to report injuries, near misses and witnessed accidents, and publishes cases without identifying individuals ^[8]. Its publication library holds annual reports analysing incidents and fatalities reported to it from 1988 onwards ^[9].

The NSS-CDS Accident Analysis Committee today

The NSS-CDS re-formed its Accident Analysis Committee to continue systematic analysis of cave diving accidents in North America, and publishes case reports on its website ^[10]. Several recent reports concern trained divers using rebreathers, consistent with the rising share of trained divers among the dead that the 2016 review described ^[6]. Reports on deaths in the Devil's Eye system in 2021 and at Cow Springs in 2023 concern trained rebreather divers diving alone; in both, the committee concluded that a buddy might have changed the outcome ^[11][12]. The 2022 Manatee Springs report concerns introductory-level cave divers attempting a traverse beyond their training in a cave none of them knew ^[13]. The 2022 Roaring River report concerns a breathing gas unsuited to the planned depth and an oxygen-toxicity seizure ^[14]. The 2023 Phantom Lake Spring report concludes that a very experienced explorer most likely died of a sudden cardiac event at about 140 metres, and urges deep and overhead divers to take medical fitness seriously ^[15].

The themes are familiar from the older studies (gas, navigation, staying within training) but newer ones are prominent: rebreather set-up and monitoring, solo diving, task loading during survey, and health. A 2022 review in Diving and Hyperbaric Medicine adds a forward-looking caution, noting that the growing practice of carrying a second rebreather as bailout on long cave dives brings risks that remain to be clearly defined ^[16].

Beyond the United States

The large fatality analyses summarised above are American, and most of the deaths they count occurred in Florida ^[6]. A study from Australia takes a different approach, trying to measure risk per dive: a 2015 Australian paper on decompression illness among cave divers in the Mount Gambier karst of South Australia. It found 16 cases treated over ten years (2002-2012) and, using an estimate of about 57,000 dives built from permit data and stated assumptions, put the incidence at about 2.8 per 10,000 dives, possibly as high as 5 per 10,000 ^[17].

What the analyses cannot tell us

The fatality studies count deaths, not dives. Without a reliable count of dives, a falling death toll could reflect safer practice, changes in how many people dive caves, or both; even the Australian study had to build its denominator from permit numbers and stated assumptions ^[17]. Reporting also depends on people choosing to report: DAN's database relies on voluntary submissions ^[8]. Case reports without witnesses rest on reconstruction; the NSS-CDS committee says so plainly in its Cow Springs report ^[12]. And a published rule set describes what has gone wrong in the past, not everything that can go wrong in a particular cave. The lesson the authors themselves draw is not that cave diving has been made safe, but that the same few hazards keep recurring and deserve continued attention ^[6].