I agree wholeheartedly with Lamont, however, that is only one piece of the OOA problem. The entire issue is a bit more complex. In the DAN workshop report it is noted that:
Fundamental problems associated with diving fatalities have not changed significantly in recent history (Denoble et al. 2011; Lippmann 2011; Cumming et al. 2011; Richardson 2011). The most frequently cited root cause among the independent population samples was insufficient gas or running out of gas. Other common factors included entrapment or entanglement, buoyancy control, equipment misuse or problems and rough water. Emergency ascent was also common. The principal injuries or causes of death included drowning or asphyxia due to inhalation of water, air embolism and cardiac events.
Now, look at that statement closely for a moment. It says, "The most frequently cited root cause among the independent population samples was insufficient gas or running out of gas." Now, to my way of thinking, for "insufficient gas or running out of gas," to kill a diver, requires a failure of an emergency ascent .
Similarly, the DAN report states:
Equipment problems seldom directly cause the divers death, but the divers response to an equipment malfunction, especially if the response is panic and perhaps a rapid ascent to the surface, may be the catastrophic event in the sequence that results in a diving-related fatality.
So now we can add pretty much all equipment problems as well as insufficient gas or running out of gas to the, "failure of emergency ascent," category.
The DAN report, notes:
Gas-supply problems resulted mainly from inappropriate gas management. Most of the time, the trigger appeared at depth, and the diver drowned or suffered an AGE during emergency ascent.
I feel confident, at this point, to suggest that the most important unsolved issues is not gear manipulation (though that's important), nor is it gas planning (though good gas planning would remove a sizable chunk of cases), the problem is that when things go wrong divers are not able, at least some of the time, to make it to the surface without damage. So let's ask the $64,000 question, "why can't divers make a safe free ascent?"
The DAN Accident Workshop provides some insight into this:
A study in Belgium found 100400 times increased risk of pulmonary barotrauma (PBT) during training dives, while emergency free-ascent training was associated with 5001,500 times greater risk. These findings prompted Belgian sport diver federations to ban free-ascent training in 2006, after which there have been no further PBT cases related to training (Lafére et al. 2009).
I was party to many conversations in the USA that followed a similar path. Most recreational instructors are poorly trained in how to conduct free ascent training, when that is combined with the fact that recreational divers receive little training (usually one tense, frightened, trial) in how to do a free ascent, that combination is, understandably, deadly. There were a lot of free ascent training incidents, and the result was that the agencies tightly controlled the way in which free ascent training was conducted. During this process I pointed out that in the science community we did lots of free ascent training and had never had a single incident, but the agencies did not want to hear that. I think that the subtext of this entire topic was how to reduce the incidence of free ascent training accidents while reducing the required course hours at the same time (the shorter course hour discussion was going on at the same time).
So we are left with a situation in which we are prevented from taking the single biggest step to increase diving safety, effective free ascent training, by agency blindness and the fact that free ascents have not been a central pillar of diver training since the 1060s and the current instructor cadre has not been trained to conduct such training, but rather has been systematically terrified to do so.