Chapter 7
Drug-related infectious diseases and drug-related deaths

Drug-related infectious diseases

One of the more serious health consequences of the use of illicit drugs, and in particular of drug injection, is the transmission of HIV and other infectious diseases, notably hepatitis C and B. The relationship between drug injecting and the transmission of infection is well established. Reducing drug injecting and the sharing of injecting equipment has therefore become a primary goal of public health interventions in this area. Studies also point to a relationship between drug use and high-risk sexual activity; this suggests a growing importance in linking drug use interventions with public health strategies aimed at sexual health. In terms of monitoring at the European level, data on infectious disease are collected by regular notification sources, where drug injecting may be recorded as a risk factor, and during special studies of drug-using populations in different settings.


Recent trends in newly reported HIV cases

At present, most countries report low rates of newly diagnosed HIV infection attributed to injecting drug use. Countries that have maintained consistently low rates of HIV infection among injecting drug users (IDUs) provide an opportunity to explore what factors can account for this, a question posed by the current EU drugs action plan and currently a focus of an EMCDDA coordinated study. Complacency should be avoided, however; neither Spain nor Italy, both of which have experienced HIV epidemics among IDUs, provides national HIV case reporting data, and this has a significant negative impact on the value of these data for describing the overall EU picture. Furthermore, data emerging from some countries are raising concerns that HIV infection may be increasing, at least among some populations of IDUs.

In France, where HIV cases have been recorded only since 2003, there has been an increase in the incidence of HIV among IDUs (albeit from a low level), from an estimated 2.3 cases per million population in 2003 to 2.9 cases in 2004. Although this generally agrees with the available study data (see below), it must be remembered that new reporting systems are often initially unstable. In Portugal, an apparent decline in newly diagnosed cases of HIV among IDUs recorded previously is called into question by the 2004 data, which reveal an incidence of HIV infection of 98.5 cases per million population, the highest in the EU (186). In the United Kingdom, the incidence of HIV among IDUs has slowly increased but is now stable at just under 2.5 cases per million population per year. In Ireland, the incidence increased during the late 1990s to a peak of 18.3 cases per million per year in 2000, fell to 9.8 per million in 2001, and subsequently increased to 17.8 cases per million in 2004.

HIV outbreaks related to injecting drug use occurred as recently as 2001 in Estonia and Latvia, and 2002 in Lithuania. Since then rates have strongly declined; a decline in the rates of newly reported cases is to be expected after an initial epidemic phase as an endemic level of infection becomes established (see below).

HIV seroprevalence among tested IDUs

Seroprevalence data from IDUs (percentage infected in samples of IDUs) are an important complement to HIV case reports. Repeated seroprevalence studies and routine monitoring of data from diagnostic tests can support conclusions drawn from case reporting data and can provide more detailed information on specific regions and settings. However, prevalence data come from a variety of sources and, in some cases, may be difficult to compare; thus, they should be interpreted with caution.

The recent increases in HIV in some countries recorded in the case reporting data are mostly confirmed by the available seroprevalence data, although the latter would suggest that these are not the only countries in which increased vigilance is necessary.

In the Baltic states, the available seroprevalence data indicate that transmission among IDUs may still not be under control (Figure 10). In Estonia, a recent study suggests that prevalence in IDUs is increasing in one region (Tallinn: from 41 % of a sample of 964 in 2001 to 54 % of 350 in 2005) and exceptionally high in another (Kohtla-Järve: 90% out of 100). In Latvia, two time series of seroprevalence data among IDUs show a continued increase until 2002/03 while a third series suggests a decrease since the peak in 2001. In Lithuania, data for 2003 suggest an increase in HIV among tested IDUs in drug treatment, needle exchanges schemes and hospitals, from between 1.0 % and 1.7 % during 1997–2002 to 2.4 % (27/1 112) in 2003.

Figure 10 HIV prevalence in tested injecting drug users, 2003–04

Figure 10


Figures in (brackets) are local data. Colour indicates midpoint of national data or, if not available, of local data.

Data for Italy and Portugal include non-injectors and are likely to underestimate prevalence in injectors.

* Data in part or totally before 2003 (Spain 2002–03; France 2002–03; Latvia 2002–03; Netherlands 2002) and from 2005 from Estonia.


Reitox national focal points. For primary sources, study details and data before 2003 or after 2004, see Table INF-8 in the 2006 statistical bulletin.


In the countries that have historically had high rates of HIV infection among IDUs (Spain, France, Italy, Poland and Portugal) there are new signs of continuing transmission at national level or in specific regions or among specific subgroups of IDUs (187). In these countries, it is important to note that the high background prevalence, resulting from the large-scale epidemics that occurred in the 1980s and 1990s, increases the likelihood that high-risk behaviour will lead to infection.

In Spain and Italy, national data from case reporting are not available. Data from routine diagnostic tests are difficult to interpret as they may be affected by different selection biases; however, they are a cause for concern in these countries. In Spain, HIV prevalence among young IDUs (under 25) tested in drug treatment was stable until 2002, at over 12 %, and among new IDUs (defined as injecting for less than 2 years) increased from 15 % to 21 % in 2000–01, suggesting a high incidence up to 2002. In Italy, trends in prevalence among IDUs differ strongly between regions, and some regions still show pronounced and recent increases (Bolzano, Liguria, Molise, Toscana and Umbria).

In Poland, local studies suggest that HIV spread has continued at least until recently, based on high prevalence rates among young IDUs (15 % in 2002 in one region, 4–11 % in two other regions in 2004). In the 2002 study, four cases (9 %) of HIV were reported among the 45 new injectors in the sample; however, in the 2004 study no cases were found among the 20 new injectors in the sample.

Finally, in some countries that have never experienced large-scale epidemics among IDUs, some recent prevalence data suggest that vigilance is warranted. This would seem to be the case in Luxembourg, Austria and the United Kingdom, although the increases remain limited and are not confirmed by case reporting data.

Low HIV prevalence countries

HIV prevalence among tested IDUs continues to vary widely between countries in the EU (Figure 10). In a number of countries HIV prevalence among IDUs has recently increased or has been high for many years. In contrast, in several countries, HIV prevalence among IDUs remained very low during 2003–04: HIV prevalence was less than or around 1 % in the Czech Republic, Greece, Hungary, Malta, Slovenia (based on national samples), and in Slovakia, Bulgaria, Romania, Turkey and Norway (based on subnational samples). In some of these countries (e.g. Hungary), both HIV prevalence and hepatitis C virus (HCV) prevalence are among the lowest in Europe, suggesting low levels of injecting (see ‘Hepatitis B and C’), although in some countries (e.g. Romania) there is evidence that the prevalence of hepatitis C is increasing.

Sex differences in HIV prevalence among tested IDUs

Available seroprevalence data for 2003–04 reveal differences between male and female tested IDUs (188). Combined data from Belgium, Estonia (2005), Spain (2002), France, Italy, Luxembourg, Austria, Poland and Portugal resulted in a total sample of 124 337 males and 20 640 females, tested mostly in drug treatment centres or other drug service provision sites. Overall prevalence was 13.6 % among males and 21.5 % among females. Differences between countries are marked, with female to male ratios being highest in Estonia, Spain, Italy, Luxembourg and Portugal, while Belgium shows the opposite trend, with prevalence being higher among males.

AIDS incidence and availability of HAART

As highly active antiretroviral treatment (HAART), available since 1996, effectively stops progression of HIV infection to AIDS, AIDS incidence data have become less useful as an indicator for HIV transmission. However, they still show the overall burden of symptomatic disease and in addition form an important indicator of the introduction and coverage of HAART among IDUs.

WHO estimates that coverage of HAART among patients in need of treatment was high in Western European countries (over 70 %) in 2003, but more restricted in most Eastern European countries, including Estonia, Lithuania and Latvia (189). More recent data on the coverage of HAART suggest that the situation has markedly improved, with all EU and candidate countries now achieving at least 75 % coverage. Specific data regarding availability of HAART among IDUs are not available, however, and it remains to be seen whether improved coverage will be reflected in a reduction in the incidence of AIDS among IDUs in Estonia and Latvia.

In all four countries in Western Europe most affected by AIDS, i.e. Spain, France, Italy and Portugal, the incidence has declined, since about 1996 in the case of the first three countries, but only since 1999 in Portugal. Portugal is still the country with the highest incidence of IDU-related AIDS, with 31 cases per million population in 2004. However, in Latvia the incidence is similar, 30 cases per million.

EuroHIV data up to 2004 (corrected for reporting delay) suggest that the incidence of IDU-related AIDS is increasing in both Estonia and Latvia (190).

Hepatitis B and C

Hepatitis C

The prevalence of antibodies against hepatitis C virus (HCV) among IDUs is, in general, extremely high, although there is wide variation both within and between countries. Prevalence rates of over 60 % among various IDU samples tested in 2003–04 are reported from Belgium, Denmark, Germany, Greece, Spain, Ireland, Italy, Poland, Portugal, the United Kingdom, Romania and Norway, while prevalence rates less than 40 % have been found in samples from Belgium, the Czech Republic, Greece, Cyprus, Hungary, Malta, Austria, Slovenia, Finland and the United Kingdom (191).

HCV antibody prevalence data among young IDUs (aged under 25) are available from 14 countries, although in some cases sample sizes are small. There is wide variation in results, with countries reporting both high and low figures from different samples. The highest prevalence rates among young IDUs in 2003–04 (over 40 %) were found in samples from Belgium, Greece, Austria, Poland, Portugal, Slovakia and the United Kingdom, and the lowest prevalence (under 20 %) in samples from Belgium, Greece, Cyprus, Hungary, Malta, Austria, Slovenia, Finland, the United Kingdom and Turkey. Considering only studies of young IDUs with national coverage, the highest prevalence rates (over 60 %) are found in Portugal and the lowest (under 40 %) in Cyprus, Hungary, Malta, Austria and Slovenia. Although the sampling procedures used may be result in bias towards a more chronic group, the high prevalence of HCV antibodies found in a national sample in Portugal (67 % among 108 IDUs under 25 years) is still worrying and may be indicative of continuing high-risk behaviour among young IDUs (see also 'Recent trends in newly reported HIV cases').

Data on HCV antibody prevalence among new injectors (injecting for less than 2 years) are scarce and suffer from small sample size, but they may provide a better proxy indicator of very recent incidence rates than data on young injectors. What information is available for 2003–04 shows that the highest prevalence rate among new injectors (over 40 %) was found in samples from Greece, Poland, the United Kingdom and Turkey and the lowest prevalence (under 20 %) in samples from Belgium, the Czech Republic, Greece, Cyprus and Slovenia. Low prevalence rates have been found in small but national samples of new injectors in Cyprus (only two of 23 injectors tested positive for HCV antibodies, a rate of 9 %), and Slovenia (two out of 32 tested positive, or 6 %).

Hepatitis B

The prevalence of hepatitis B virus (HBV) markers also varies greatly both within and between countries. The most complete data are for anti-HBc, which indicates a history of infection. In 2003–04, prevalence rates of over 60 % among IDU samples were reported from Italy and Poland, while samples with prevalence rates of less than 20 % were recorded in Belgium, Ireland, Cyprus, Austria, Portugal, Slovenia, Slovakia and the United Kingdom. Hepatitis B notification data for the period 1992–2004, for those countries from which data are available, show a very diverse picture (192). In the Nordic region, the great majority of notified acute cases of hepatitis B occur among IDUs, and hepatitis B outbreaks have coincided with increases in drug injecting in several countries. For example, the data for Norway suggest a strongly increasing incidence of hepatitis B infections among IDUs between 1992 and 1998 with a decline thereafter. In Finland, hepatitis B notifications among IDUs have fallen steeply in recent years, possibly as a result of vaccination programmes and a comprehensive needle and syringe exchange system.

Preventing infectious diseases

Effective responses

A number of public health interventions have been shown to be useful in reducing the spread of infectious diseases among drug users and there is a growing consensus that a comprehensive approach to service provision in this area is most likely to be successful. Historically, the debate has largely focused on the prevention of HIV infection among drug injectors, but the need for effective measures to inhibit the spread of hepatitis has increasingly become recognised, as has the need to prevent the spread of infectious diseases among non-injecting drug users.

It must be noted that evidence on effectiveness is strongest for the prevention of HIV infection among drug injectors. A robust body of available evidence shows that interventions can be considered effective and that access to treatment of all types provides protection (Farrell et al., 2005; WHO, 2005). Since the mid-1990s, the European response has been characterised by an increasing provision of drug dependence treatment (see Chapter 2), and overall this appears to be one of the elements that have contributed to the relatively encouraging picture now seen in Europe in terms of epidemic spread of HIV among injectors.

Figure 11 Priority for and extent of use of selected measures to prevent infectious diseases in drug users, according to national expert opinion


* For infectious diseases.

Priority ratings from 23 EU countries plus Bulgaria and Norway. Countries not supplying these ratings were Ireland, Cyprus, Lithuania and the Netherlands.

Ratings on the 'extent of use' were provided by experts from all 25 EU countries plus Bulgaria and Norway. The French and Flemish Communities in Belgium provided separate ratings, bringing the total number of replies to 28.


National focal points expert survey, SQ 23 (2004) question 5.


Treatment is only one part of a comprehensive approach to HIV prevention. Other elements include a range of information, education and communication techniques, voluntary infectious disease counselling and testing, vaccination and the distribution of sterile injecting equipment and other prophylactics. These measures, as well as the provision of medical treatment services at low-threshold agencies, or even sometimes at street level, can help to establish or improve communication with active drug users and their sexual partners regarding the risk of and prevention of health consequences of drug use.

A general commitment to a comprehensive approach does not mean that all these service elements are equally developed or supported at national level. However, some consensus appears to be emerging. In a survey among NFPs, three out of four respondents identified needle and syringe programmes combined with counselling and advice as a priority in the national policy that addresses the spread of infectious disease among drug injectors (Figure 11). That so many countries now explicitly recognise of role of providing clean injecting material as part of their HIV prevention strategy illustrates how this form of provision has become mainstream in most of Europe and is no longer regarded as a controversial issue in most countries. That is not to say that there is uniform agreement on the benefits of this kind of provision. Greece and Sweden, for example, did not rate it as a policy priority, although overall a relatively homogeneous picture emerges across the EU in terms of the implementation of interventions in this area (193), with all countries except Cyprus reporting the existence of programmes for the exchange or distribution of sterile needles and syringes (194).

Types of needle and syringe programmes in European countries

Although most European countries now distribute sterile injecting equipment, the nature and range of provision vary between countries. The most common model is to provide the service in a fixed location, usually a specialised drugs service, but often this type of provision is complemented by mobile services that attempt to reach out to drug users in community settings. Syringe exchange or vending machines complement the available NSP services in eight countries (195), although provision appears to be restricted to a handful of sites, with only Germany and France reporting substantial activities (around 200 and 250 machines respectively). Spain is the only EU country where needle and syringe exchange is regularly available in a prison setting, with provision available in 27 prisons in 2003. The only other EU country reporting activity in this area is Germany, where provision is limited to one prison.

Box 15: Syringe coverage in Europe: is it sufficient?

Although nearly all Member States report some availability of needle and syringe programmes (NSPs), the impact of this kind of intervention depends on the level of provision being adequate to meet the needs of IDUs.

Recent estimates of the number of injecting drug users and on the number of syringes distributed through NSPs are available for nine European countries. From these data, it is possible to make a crude estimation of the annual number of syringes available per injector (1). Based on the most recent data available, coverage rates of NSPs vary considerably, with the number of syringes distributed per estimated IDU per year varying from 2–3 in Greece, through 60–90 in the Czech Republic, Latvia, Austria and Portugal, to approximately 110 in Finland, 210 in Malta and more than 250 in Luxembourg and Norway. In addition, syringes are also available from pharmacies, and data from the Czech Republic and Finland allow overall syringe availability to be estimated. Combining distribution and sales data suggests that in a year drug injectors obtain on average 125 syringes in the Czech Republic and 140 syringes in Finland.

Many factors are known to influence injecting frequency among those using drugs, including patterns of use, level of dependency and type of drug used. A recent study exploring the relationship between HIV prevalence and the coverage of syringe distribution suggested that behavioural factors, e.g. injecting frequency and personal re-use of syringes, strongly influence the level of syringe distribution required to achieve a substantial decrease in HIV prevalence (Vickerman et al., 2006).

The measurement of syringe coverage is an important component of understanding the likely effects of syringe distribution in disease prevention and for assessing unmet needs. However, it is important to take account of the availability of syringes through pharmacies sales (prices, density of pharmacy network) as well as drug injectors’ behavioural patterns and environmental factors in interpreting these data. This issue is further discussed in the 2006 statistical bulletin.

(1) See the 2006 statistical bulletin for technical notes.

Pharmacy-based exchange schemes also help to extend the geographical coverage of the provision and, in addition, the sale of clean syringes in pharmacies may increase their availability. The sale of syringes without prescription is permitted in all EU countries except Sweden, although some pharmacists are unwilling to do so and some will even actively discourage drug users from patronising their premises. Formally organised pharmacy syringe exchange or distribution networks exist in nine European countries (Belgium, Denmark, Germany, Spain, France, the Netherlands, Portugal, Slovenia and the United Kingdom), although participation in the schemes varies considerably, from nearly half of pharmacies (45 %) in Portugal to less than 1 % in Belgium. In Northern Ireland, needle and syringe exchange is currently organised exclusively through pharmacies.

The purchase of syringes through pharmacies may be a major source of contact with the health service for some injectors, and the potential to exploit this contact point as a conduit to other services clearly exists. Work to motivate and support pharmacists to develop the services they offer to drug users could form an important part of extending the role of pharmacies, but to date only France, Portugal and the United Kingdom appear to be making significant investments in this direction.

Mortality and drug-related deaths

Mortality among problem drug users

Most information on mortality among problem drug users in Europe refers to opioid users. Mortality related to other forms of drug use is less well known but remains an important public health issue.

A collaborative study that started within an EMCDDA project examined mortality among opioid users recruited in treatment in eight European locations (196). The study found a very high mortality among opioid users compared with their peers: 6–20 times higher among males and 10–50 times higher among females. It was estimated that in six of the locations (Amsterdam, Barcelona, Dublin, London, Rome and Vienna) 10–23 % of the overall mortality among adults aged 15–49 years could be attributed to opioid use, mainly overdoses, AIDS and external causes (accidents, suicides). Roughly one-third of these drug-related deaths were due to overdoses, although this proportion was higher in cities with a low prevalence of HIV infection among drug injectors, and is likely to increase once highly active antiretroviral treatment (HAART) becomes more widely available.

A mortality cohort study carried out in the Czech Republic found that the mortality of stimulant users was 4–6 times higher (standardised mortality ratio – SMR) than that of the general population, while that of opioid users was 9–12 times higher. A French cohort study that followed individuals arrested for heroin, cocaine or crack use found that male mortality was five times higher and female mortality 9.5 times higher than in the general population, but with a decreasing trend.

As opioid users age, mortality resulting from chronic conditions (cirrhosis, cancer, respiratory diseases, endocarditis, AIDS) adds to mortality due to external causes other than overdoses, such as suicide and violence (Dutch national reports, 2004 and 2005, from Municipal Health Service Amsterdam). The living conditions of drug users (for example, homelessness, mental illness, violence, poor nutrition) may also contribute substantially to the high mortality in this group.

In addition, AIDS related to intravenous drug use accounted for 1 528 deaths in 2002 (197), although this is probably an underestimate. Other causes of drug-related deaths, such as illness (e.g. hepatitis), violence and accidents, are more difficult to assess, but it is likely that they account for an important number of deaths. It has been estimated that 10–20 % of deaths among young adults in European cities can be attributed to opioid use (see above). To this should be added mortality related to other forms of drug use, although this is very difficult to quantify.

Drug-related deaths

Drug-related death is a complex concept. In some reports it refers only to deaths caused directly by the action of psychoactive substances, while in other cases it includes also deaths in which drug use played an indirect or circumstantial role (traffic accidents, violence, infectious diseases). A recent report that analysed the types of harm caused by illegal drug use in the United Kingdom estimated that drug-related death was the main harm related to drug use (MacDonald et al., 2005).

In this section, and in the EMCDDA protocol, the term ‘drug-related deaths’ refers to those deaths caused directly by the consumption of one or more drugs and, generally, occurring shortly after the consumption of the substance(s). Other terms used to describe such deaths include ‘overdoses’, ‘poisonings’, ‘drug-induced deaths’ or 'acute drug deaths' (198).

Between 1990 and 2003, from 6 500 to over 9 000 deaths were reported each year by the EU countries, adding up to more than 113 000 deaths during this period. These figures can be considered as a minimum estimate owing to probable underreporting in many countries (199).

Population mortality rates due to drug-related death varied widely between European countries, ranging from 0.2 to over 50 deaths per million habitants (average 13). In most countries the figure lies in the range of 7–30 deaths per million inhabitants, with rates of over 25 being found in Denmark, Estonia, Luxembourg, Finland, the United Kingdom and Norway. Among males aged 15–39 years, mortality rates are typically three times higher (averaging 40 deaths per million), with seven countries presenting rates over 80 deaths per million. Drug-related deaths accounted for 3 % of all deaths among Europeans aged 15–39 in 2003–04, and for more than 7 % in Denmark, Greece, Luxembourg, Malta, Austria, the United Kingdom and Norway. These figures should be considered minimum estimates, and it should also be taken into account that, despite improvements, there are still important differences in quality of reporting between countries, such that direct comparisons should be made with caution (200).

Opioid deaths

Opioids are present in most cases of ‘acute drug-related deaths’ due to illegal substances reported in the EU, although in many cases other substances are also identified during the toxicological examination, in particular alcohol, benzodiazepines and, in some countries, cocaine. In Europe most cases of opioid deaths are related to heroin, but other opioids play a role (see below) (201).

Opioid overdose is one of the leading causes of death among young people in Europe, particularly among males in urban areas. At present, overdose is also the main cause of death among opioid users in the EU as a whole, in particular in countries with a low prevalence of HIV among injectors (see 'Mortality among problem drug users').

The majority of drug users who overdose are men (202), accounting for 65–100 % of cases, and in most countries the proportion ranges between 75 % and 90 %, with the highest proportion of females in the Czech Republic, Poland and Finland and the lowest in Greece, Italy and Cyprus. These finding must be interpreted in the context of differential rates of opioid use and injecting between men and women.

Figure 12 Proportion of acute drug-related deaths occurring under the age of 25 years in 2002


ONS, Office of national statistics, DSD, drug strategy definition.

2002 was taken as reference as it is the year for which information is available for most countries.


Reitox national reports (2005), taken from national mortality registries or special registries (forensic or police). Based on 'national definitions' as presented in methodological notes on drug-related deaths in the 2006 statistical bulletin.


Most overdose victims are between 20 and 40 years old, with the mean age in most countries lying in the mid-30s (but ranging from 20 to 44 years). The mean age of overdose victims is lowest in Estonia, Slovenia, Bulgaria and Romania and highest in the Czech Republic, the Netherlands, Poland and Finland. There are very few reported overdose deaths among people under age 15 years (17 cases out of a total number of deaths of 7 516, based on most recent available for each country), although drug deaths in this age group could be underreported. The EMCDDA figures include a few deaths among those over 65 years, with only seven countries reporting that more than 5 % of cases fall into this age group (203).

In several new Member States and candidate countries the mean age at death is comparatively low (Estonia, Cyprus, Latvia, Slovakia, Bulgaria and Romania), and there is a high proportion of overdose cases younger than 25 years, which may signal a younger heroin-using population in these countries. The high mean age in the Czech Republic is related to the inclusion of many deaths due to psychoactive medicaments (Figure 12).

In many Member States, the age of overdose victims is increasing, suggesting a decrease in initiation to heroin use among young people. This trend is common in EU-15 Member States and has been observed since the early 1990s, although in Sweden and the United Kingdom it is less marked. In new Member States the trend is less clear and a decrease in mean age is even observed in many cases (204).

Methadone deaths

Several countries reported the presence of methadone in a substantial proportion of drug-related deaths in their 2005 Reitox reports. The terminology used varies between countries, and in some cases it is difficult to determine what role methadone played in the death.

Denmark reported that methadone was the cause of poisoning (alone or in combination) in 44 % of deaths (95 out of 214 in 2004), a similar proportion to 2003, but with a clear increase from 1997; Germany reported that 345 cases were attributed to ‘substitution substances’ (46 alone and 299 with other narcotics in 2004) with a clear decrease since 2002; and the United Kingdom reported 216 cases with ‘mention’ of methadone (England and Wales, in 2003), also with a clear decrease from 2002. Spain reported that there were few overdose cases involving methadone in isolation (2 %), but that it was frequently present in combination in opioid deaths (42 %) and cocaine deaths (20 %). Other countries did not report methadone deaths or the numbers reported were very small. It is unclear what factors lie behind these differences, and whether there is underreporting of cases in some countries (205).

Although research shows that substitution treatment reduces the risk of fatal overdose, it is important to monitor the number of deaths related to methadone and the circumstances surrounding the death (the source of the substance, whether it was consumed in combination with other substances, the point in the treatment process at which intoxication occurred) as part of the quality assurance monitoring for substitution programmes.

Buprenorphine and fentanyl deaths

Deaths due to buprenorphine poisoning appear to be rare, a fact that is attributed to the agonist–antagonist pharmacological characteristics of this drug. However, some deaths have been reported by European countries.

In the 2005 national reports, only France and Finland recorded deaths related to this substance. In Finland, buprenorphine was found in 73 drug-related deaths in 2004, the same number as in 2003, and was generally combined with benzodiazepines, sedatives or alcohol. These high figures parallel increases in buprenorphine treatment in Finland, although the numbers treated are much lower than the estimated 70 000 to 85 000 people receiving buprenorphine in France. It is therefore interesting that in France only four cases of buprenorphine overdose were reported in 2004 (compared with eight cases in 2003). Even taking into account a possible underreporting of poisonings in France, the scale of the differences is striking. In addition to France and Finland, three other countries reported cases (only two or three in each case) of death related to buprenorphine, but without evidence that the substance was the main causal agent.

In previous years deaths due to fentanyl have been reported in the countries surrounding the Baltic Sea, but no such reports were included in the 2005 national reports.

Trends in acute drug-related deaths

National trends in drug-related deaths can provide some insight into developments in patterns of problematic drug use in each country such as heroin epidemics and high-risk behaviours (e.g. injection), as well as treatment provision and even differences in heroin availability. They can also, of course, reflect the success of medical emergency services policies in preventing fatal overdoses (206).

Data available from the EU reveal some general trends in drug-related deaths. Among the EU-15 Member States, a sharp increase was apparent during the 1980s and early 1990s, possibly paralleling the expansion of heroin use and injecting. Drug deaths continued to increase between 1990 and 2000, although less sharply (Figure 13). The total annual number of drug-related deaths in those Member States providing information (most Member States, old and new) increased by 14 %, from 8 054 in 1995 to 9 392 in 2000.

Figure 13 Long-term trend in acute drug-related deaths, 1985–2004


New Member States and candidate countries are not included in this figure owing to the lack of retrospective data in most cases.

See Table DRD-2 in the 2006 statistical bulletin for numbers of deaths in each country and notes on methodology.

Ten countries provided information for 2004 and six did not. Therefore, the figure for 2004 is provisional, based on a comparison of 2003 and 2004 data only for those countries providing data for both years.


Reitox national reports (2005), taken from general mortality registries or special registries (forensic or police).


Since 2000, many EU countries have reported decreases in the numbers of drug-related deaths, possibly related to increases in treatment availability and harm reduction initiatives, although declines in the prevalence of drug use may also be important. At European level, drug deaths fell by 6 % in by 2001, 13 % in 2002 and by 7 % in 2003. Despite these improvements, there were still almost 7 000 reported drug-related deaths in 2003 (data from Belgium, Spain and Ireland are missing). However, among countries reporting data in 2004 (19), there was a small increase of 3 %. Although inferences about 2004 should be made with caution, 13 out of the 19 countries that reported information recorded an increase of some degree.

There is a marked discrepancy between trends in the old and new Member States in the number of deaths among people younger than 25 years. Among the EU-15 Member States there has been a steady decrease since 1996, suggesting a decrease in the number of young opioid injectors, while in new Member States a sharp increase was observed until 2000–02, with an apparent decrease beginning only in 2003 (207).

Gender differences are also observable. The number of deaths in males increased progressively from 1990 until 2000, followed by a clear decrease (a 30 % decline by 2003), whereas the number of reported deaths among females remained roughly stable between 1990 and 2000, oscillating between 1 700 and 2 000 per year, and has fallen by only 15 % since then. This could be due to a number of factors, including differential efficacy of interventions or differences in risk factors between the sexes (208).

In countries with a longer series of data, different patterns of drug-related deaths can be detected. In some countries drug-related deaths peaked during the early 1990s and subsequently decreased, for instance in Germany, where drug-related deaths peaked in 1991–92, Spain (1991), France (1994) and Italy (1991). In other countries, for instance Greece, Ireland, Portugal, Finland, Sweden and Norway, drug deaths peaked later, between 1998 and 2001, and again subsequently declined. In other countries, the pattern was less clear or numbers were stable. Although interpretations should be made with caution, because of the relatively low numbers of drug-related deaths in some countries, these patterns could be related to the trends in heroin injection (209).

Deaths related to ecstasy and amphetamines

Deaths related to ecstasy started to be reported in Europe during the 1990s as the drug became popular. Ecstasy deaths cause considerable concern as they often occur unexpectedly among socially integrated young people.

Information on ecstasy deaths is limited, but data from 2005 Reitox national reports suggest that deaths involving ecstasy remain relatively unusual compared with opioid deaths, although in some countries the number is not negligible. In Europe as a whole, there were references to 77 deaths, which should be considered as a minimum estimate (210). Cases were reported from Denmark (2), Germany (20), France (4), Hungary (3), and the United Kingdom (48 cases with ‘mentions’ – 33 in England and Wales), where probably reporting is better than in other countries. In Spain, ecstasy was present in 2.5 % of drug poisonings.

The issue of the risk of ecstasy use has often been raised. Bearing in mind the margin of error in survey-based estimates of prevalence and the difficulties in reporting drug deaths, dividing the number of fatalities observed by the number of users per year (211) (people at potential risk) yields rates of 5–8 cases and 2–5 cases per 100 000 users in the two countries for which this calculation can be made.

Amphetamine deaths are also infrequently reported, although in the Czech Republic 16 deaths were attributed to pervitin (methamphetamine) in 2004, almost double the 2003 figure, correlating with an increase in the estimated number of problem pervitin users and treatment demands. For GHB deaths see Chapter 4.

Deaths related to cocaine

There is increasing concern about the health risks of cocaine use, following increases in recreational use observed in some countries among young people, among people being treated for addiction and among marginalised populations.

Cocaine use is frequent among opioid users, and it is common to find cocaine in toxicological analyses of opioid overdoses, as well as other substances such as alcohol and benzodiazepines Cocaine is commonly consumed together with alcohol, a combination that may result in increased toxicity

Current statistics available in Europe are limited, and variation in the criteria used to identify cocaine-related cases means that figures are not comparable; in addition, some cocaine-related deaths may go unrecognised or unreported, resulting in underreporting. The data that do exist indicate that many deaths involving cocaine also involve opioids.

Among the countries supplying data, over 400 cocaine deaths were identified in the 2005 national reports; this is a minimum estimate. In most of these cases, cocaine seems to have played a causal role, although this is not always entirely clear from the reports. Nine countries did not explicitly mention the existence or absence of cocaine deaths. Cocaine accounted for 0–20 % of reported acute drug deaths, representing between 10 % and 20 % of such deaths in Germany, France, Spain, the Netherlands and the United Kingdom. Deaths in which cocaine played a causal role (alone or in combination) were reported by Germany (166), Spain (53), France (14), Netherlands (20), and United Kingdom (142 ‘mentions’ – 113 in England and Wales). Nine other countries reported from zero to two cases. In addition, cocaine is commonly found in toxicological analysis of opioid overdoses in some countries. From the limited data available, it is difficult to identify trends with certainty, but an increasing trend seems to exist in all countries with larger number of cases, i.e. Germany, Spain, France, the Netherlands and the United Kingdom, although in the Netherlands increases have halted in the last 2 years.

In addition, cocaine may be a contributor to deaths due to cardiovascular problems (arrhythmias, myocardial infarction, cerebral haemorrhages), particularly in users with predisposing conditions or risk factors (tobacco, hypertension, angiomas) or with increasing age. Many of these cases may go unnoticed at present because of a lack of awareness. Further research is needed in this area.

Reducing drug-related deaths

Effective responses

Reaching out to untreated populations of drug users and establishing links for communication is a precondition for risk education and management, and for mediating access to services, including treatment.

Research into the circumstances of overdoses has supported the development of interventions that target high-risk situations or high-risk individuals. Such measures may achieve an important reduction in the deaths attributable to the immediate effects of drug taking. The role of different interventions in reducing acute drug-related overdose deaths was summarised in a recent EMCDDA policy briefing (EMCDDA, 2004d).

As most overdose deaths in Europe involve heroin, increasing the proportion of heroin users in treatment can be viewed as an overdose prevention measure. A number of factors may be responsible for recent modest reversals in the trend in overdose deaths observed in some Member States. These include decreases in prevalence and injecting rates, increased prevention efforts, increased availability and uptake of, and retention in, treatment and possibly reductions in risk-taking behaviour.

Profile of responses

In most countries, expert opinion on the use of different intervention strategies to reduce drug overdose deaths considers opioid substitution treatment to be the most valuable approach (212). In Hungary and Sweden, although this type of treatment is available, it is not considered a means of reducing drugs deaths. And in Estonia and Poland, the low level of substitution treatment provision means that methadone substitution treatment is not currently considered a major response to reduce overdose deaths.

Information, education and communication (IEC)-oriented responses are further important measures in most European countries. The dissemination of risk awareness messages and overdose management instructions via specifically developed printed materials or other media (flyers, websites, mass media campaigns) is common or predominant in 19 countries. However, seven countries (Estonia, France, Ireland, Latvia, Hungary, Malta, Finland) use such measures rarely, and one country (Sweden) not at all.

According to the NFPs, the approach of systematically integrating an individual risk assessment into counselling and treatment routines and organising group sessions on risk education and response for drug users is less common.

A broad category of activities can be defined as ‘prison pre-release interventions’. These ranged from simple information dissemination, through counselling on overdose risks and prevention, to initiation or continuation of substitution treatment in prison. However, activities falling into this spectrum of responses were rarely used in 13 countries and not in use at all in another five (Latvia, Hungary, Poland, Romania and Sweden). In Spain, Italy and the United Kingdom, prison interventions are among the predominant approaches to a reduction in acute drug deaths.

Local conditions of risk related to public injecting have led to the opening of professionally supervised drug consumption facilities in four EU countries and Norway (213). Their target groups are highly marginalised and risk-prone street injectors (EMCDDA, 2004c).

(186) See Figures INF-2 (part i) and INF-2 (part ii) in the 2006 statistical bulletin.

(187) See Figure INF-5 (part ii) and (part vi) in the 2006 statistical bulletin.

(188) See Figure INF-3 (part v) in the 2006 statistical bulletin.

(189) See Figure INF-14 (part iii) and (part iv) in the 2006 statistical bulletin.

(190) See Figure INF-1 (part i) in the 2006 statistical bulletin.

(191) See Figure INF-6 (part i) in the 2006 statistical bulletin.

(192) See Figure INF-12 (part i) in the 2006 statistical bulletin.

(193) See Figure NSP-3 in the 2006 statistical bulletin.

(194) See the 2005 annual report for a brief summary of the evidence on effectiveness of needle and syringe programmes (p. 68).

(195) See Table NSP-2 in the 2006 statistical bulletin.

(196) Amsterdam, Barcelona, Dublin, Denmark, Lisbon, London, Rome and Vienna. Estimates of population mortality did not include Lisbon and Denmark. See EMCDDA (2002b).

(197) See EuroHIV (2005). The figure refers to the West and Centre areas of WHO Europe, which include some non-EU countries, and the total deaths for Estonia, Latvia and Lithuania (East area).

(198) This is the agreed common definition by the EMCDDA group of national experts. At present, most national case definitions are the same as the EMCDDA definition or very similar, although some countries still include cases due to psychoactive medicines or non-overdose deaths, generally as a limited proportion (see the 2006 statistical bulletin methodological note 'Drug-related deaths summary: definitions and methodological issues’. Section 1: EMCDDA definition and Section 2: National definitions and 'DRD Standard Protocol, version 3.0').

(199) See Tables DRD-2 (part i), DRD-3, DRD-4 in the 2006 statistical bulletin.

(200) See Table DRD-1 (part iii) and (part iv) in the 2006 statistical bulletin.

(201) See Figure DRD-1 in the 2006 statistical bulletin.

(202) As most cases reported to the EMCDDA are opioid overdoses, general characteristics of acute drug-related deaths are used for description of opioid cases.

(203) See Table DRD-1 in the 2006 statistical bulletin.

(204) See Figures DRD-3 and DRD-4 in the 2006 statistical bulletin.

(205) An ongoing EMCDDA field trial aims to improve the quality of information provided on substances involved in drug-related deaths, including substitution substances.

(206) See Figure DRD-7 in the 2006 statistical bulletin.

(207) See Figure DRD-5 in the 2006 statistical bulletin.

(208) See Figure DRD-6 in the 2006 statistical bulletin.

(209) See Figure DRD-7 in the 2006 statistical bulletin.

(210) Depending on country, figures refer to 2003 or 2004, for ecstasy and cocaine.

(211)  Last 12 months’ use in population surveys.

(212) Results based on a survey conducted through 27 NFPs in 2004. The instrument can be downloaded at

(213) The EU countries are Germany, Spain, Luxembourg and the Netherlands.