Alcohol is the most damaging drug to the drinker and others overall, heroin and crack are the second and third most harmful, Professor David Nutt and colleagues wrote in the medical journal The Lancet today. When all factors related to self harm and harm to others are considered, alcohol comes out top. The authors explain that drugs, including tobacco products and alcohol are major contributors to damage to individuals as well as society as a whole.
The harms that are caused by drugs need to be comprehensively assessed so that policy makers can be properly advised regarding health, social care and policing, the authors write; not an easy undertaking because drugs can cause damage in so many different ways.
Professor Nutt and colleagues had previously tried to do this (Lancet 2007) by asking experts to give each drug a score according to nine criteria of harm, which included the drug's intrinsic harms as well as the social and health care burdens. The report triggered widespread debate and interest. However, there were doubts regarding the differential weights of each criterion used.
In this latest report, Nutt and colleagues say they have addressed these concerns by using a multicriteria decision analysis (MCDA) when reviewing drug harms. MCDA technologies have been effectively used to help decision making in areas where factors, features and characteristics are complex and often conflicting, as may be the case when deciding policy on nuclear waste disposal.
Nine criteria related to harm to an individual from a drug, while six looked at harm to others - both in the United Kingdom and other countries. The harms were gathered into five subgroups that covered social, psychological and physical harms. Scoring was done with points up to 100, with 100 being the most damaging and zero no damage. Weighting then compared the impact a score of 100 had on all the other criteria, thus identifying the 100-points-scoring-drugs which were more harmful than other 100-points-scoring-drugs.
The authors wrote (in explanation of their model):
In scaling of the drugs, care is needed to ensure that each successive point on the scale represents equal increments of harm. Thus, if a drug is scored at 50, then it should be half as harmful as the drug that scored 100.
The nine harm-to-self categories of a drug were:
- drug-related damage
- drug-related impairment of mental functioning
- drug-related mortality
- drug-specific damage
- drug-specific impairment of mental function
- loss of relationships
- loss of tangibles
- decline in community cohesion
- economic cost
- environmental damage
- family conflict
- international damage
- Alcohol, overall harm score 72
- Heroin, overall harm score 55
- Crack, overall harm score 54
- Crystal meth, overall harm score 33
- Cocaine, overall harm score 27
- Tobacco, overall harm score 26
- Speed/amphetamines, overall harm score 23
- Cannabis, overall harm score 20
- GHB, overall harm score 18
- Valium (benzodiazepines), overall harm score 15
- Ketamine, overall harm score 15
- Mephedrone, overall harm score 13
- Butane, overall harm score 10
- Khat, overall harm score 9
- Ecstasy, overall harm score 9
- Anabolic steroids, overall harm score 9
- LSD, overall harm score 7
- Buprenorphine, overall harm score 6
- Mushrooms, overall harm score 5
- The most harmful drugs to the individual are heroin, crack and crystal meth
- The most harmful drugs to others are alcohol, heroin and crack
Mephedrone, which was recently a legal-high in the UK before it was re-categorized as a Class B controlled drug this year. Alcohol is over five-times as harmful as mephedrone.
Ecstasy is just one-eighth as harmful as alcohol, despite all its media attention and public concerns.
Professor Nutt said (direct quote, not found in article):
What a new classification system might look like would depend on what set of harms-to self or others-you are trying to reduce. But if you take overall harm, then alcohol, heroin and crack are clearly more harmful than all others so perhaps drugs with a score of 40 or more could be class A; 39 to 20 class B; 19-10 class C and 10 or under class D.
The MCDA procedure is an effective and powerful means for dealing with the complex issues related to drug misuse, the authors wrote.
The issue of the weightings is crucial since they affect the overall scores. The weighting process is necessarily based on judgment, so it is best done by a group of experts working to consensus.
(conclusion) Our findings lend support to previous work in the UK and the Netherlands, confirming that the present drug classification systems have little relation to the evidence of harm. They also accord with the conclusions of previous expert reports that aggressively targeting alcohol harms is a valid and necessary public health strategy.
In an associated Comment, also in The Lancet, Dr. Jan Van Amsterdam, National Institute for Public Health and the Environment, Netherlands, and Dr Wim van den Brink, Amsterdam Institute for Addiction Research, Academic Medical Center, University of Amsterdam, Netherlands, wrote:
A major point not addressed in the study, because it was outside their scope, is polydrug use, which is highly prevalent among recreational drug users. Notably, the combined use of alcohol with other drugs often leads in a synergistic way to very serious adverse effects.
They also explain that consuming combinations of these drugs can significantly alter their adverse events and harm impacts. For example, magic mushrooms on their own have a very low incidence of adverse events, but individuals who consume mushrooms as well as alcohol have a much higher risk of accidents that result in death. Other examples of combinations mentioned include alcohol with cocaine, leading to cocaethylene - an extremely toxic compound, or alcohol with cannabis which can seriously affect an individual's ability to drive properly.
The Comment authors concluded:
Nutt and colleagues' ranking of the licit and illicit drugs is certainly not definitive, because the pattern of recreational drug use is dynamic: the popularity and availability of the drugs, and the pattern of polydrug use, might change within a decade. The ranking of the drugs. should therefore be repeated at least every 5-10 years. Finally, for the discussion about drug classification, it is intriguing to note that the two legal drugs assessed-alcohol and tobacco-score in the upper segment of the ranking scale, indicating that legal drugs cause at least as much harm as do illegal substances.
"Drug harms in the UK: a multicriteria decision analysis"
Prof David J Nutt FMedSci a Corresponding AuthorEmail Address, Leslie A King PhD b, Lawrence D Phillips PhD
The Lancet, Early Online Publication, 1 November 2010
Written by Christian Nordqvist