Several papers yesterday headlined claims that accidental deaths from carbon monoxide poisoning rose sharply in the 12 months to the end of June 2011.

“CO poisoning deaths treble in a year” is The Press and Journal’s take on the story.  The Belfast Morning Telegraph reports that Ulster tops the UK deaths table, while the BBC says that Devon has been named as the top “hot spot” for incidents of CO poisoning.

The reports derived from the Gas Safety Trust’s Carbon Monoxide Hotspot Report for 2011, which records 25 fatalities in 2010-11 against just seven the year before. It further finds that while deaths have risen the number of incidents has fallen from 72 to 50, and the number of casualties from 138 to 80.

Ulster comes high on the list because of three incidents in which a total of seven people died. In Devon, there were five incidents and two deaths.  These numbers are simply too small and randomly distributed to allow a league table of hot spots to be constructed. But are they even right?

The figures are based entirely on press reports. Such sources aren’t to be despised, but it says something for the dearth of reliable information about carbon monoxide deaths that they should be the sole source. While deaths that are reported are likely to be accurate as they generally originate from coroners’ inquests, there may be others that are not reported at all.

Other sources of data are the Health and Safety Executive and the charity CO-Gas Safety, and both tell a very different story. The HSE data cover only those incidents caused by flammable gas, mainly piped gas but also LPG. They record 16 deaths in 2005-06, 10 in 2006-07, 13 in 2007-08, 15 in 2008-09, and (provisionally) nine in 2009-10. Delays in coroners’ inquests mean that this figure is likely to be too low. 

CO-Gas Safety’s figures are far more convincing, although it is a small charity operating on a modest budget.  It records (with names and dates) 594 people who have died in the past 15 years (an average of around 40 a year) and its data includes deaths involving gas mains, portable gas, solid fuel and petrol. It too uses press reports but it also contacts coroners to check details. The best summary of its work appears in this press pack.

The charity’s most recent data (see Table) show reductions in deaths since the 1990s but a pretty steady figure for the past decade of around 30 deaths a year. In a 2009 report for the Department of Communities and Local Government, the CO-Gas Safety figures are quoted and compared with a very similar estimate for 2007 made by the Office for National Statistics of 35 deaths due to CO that year. This estimate appears to have been made at the DCLG’s request and is not part of the normal ONS output.

However, these figures are in sharp disagreement with the data from the Hot Spot report. There’s no reason to believe that deaths in 2009/10 were as low as seven, as the Hot Spot report claims – CO-Gas Safety counted 25. So the purported increase to 25 this year found by the Hot Spot report is probably not an increase at all. Nor is it the case, if one considers all the deaths recorded by CO-Gas Safety over the past 15 years, that either Devon or Northern Ireland stand out as in any way exceptional. Both fall into the second-highest quintile for CO-related accidental deaths over that period.

What does stand out is the absence of official statistics on this cause of death. My guess is that the CO-Gas Safety figures are about as good as there are. It’s extraordinary that the founder and director of the charity, Stephanie Trotter OBE, assisted by a friend, should be able to collect figures that appear beyond the heating industry. The figures are regularly updated, mostly by adding deaths for recent years because it can take three years for an inquest to be held, but very occasionally removing deaths that turn out later to have been suicides.

As she puts it on her website: “We are shocked that our data is better than Government's. We try to check most deaths with Coroners and we have built up a good relationship with them over the 15 years we have been doing this. We also check with other bodies, such as the Solid Fuel Association, which has always been extremely helpful to us.”

It’s also rather shocking that nobody is trying harder to fill this gap. ONS has demonstrated that it has the capacity to make an estimate, but doesn’t do so on a regular basis.  The evidence is that the number of deaths has fallen since the 1990s, but it is still unacceptably high. There are also many "mear-misses" and in September the Department of Health issued experimental data derived from Hospital Episode Statistics suggesting that 4,000 people a year attend A&E departments with CO poisoning. The DH also believes that there are 50 deaths a year, and 200 injuries tnat require admission to hospital. This suggests that the deaths recorded by CO-Gas Safety may be an underestimate.

The Hot Spots report was researched and compiled by an undergraduate at St John’s College, Cambridge. No offence to him, and I’m sure he did the best with the data at his disposal, but I think it’s time greater efforst were made to measure this significant cause of death. The Gas Safety Trust records that it has spent more than £170,000 on data collection and analysis since September 2007, employing a number of consultancies to do the work. The Hot Spots report is not the only publication it produces using this data, but it is the one most likely to be read by gas consumers.

CO-Gas Safety, meanwhile, is seeking grants to continue its work.