Elsevier

The Lancet

Volume 362, Issue 9388, 20 September 2003, Pages 941-944
The Lancet

Articles
Drug-related deaths by region, sex, and age group per 100 injecting drug users in Scotland, 2000–01

https://doi.org/10.1016/S0140-6736(03)14362-0Get rights and content

Summary

Background

In 2000 and 2001, there were 292 and 332 drug-related deaths in Scotland, respectively. Of the 332 people who died in 2001, 65 were of female sex, 80 were younger than 25 years, and 112 were older than 34 years. We assessed the effect of region, sex, and age group on drug-related deaths in Scotland per 100 injecting drug users.

Methods

We used published regional estimates of current injecting drug users (n=22 805), their sex (30% female) and age distribution (8% older than 34 years in mid-1990s), or corresponding data for problem drug users (21% older than 34 years in 2000) to compare observed with expected drug-related deaths by region, sex, and age group per 1000 population or per 100 injecting drug users.

Findings

Regional population-based expectations fitted poorly to observed drug-related deaths, but those based on injecting drug users fitted tolerably. Drug-related deaths in 2001 per 100 injecting drug users were significantly lower in female (0·9 [95% CI 0·6–1·1]) than in male drug users (1·5; 1·3–1·7). They were between two and six times more frequent per 100 injecting drug users older than 34 years than younger than 25 years, depending on whether current injecting drug users' age distribution was assumed to be the same as problem drug users' or as injecting drug users' in the mid-1990s (2·1 [1·7–2·5] or 5·3 [4·3–6·3], respectively).

Interpretation

Older and male injecting drug users were at highest risk of drug-related death. Injecting drug users' age distribution should be assessed and compared across countries.

Introduction

Hall and Darke1 reported trends in opiate overdose deaths per million population in Australia during 1975–95 by region, sex, and age group. 85% of 953 heroin-related deaths in New South Wales during 1992–96 were in drug users of male sex; 2 only ten (1%) deaths resulted from non-injection routes of heroin administration.3 50 (5%) deaths had occurred within 1 month of release from prison.2 2 years earlier, Seaman and colleagues4 had quantified male, HIV-infected injecting drug users' overdose deaths in the first 2 weeks after release from prison as eight times more likely than in comparable other fortnights at liberty.4 More recently, Bird and Hutchinson5 estimated that drug-related deaths in the fortnight after release from 14 or more days' incarceration accounted for one in nine drug-related deaths in the 15–35-year-old male population in Scotland during 1996–99.

The UK Advisory Council on the Misuse of Drugs6 called for internationally comparable research on drug-related mortality rates and hinted that the preferred denominator for such comparisons was the number of current injecting drug users. The validity of denominators based on numbers of injecting drug users is strongly supported by non-injection routes causing only 1% of heroin-related deaths.

We investigated drug-related deaths in Scotland in 2000 and 2001 per 100 current injecting drug users (defined as injecting illicit, non-prescribed drugs) by region, sex, and age group, using regional injecting drug user prevalences in 2000;7 the sex8, 9 and age distribution7, 8 of injecting drug users and of problem drug users— defined as misusers of opiates or benzodiazepines7—and drug-related deaths in Scotland for 1996–200110 according to the consistent definition across England, Wales, Scotland, and Northern Ireland.6, 10

Section snippets

Methods

We describe key sources of numerators10 and denominators7, 8, 9 that we used to relate drug-related deaths in Scotland in 2001 to regional health board, sex, and age-group specific estimates of current injecting drug users.

Results

Of 332 drug-related deaths in Scotland in 2001, 65 (20%) were in girls or women, 80 were in people younger than 25 years, 140 in people aged 25–34 years, and 112 (34%) in people older than 34 years.

Table 1 shows the observed numbers of drug-related deaths in 2000 and 2001 by health board area, the regional population aged 15–54 years in mid-2000, and Hay and colleagues'7 estimated numbers of current injecting drug users for 2000. Table 1 also shows the expected numbers of drug-related deaths by

Discussion

Our results show the importance of using an appropriate denominator for drug-related deaths. This finding has been presaged in comparisons between Greater Glasgow and the rest of Strathclyde of methadone-related deaths per 100 patient-years of treatment.6, 15

Regional variation in drug-related deaths in Scotland in 2000 and 2001 was reduced, but not wholly resolved, when drug-related deaths were expressed per 100 injecting drug users rather than per thousand population. Because Hay and colleagues

References (18)

  • HallW et al.

    Trends in opiate overdose deaths in Australia 1979–1995

    Drug Alcohol Depend

    (1998)
  • DarkeS et al.

    Heroin-related deaths in New South Wales, Australia, 1992–1996

    Drug Alcohol Depend

    (2000)
  • DarkeS et al.

    Fatal heroin overdoses resulting from non-injecting routes of administration, NSW, Australia, 1992–1996

    Addiction

    (2000)
  • SeamanSR et al.

    Mortality from overdose among injecting drug users recently released from prison: database linkage study

    BMJ

    (1998)
  • BirdSM et al.

    Male drugs-related deaths in the fortnight after release from prison: Scotland, 1996–99

    Addiction

    (2003)
  • Reducing drug related deaths

    (2000)
  • HayG et al.

    Executive report: estimating the national and local prevalence of problem drug use in Scotland

    (2001)
  • HutchinsonSJ et al.

    Extent and contributing factors of drug expenditure in injectors in Glasgow: multi-site city-wide cross-sectional study

    Br J Psychiatry

    (2000)
  • BirdSM et al.

    Projecting severe sequelae of injection-related hepatitis C virus epidemic in the UK: part 1—critical hepatitis C and injector data

    J Epidemiol Biostatistics

    (2001)
There are more references available in the full text version of this article.

Cited by (43)

  • Epigenome-wide association analyses of active injection drug use

    2022, Drug and Alcohol Dependence
    Citation Excerpt :

    In addition, there were an estimated 5.9 million cocaine users, 745,000 heroin users, and 10.1 million opioid misusers (including the misuse of prescription pain relievers) in 2019 in the US (Substance Abuse and Mental Health Services Administration, 2020). Active injection drug use (IDU) is associated with substantial risk of blood-borne infections, morbidity, and mortality (Bargagli et al., 2001; Bird et al., 2003). Studies have shown that people who inject drugs (PWID) have a higher risk of HIV acquisition than the general population (Schoenbaum et al., 1989) and worse HIV outcomes during treatment initiation compared with people who do not inject drugs, primarily due to treatment non-adherence.

  • Increasing drug-related mortality rates over the last decade in Scotland are not just due to an ageing cohort: A retrospective longitudinal cohort study

    2021, International Journal of Drug Policy
    Citation Excerpt :

    The last covariate was included because Scotland's HCV Action Plan Phase II (launched in May 2008) encouraged former PWID to be tested and led to increased diagnosis rates (McLeod et al., 2014). As previous research observed strong effect modification by sex of the association between age and drug-related mortality (Bird, Hutchinson, & Goldberg, 2003; Gao et al., 2016; Gao, Robertson, & Bird, 2020; Pierce et al., 2015; White, Bird, Merrall, & Hutchinson, 2015), we investigated a model variant specifying a sex x attained age-group interaction term. A linear time trend was also tested in place of the categorical variable for calendar time, with model fits compared using a likelihood ratio test.

  • Effect of ageing and time since first heroin and cocaine use on mortality from external and natural causes in a Spanish cohort of drug users

    2018, International Journal of Drug Policy
    Citation Excerpt :

    Moreover, it is generally accepted that fatal and non-fatal overdose occur not only among the youngest and most inexperienced opioid users, but also in older users with many years of drug use (Darke & Hall, 2003; UNODC, 2015; Warner-Smith, Darke, Lynskey, & Hall, 2001). Many studies have found that fatal overdose risk does not decrease with age (Bartu, Freeman, Gawthorne, Codde, & Holman, 2004; Bird, Hutchinson, & Goldberg, 2003; Buster, van Brussel, & van den Brink, 2002; Clausen et al., 2009; Merrall, Bird, & Hutchinson, 2012; Odegard, Amundsen, & Kielland, 2007; Pierce et al., 2015) or time since first opioid use (Odegard et al., 2007). Mortality risk from natural causes usually increases with age, especially after age 50 (Brugal et al., 2005; Clausen et al., 2009; EMCDDA, 2015; Larney et al., 2015; Pierce et al., 2015), the same as in the GP, whereas fatal-injury risk shows inconsistent results between studies, with increases (Clausen et al., 2009), stability (Larney et al., 2015; Pierce et al., 2015), and decreases (Copeland, Budd, Robertson, & Elton, 2004; Degenhardt et al., 2014) with age.

  • Mortality of those who attended drug services in Scotland 1996-2006: Record-linkage study

    2012, International Journal of Drug Policy
    Citation Excerpt :

    The age effect on DRD risk appeared specific to HCV-diagnosed individuals, consistent with the hypothesis that HCV-related liver damage enhances the risk of overdose (Darke, Degenhardt, & Mattick, 2007). The observed interaction may explain conflicting evidence on how strongly age influences DRD risk if there is confounding by HCV diagnosis (Bird, Hutchinson, & Goldberg, 2003; COSMO Workshop, 2010). The SDMD cohort was sufficiently large to investigate HR influences on other major causes of death besides DRDs, and for all non-major causes.

  • Self-reported drug use and mortality among a nationwide sample of Swedish conscripts - A 35-year follow-up

    2011, Drug and Alcohol Dependence
    Citation Excerpt :

    The large differences in mortality risk among male drug users in the general population show that the high mortality found in clinical studies cannot be extrapolated to the general population. The HR of 2.57 for death found in the multivariate analysis of drug users with repeated drug injections confirms earlier findings of an increased risk of death associated with injection (Bird et al., 2003; Copeland et al., 2004; van Haastrecht et al., 1996). Among cannabis users at conscription, about 5.5% were hospitalised with a drug-related diagnosis, which may indicate that, for some youths, use of cannabis precedes more severe drug use (Degenhardt et al., 2010; Fergusson et al., 2008; Fuller et al., 2001; Stenbacka et al., 1992a,b), poly drug use (Degenhardt et al., 2001), and mortality (Calabria et al., 2010).

View all citing articles on Scopus
View full text