Methodology (FAQ drug overdose deaths in Europe)
Introduction
The national focal points and their experts in the 27 EU Member States, Norway and Türkiye report data annually to the EUDA through standard tables and workbooks. The annual Statistical Bulletin contains the most recent available data on the drug situation in Europe provided by the EU Member States, Norway and Türkiye. These data sets underpin the analysis presented in the European Drug Report. This section provides information on the methods, sources and definitions used to monitor drug-induced deaths in Europe, as well as the strengths and limitations of the surveillance in place. This section also provides methods and references to recently published findings on the overall mortality – due to drug-induced deaths and other causes – among people who use drugs.
This page is part of the Frequently asked questions (FAQ): drug overdose deaths in Europe.
Where do the data come from?
There are two kinds of mortality registers from which cases of drug-related deaths can be retrieved and reported: general mortality registers and special mortality registers.
- General mortality registers are usually maintained by national statistical offices or health departments. They are based on the mandatory death certificates issued for all deaths by a certifying doctor. In cases of death reported for legal scrutiny, the certifying person is usually a forensic doctor or a coroner. These registers are limited, as often they do not have specific information on which drugs were consumed. Limitations are also due to the use of broad categories such as ‘other synthetic opioids’ and ‘other opioids’ in the coding of the cause of death in these registers.
- Special mortality registers are ideally developed specifically for drug mortality monitoring through a combination of different sources (e.g. forensic, police and other sources), which allows a high degree of detection of drug-induced deaths in a country. Alternatively, these registers are included in and maintained by existing information systems of police or medico-legal institutions (e.g. forensics institutes, coroners) for all unnatural deaths that required investigation.
What cases are included?
The European definition of drug-induced deaths is translated into a set of codes and criteria to allow the extraction of the relevant cases from mortality registers.
The extraction of cases from general mortality registers is explained in detail in the DRD methods page of the Statistical Bulletin. Cases are counted when their underlying cause of death is mental and behavioural disorders due to psychoactive substance use or poisoning (accidental, intentional or by undetermined intent) (EMCDDA, 2010).
What sources of information are used in different countries?
Most countries report data from both sources (general and special registers – see the map below), which allows triangulation and validation of the data.
While the trends are usually consistent between both sources (as can be seen in the Statistical Bulletin), there can be discrepancies between the general mortality register and the special mortality register data in some countries. This can be because the coding of the causes of deaths in the general mortality register is not sufficiently sensitive, or because the data reported from the special register do not cover the whole country.
Depending on the certification and coding procedures in the registries, and the flux of information between the special and the general mortality registries, there are reporting delays in some countries. The most recent data are from 2022 in 18 countries, 2021 in 7 countries, 2020 in 3 countries, 2017 in 1 country (see the map below).
Figure 3.1. Sources used by the countries to report drug-induced deaths to the EUDA, 2022 (or most recent data available)
What is the preferred source of data chosen by the countries for use in EUDA publications?
When both sources are available, the countries are asked to choose which is more valid and which should be taken as the reference for the EUDA Statistical Bulletin. The data from the selected source are then used to compute the European figures and European trends.
Meanwhile, at national level, figures from both sources (the selected source or ‘national definition’ and the other source) are given in the Statistical Bulletin.
Eighteen countries indicated that the general mortality register was their preferred source, while 11 preferred the special register (see the map below). Twelve countries report data from both sources.
In the Statistical Bulletin, when the EUDA definition is fully applied, the cases extracted from the general mortality register are shown under the category ‘EMCDDA definition for the general mortality registers’ and those extracted from the special mortality registers are shown under the category ‘EMCDDA definition for the special mortality registers’ (1).
(1) On 2 July 2024, the European Monitoring Centre for Drugs and Drug Addiction (EMCDDA) became the European Union Drugs Agency (EUDA).
Figure 3.2. Sources preferred by the countries to report drug-induced deaths to the EUDA, 2022 (or most recent data available)
Are the data comparable between countries?
The differences in drug-induced death rates between countries may be explained by several factors, including differences in the size of the population at risk, differences in drug taking and risk-taking patterns – including injecting drug use – differences in practices among drug users, and also differences in age distribution and health status among drug users. There are also differences in access to health care, drug treatment and harm-reduction interventions.
In addition to these, there are varying rates of post-mortem examinations, varying quality in the data available about toxicology and mortality, and varying practices in the coding of the cause of death, including the use of ‘unknown cause of death’. These are some known influential factors that may impact on the comparability of the data.
In particular, there are differences between (and within) countries with regard to the identification and certification of the cause of death, and the classification and reporting of the number of drug-induced deaths.
This relates to:
- the quality and frequency of post-mortem investigations;
- the availability of this information for the determination and codification of the cause of death;
- the classification system used;
- the quality of classification;
- the coverage and quality of the overall reporting system.
Countries also differ in their levels of forensic laboratory capacity and standard procedures for post-mortem toxicological investigations of suspected drug-induced deaths (EMCDDA, 2019a, 2021c; Leifman, 2017; Millar and McAuley, 2017). These factors affect the sensitivity of analyses and therefore the comparability of the data within and between countries.
Caution is thus advised when interpreting and comparing data on drug-induced deaths over time and between countries.
Insights into the differences in post-mortem toxicology investigations are available in an EMCDDA publication (2019a).
How do we measure the overall and the cause-specific mortality rates related to drug use?
Monitoring overall mortality among high-risk drug users is the second component of the DRD epidemiological indicator. The overall or ‘all-cause’ mortality among high-risk drug users is investigated by means of cohort studies, which link records from death registers with records of individuals – typically from treatment registers – who are or have been using drugs (see the guidelines for carrying out, analysing and reporting key figures (EMCDDA, 2012)). In this way, it is possible to check the vital status of the individuals who enrolled in treatment at some point. Follow-up cohort studies also allow measurements of behaviour over time, for example the duration of treatment and interruptions, if any.
Mortality cohort studies measure the overdose mortality risk among drug users. This risk can, in turn, be applied in the estimated population of high-risk drug users in countries, in combination with the reported number of drug-induced deaths. Overdose mortality rates from cohorts can allow researchers to estimate the ‘expected’ number of overdose deaths in countries and can help cross-validate the reported number of overdoses in the national statistics (2).
The findings of a provisional mapping conducted in 2021 suggest that, for the majority of countries, there are some recent data based on cohort studies. More information is also available from previous reviews of cohort studies published by the EMCDDA (2015, 2019b).
Recent studies are available from Greece (Roussos et al., 2024), Italy (Pavarin et al., 2023), Norway (Eide et al., 2023; Skeie et al., 2022). These studies confirm the available evidence that people who use drugs have an excess risk of mortality compared to their peers of the same age and sex in the general population (Degenhardt et al., 2019; Peacock et al., 2021; Santo et al., 2021). Most of the excess risk of mortality relate to drug-induced deaths. However, other causes of morbidity and mortality, including cancer, are more common among people who use drugs compared to the general population. In Norway, cancer incidence and cancer-related mortality appear to be elevated among individuals with opioid use disorders. The study also showed that the average 1-year survival rate after a cancer diagnosis was lower in liver, pancreas and colon cancer among these people compared to the survival rate observed in patients in the general population, ranging from 10 % to 15 % less than that of the general population (Kostovski et al., 2024).
Figure 3.3. Mortality cohort studies among people using drugs in Europe: countries with studies conducted within the last 10 years or earlier
Note: As of June 2024, provisional findings.
(2) More information on the methods was presented during the annual DRD expert meeting in 2019 (Vicente and Giraudon, 2019).
Resources and references
Consult the methodological information and list of references on the Frequently asked questions (FAQ): drug overdose deaths in Europe main page.
Source data
The data used to generate infographics and charts on this page may be found below.