Lies, damned lies, and statistics…

The last 3 months has been pretty horrendous in the UK diving community with 11 fatalities occurring.  Unsurprisingly there have been a number of social media posts questioning what is going on when the mathematical mean is 16.45 per year, and yet the MCA stated that the reported incidents had fallen to the lowest level on 21 years in 2013. So how risky is diving? 

What is Risk?

The HSE defines risk as the probability of a hazard causing an effect. A hazard is something (e.g. an object, a property of a substance, a phenomenon or activity) that can cause adverse effects. For example:

  • Excessively high pO2 is a hazard because it can cause oxygen toxicity.
  • The poison in marine life is a hazard because it can cause injury if you touch the toxin-containing object.
  • Over-saturation of inert gas above a certain threshold in the tissues and bodily fluids is a hazard because it can cause Decompression Sickness (DCS).

Risk can be measured in both quantitatively and qualitatively terms. e.g. in the UK, there is a risk that 1:X [probability] dives will end in a fatality [effect] due to DCS [hazard], or there is more [probability] risk of dying [effect] from crashing [hazard] whilst driving to the dive site than a fatality [effect] due to DCS [hazard] during a dive.

The Quantitive Measurement of Risk

To define a quantitive risk there is a need to have a numerator (event) and a denominator (sample population size) to make the comparison against. e.g. 1 in 1000 events will end up as an adverse event. (There is also a need to define what an adverse event is but that will be the subject of another blog!). If we examine aviation, we know how many flights occur, how many passengers fly and how many miles they fly. We also know how many aircraft crash and as a consequence, how many fatalities there are. Therefore, the risk of 1:XXXXX fatality per mile flown globally or 1:XXXX (in USA) is possible to calculate. It is also possible to calculate the driving fatality rate of 1:XXXXX per billion miles driven in the UK. These can all be calculated because both denominator and numerator are fairly accurately known, mainly because the environments are controlled and legislated and the number of accidents (effects) is accurately captured. As long as the numerator and denominator are both recorded (or can be inferred with a high level of confidence), a number of other metrics can be determined e.g. hull losses per year in general aviation, injuries per car owned, crashes per hour driven or major injuries per hour flown.

Accuracy and Availability of Numerical Data with Sports Diving in the UK (or globally for that matter!)

Fatality Numerator: We have a pretty good idea about the number of UK diving fatalities as they are relatively widely reported in the media, however, their causes are not so widely known.  The majority are the subject of a Coroner’s inquiry, some are covered by a Sheriff’s Fatal Accident Inquiry. Coroner’s inquest reports are only available to interested parties, the definition of which is down to the Coroner, but NGBs, academia or research organisations are not normally considered. In the case of a FAI, some are made public, but only three are listed here with the keyword “SCUBA”  There are a number of reasons why the detail is not available but that isn’t the subject of this blog.  The following table provides a summary of the fatalities since 1998. The Rec and Rec CCR are from the BSAC Annual Report and the CCR attribution has been determined from the narrative in each entry therefore it may not be correct.  I have not yet gone through the annual reports by month to determine frequencies per month to show clustering.

Table of UK Diving Fatalities 1998-2013


Incident Numerator: We don’t have a good idea of the number of incidents occurring at all. Not least because there is no clear definition of incident.  If you consider that safety is the reduction (or elimination) of risk, and we dive in a sport where there are ’no rules’ (outside of HSE ‘At Work’ ACoPs) therefore the level of acceptable risk is down to the individual diver/team. Each diver has their own acceptable qualitative value which varies from day to day, buddy to buddy, year to year, and therefore determining what should be reported is also an issue.  I believe that the numbers of DCI incidents alone is in the order of 10 times more than reported to BSAC based on BHA figures, DDRC research and my own research into why incidents aren’t reported.

Diving Denominator: This is the biggest problem: how much diving actually goes on? What depths, gases, types of diving (OC, CCR, Overhead etc), bottom & deco times etc.  So the quote above from the MCA saying that incidents were the lowest for 21 years, could be down to the fact that people weren’t diving.  The weather in 2013 wasn’t great and a number of trips were cancelled (I’m waiting for some data back from some boat operators on this). Maybe down to the economy where people didn’t have as much disposal income as they used to and therefore weren’t diving. Conversely, the rise this year in fatalities might be because the economy is picking up and divers are getting back in the water but the early part of 2014’s weather precluded sufficient work-up diving. It might be that health issues may be surfacing (no pun intended), or Reason’s ‘frequency gambling’ odds are coming in

“One possible explanation is that crews lack the relevant experience or are unable to retrieve the knowledge needed to assess risk appropriately in those specific circumstances (cf. Klein, 1993b). Another arises from pilots’ routine experience. If similar risky situations have been encountered in the past and a particular course of action has succeeded, the crew will expect to succeed the next time with the same response, a phenomenon Reason (1990) called ‘‘frequency gambling.’’ – Orasanu, J.M., 2010, Flight crew decision-making, Crew Resource Management, pp. 147-79


The problem is we just don’t know what the quantitative risks are or what the detailed causality is, and therefore speculation or sensationalism doesn’t help in informing the community about the risks.  What would help is capturing better data and making it more widely available, not necessarily because it would help understand the absolute risk, but it would certainly inform the relative risks.

Ultimately, trying to work out if diving is a risky sport when only looking at the fatality rate is seriously flawed and is unlikely to help anyone improve safety. A better way is to talk about or report the silly mistakes we made, the ones we are embarrassed to admit to as we were breaking the ‘rules’ of best practice, thereby allowing other’s to learn from them.  Hard I know…

Safe Diving…


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s