Scuba diving

Is scuba diving safe?

For a fit, well-trained diver, mostly yes, but each dive carries a risk in the same range as a skydive. The best recent figure comes from the US: 563 recreational diving deaths in 2006–2015, from about 306 million dives, or 1.8 per million. Older data and training dives come out closer to 5 per million. We use 2 to 5 micromortsMicromortA one-in-a-million chance of death. It lets you compare the risk of very different activities on one scale.More in the glossary per dive.

What makes diving unusual is how most deaths unfold. They rarely start with something sudden and violent. They start with a problem that could have been fixed, often running low on gas, and get worse underwater.

What goes wrong underwater

Divers Alert Network (DAN) researchers worked through 947 diving deaths from 1992 to 2003, tracing each from the first problem to the cause of death. The first problem, the "trigger", was:

Trigger	Share of deaths
Insufficient gas	41%
Entrapment (wrecks, caves, kelp, lines)	20%
Equipment problems	15%

The injuries that then disabled the diver were asphyxia (33%), arterial gas embolism (29%) and heart events (26%). Death certificates then record drowning in about 70% of cases, which hides the real story. A diver who runs out of gas, bolts for the surface holding their breath and suffers a lung injury ends up recorded as drowned.

The lesson is that the gauge and the buddy matter more than the shark. Running out of gas is almost always a planning failure.

The heart problem

About a quarter of diving deaths involve a heart event, mostly in older divers. Diving loads the heart in several ways at once: cold water, exertion against currents, carrying heavy gear on a boat deck and the shift of blood into the chest when immersed. Among insured DAN America members in 2000–2006, divers older than 49 had about 13 times the risk of a disabling heart event of younger divers.

This is the part of diving risk that skill cannot remove. A perfect gas plan does not help if an undiagnosed heart condition shows up 20 metres down.

Injuries and decompression illness

Non-fatal injuries are less common than in most sports. Buzzacott and colleagues estimated about 1,394 emergency visits a year in the US for scuba injuries, against millions of dives. The ratio is the troubling part: 47 deaths for every 1,000 emergency visits, far higher than for most sports, because problems underwater are less forgiving.

Decompression illness, caused by nitrogen bubbles after ascent, is rare on recreational dives within tables or computer limits. It needs oxygen on site and often hours in a recompression chamber. On liveaboards and in remote places, the nearest chamber can be a flight away, which is why we score the rescue environment at 50.

Who should be careful

• Divers over about 45 or with heart risk factors. Talk to your doctor before diving, and be honest on the medical questionnaire.
• Divers who have not dived in a year or more. Skills such as gas monitoring and buoyancy fade. Book a refresher dive.
• Anyone going into overhead environments. Caves and wreck penetration are where entrapment deaths cluster. They need separate training and redundant gas.
• Solo and "instabuddy" divers. A buddy who stays close and knows your gear can share gas when you need it.

How to dive more safely

1. Plan gas, then check it. Turn the dive with a reserve, commonly a third of your starting gas, and check your gauge every few minutes.
2. Ascend slowly and do a safety stop. Never hold your breath on the way up.
3. Stay within your training. Depth and overhead limits exist because those are where problems escalate.
4. Do a buddy check before every dive. Air on, weights, releases, alternate air source.
5. Get dive insurance with evacuation and chamber cover, especially on remote trips, and know where the nearest chamber is.
6. Wait before flying. Most guidelines suggest at least 12 hours after a single dive and 18 or more after several days of diving.

How we scored it

Fatality risk (76) sits in the 1–5 micromortMicromortA one-in-a-million chance of death. It lets you compare the risk of very different activities on one scale.More in the glossary band. Injury risk (80) and controllability (72) are fair, since most triggers respond to training, but the rescue environment (50) is poor. Compare skydiving, with a similar risk per outing, and marathon running, where heart events are also the main cause of death.