Yesterday users of designer drugs including mephedrone were told they were playing Russian roulette with their lives.

The Independent attributed the remark to Professor Les Iverson, Chairman of the Advisory Council on the Misuse of Drugs, in a report recording ACMD’s conclusion that mephedrone had been linked to 98 deaths.

Colourful metaphors should at least convey a correct public health message. The lethality of mephedrone is several orders of magnitude less than that of Russian roulette.

Far from causing one death is every six users, the answer is closer one in 80,000 (using data from Office for National Statistics and British Crime Survey),  or to one in 20,000 using sources cited by ACMD in its report on novel psychoactive substances, which gives an estimate of 25 mephedrone-related deaths in England and Wales between July 2009 and June 2010.

The ACMD report’s new data on mephedrone-related fatalities made headlines by ditching denominators.

Fatalities: ACMD reported that, between July 2009 and June 2011 (2 years), the former National Programme on Substance Abuse Deaths (np-SAD) was notified of 42 mephedrone-related deaths: in England and Wales, there were 90 suspected cases with toxicology, 70 of which were positive for mephedrone. Half of the 90 had also completed the legal process, of which 30 were determined as mephedrone-related deaths. Hence, over 2 years and all things being equal, England and Wales might expect 66 mephedrone-related deaths, and say 33 (up to 40) in the first 12 months.

However, the plot thickens when the curious time-sequence of the legally-determined deaths is examined (data are available for 35/42 mephedrone-related deaths). In 2009 there were four deaths: September (1), October (1), and December (2).  In 2010. before the ban, there were  18: January (6), February (6), March (6); and in the rest of  2010, eight: April (3), May (3), June (1), November (1). In the first six months of 2011 there were five: February (1), March (1), April (2), and May (1).

There are three possible explanations for this pattern. Has the ban worked; did forensic testing for mephedrone peak when media interest did; or is the time-sequence explained by the waiting time for legal processes to complete?

The twelve months from July 2009 to June 2010 accounted for 29/35 dated mephedrone-related deaths and England and Wales for 30/42 dated mephedrone-related deaths. If completed legal cases pertain to earlier deaths, then we’d expect 25 mephedrone-related deaths in England and Wales in the 12-months from July 2009 to June 2010.

Mixmag dance magazine’s surveys do not suggest the ban led to the immediate abandonment of mephedrone. Pre-ban, it found 34 per cent of respondents reported use of mephedrone in the previous month; post-bam, the figures was 25 per cent. However, not too much weight should be put on an opt-in survey without a denominator

The ACMD also noted that at least 19 inquests had commented on parents’ concern that their child’s apparent suicide (typically by hanging) was linked to having felt low or depressed after mephedrone use. Whether these19 inquests are a subset of the cited 42 mephedrone-related fatalities was not made clear.

Missing data: The ACMD report omitted to mention how many fatalities were mephedrone-only. This is an important question, to which an answer is needed. Also missing were comparative data on cocaine-related or MDMA/ecstasy –related deaths [see Tables below, as reported by a different source: Office for National Statistics by (much delayed) year of registration.]

Ditched denominators:  The British Crime Survey’s (BCS) representative sampling showed that 4.4 per cent of 16 to 24 year-olds 2010/11 reported past-year use of cocaine powder (that is: in 2009/10) and the same proportion for mephedrone. As ACMD notes, the percentage was much lower at 0.6 per cent of respondents aged 25to 59.

With around 7.3 million 16-24 year olds in England and Wales, the BCS estimates (applied to E & W) imply that England and Wales had around 320,000 past-year users of mephedrone aged 16 to 24; and some 150,000 users aged 25 to 59.  Together they represent a plausible denominator for ONS’s mephedrone-related deaths registered in 2010 (but likely to have occurred in 2009): 6 out of 470,000 past-year users.

Far from one mephedrone-related death in six (Russian roulette), the ONS answer is closer to one in 80,000. Or, using np-SAD’s maximal data for July 2009 to June 2010, around 25 to 40 out of 470,000, that is between one in 20,000 and one in 10,000.

The range between these two estimates is one order of magnitude. But the Russian roulette metaphor is wide by at least three orders of magnitude.

Conflict of interest: SMB writes in a personal capacity but is a member of Home Office’s Surveys, Design and Statistics Subcommittee.