3 Drug related problems

3.1 Introduction
The drug problem of a nation is not measured simply by the number of drug users or even by the aggregation of the adverse effects of their drug use. A nation such as Turkey or Peru, with relatively few users, may still be seen as having a major drug problem because it either serves as a major trafficking route (Turkey) or a producer (Peru), with the attendant corruption and large illegal incomes. The UNODC has attempted to take account of this by constructing an Illicit Drug Index (IDI), which measures each nation’s contribution to the global drug problem, summing those associated separately with consumption, production and trafficking (UNODC, 2005). Though the IDI has many weaknesses (e.g. its effort to turn all adverse consequences into health harms, its assumption that all grams of a given drug are equally harmful) it does represent an important step forward in aggregating across the many dimensions of drug problems.
For purposes of the UNGASS assessment it would be highly desirable to use something like the IDI and aggregate problems across nations for 1998 and 2007. Even putting aside the conceptual problems, this is impossible because of limited data availability. We were able to obtain data for only a very small number of the adverse consequences for most nations in our 18 country sample. We collected data on three measures: drug-related deaths (DRD), HIV, and crime.

3.2 Drug related deaths
These data are available for a few Western countries. Comparability across countries is limited by differences both in the definition of a DRD and also in the methods by which death certificates are generated.41 The procedure for determining whether death is the consequence of a drug overdose ranges from a full post-mortem to superficial medical check by a GP. Nations also differ in how the data are aggregated. In some countries data on overdose deaths are registered separately; in others these data are included in the general mortality register. The latter is the case in the Netherlands which guarantees national coverage but includes on the other hand only residents of the Netherlands. However, though cross-country comparisons are of doubtful validity, it is possible to make comparisons of the number of overdose deaths within a country at two different points in time. Note that we are including only deaths in which drug use was the direct, acute cause. Not included are those in which drug use is the ‘indirect’ cause, e.g. death by drug use related diseases and accidents. For example, deaths related to Hepatitis B, in which the cause of the infection was previous injecting drug use are not counted as drug related deaths.
Nor are homicides which result from drug-related disputes included in these figures.
In a large number of countries the number of such deaths has declined somewhat in the second half of our study period.
For example, in the European Union the EMCDDA estimates that the number of DRD approximately doubled from 1990 to 2000 but then fell by about 15% to 2005 (Table DRD5 http://www.emcdda.europa.eu/html.cfm/index52843EN.html) There was a decline of more than 50% in Australia over the period 2000-2005, reflecting presumably the influence of the heroin “drought” that started at the end of 2000.42
No data on drug-related deaths are available for most non-Western countries, including Brazil, China, India, Iran or Russia.
In Mexico there are regular newspaper statements about the number of drug-related homicides (approximately 4,000 in 2008) but they have not been verified and in any case fall outside the scope of drug-related deaths recorded in other nations.
Given that these countries account for such a large share of all heroin addicts globally, it is impossible to make statements about how drug-related deaths have changed globally between 1998 and 2007.

3.3 HIV
The extent of drug related HIV varies substantially across countries, both expressed as the fraction of injecting drug users who are HIV positive or the share of HIV cases for which injecting drug use is the most likely vector of transmission. Among our 18 countries, the rates of HIV in IDUs are low (i.e. less than 5%) for most Western countries. These are mostly nations that undertook aggressive campaigns against HIV early on. Though the contribution of explicitly harm reduction programs such as Syringe Exchange Programs (SEP) to keeping the rates low remains controversial, the weight of the evidence is usually read as favouring that proposition (e.g. Institute of Medicine, 2006).
For HIV we were able to make use of the relatively sophisticated data systems that have been developed in many countries, including some impoverished nations, to track the spread of the AIDS virus. However, numbers on HIV+ drug users are frequently calculations based on samples and on assumptions of the actual source of infection, either sexual behaviours or injecting drug use. The way they are collected differs substantially among countries. But recently a few global overview studies in the field of HIV prevalence and prevalence of injecting drug use were published that we considered useful for our purpose (Mathers et al., 2008; Cook & Kanaef, 2008). The HIV infection rate among IDU is high for Iran and the United States both of which started their control campaigns only after the epidemic was well established among injecting drug users.
HIV among injecting drug users is also a major problem in Russia, where IDU account for over 60% of all new HIV infections in the middle of this decade; the same was true for Iran. For India the best data suggest a modest level, about 10% among IDU; they account for a small share of the estimated 5 million HIV infections in the country. This last statement is in contrast to the situation in China where the estimated number of HIV infected drug users went up from 12,536 in 1998 to 637,000 in 2007, though again there are substantial differences between available estimates.

3.4 Crime
In Western countries there is considerable evidence from surveys of arrestees that drug use is much higher among the criminally active than it is in the general population. For example, in the United Kingdom voluntary urine tests of arrestees in eight sites found that about 30% tested positive for heroin use (Bennett and Holloway, 2005). The question of whether the relationship between drug use and crime is causal is a vexed one (Stevens, 2007) but the extent of drug use among arrestees is often cited as an indication of the extent of drug-related crime.
Much of the drug-related crime in Western countries is property crime, intended to generate income for the purchase of expensive illicit drugs. Though numerous studies show that offenders commit more property crimes when using drugs than when abstinent (e.g. Gossop, Marsden, Stewart and Kidd, 2003), there are no population level indicators that would permit tracking the change over time for a country.
In a few countries the more important crime is the violence related to drug markets themselves. For example, in Brazil it is asserted that many of the killings associated with gangs in the favelas surrounding Rio de Janeiro and Sao Paulo arise from the struggle to control drug markets (e.g. Zaluar, 2004). Similar competitive and transactional violence has been observed on a large scale in Mexico, the United States and Russia.
There is no systematic measure of this particular manifestation of drug-related crime in any country. Impressionistically, it seems that the violence has declined in the United States, perhaps because of the aging of the buyers and sellers in cocaine markets and perhaps because the more violent dealers are more likely to be incarcerated. For other countries there is no basis for making statements about change over the ten year period.

3.5 Economic cost estimates
One way of aggregating the severity of a nation’s drug problems is to try to estimate the economic costs arising from use, production and distribution. A small literature is now available for a few Western countries; for cross-country comparisons see Single et al. (2003).
Proper identification of the harms within a society is important for understanding the extent to which the individual drug problem affects individuals and the broader society vis-à-vis other societal problems. Inconsistencies in how harms are identified, how they are measured over time, and the extent to which they are measured consistently with other harms within the same geopolitical boundaries or across geopolitical boundaries makes it difficult to develop a solid understanding of the magnitude of the problem and how/whether it is changing over time. Report 3 reports the results of efforts to develop new estimates that use a consistent methodology for nine Western countries, Australia, Canada and the United States, along with the six of the largest member States of the European Union. It is based on a close examination of the published estimates for seven nations.
The most powerful finding from the study is simply that the exercise is infeasible even for the countries with the most advanced monitoring and data systems. To return to an issue previously discussed, drug related deaths (DRDs), we are faced with the following implausible comparison: DRDs are estimated to be 1,979 in the Unite Kingdom and 2612 in the United States. Yet the US is estimated to have at least five times as many cocaine and heroin dependent users as does the United Kingdom. It is simply not credible that the mortality rate amongst these users is so much lower in the United States. Similarly, the Hepatitis C. figure for IDU in Australia is 62,000 while for France it is only 1,000; given that France has more users in OST than does Australia and there are high rates of Hep C. in treatment populations, this is completely implausible.

41 The EMCDDA studied this issue extensively in the 1990s (e.g. EMCDDA, 1997) and has developed guidelines for the reporting of deaths but compliance remains low.
42 Opiod death figures rose rapidly through the 1990s to a peak of 1116 in 1999. Thereafter they fell sharply to 374 in 2005, close to the figures around 1990. http://ndarc.med.unsw.edu.au/NDARCWeb.nsf/resources/NIDIP_FactSheet_Opioid/$fil /OPIOIDS+OVERDOSE+2005.pdf [accessed February 14, 2009].