CommentaryMissing targets on drugs-related deaths, and a Scottish paradox
Introduction
A new drug strategy for England and Wales (2008–18) was published in February 2008 (HM Government, 2008). Drugs-related deaths (DRDs) were dropped as a key performance indicator, perhaps because the previous target – to reduce DRDs by 20% between 1999 and 2004 – had been missed (Morgan et al., 2006, Report, 2007). Only a 9% reduction could be claimed (Morgan et al., 2006) when DRDs were defined in accordance with recommendations by the Advisory Council on the Misuse of Drugs (2000) (ACMD).
Scotland's new drugs strategy, Road to Recovery, was published in May 2008 (The Scottish Government, 2008a). In the wake of Scotland's confidential inquiry into DRDs in 2003 (Zador et al., 2005) and specific recommendations in 2007 by a National Forum on Drug-Related Deaths (2007), DRDs have remained firmly in focus with the The Scottish Government, 2008a, The Scottish Government, 2008b. A newly designed national database was launched in January 2009 for collecting and analysing standardised information on all Scotland's DRDs.
According to ACMD's recommended definition, Scotland's DRDs numbered 1006 and 1009 in the 3-year periods of 2000–02 and 2003–05, respectively; but 876 in the most recently published two years (2006 + 2007), which gave a total of 2891 for the 8-year period of 2000–07 (Information Services Division, 2008). For England and Wales, according to the same definition but by the registration year of deaths, DRDs at all ages were 5022 in 2000–02, lower at 4534 in 2003–05; and 3177 in the most recently published two years (2006 + 2007), which gave a total of 12,733 for 2000–07 (Report, 2008).
Most DRDs are opiate-related, and they mostly occur in injecting drug users (IDUs). We therefore delve deeper—firstly, to analyse by age-group because the epidemic history of injection drug use in the UK requires it (Bird, Hutchinson, & Goldberg, 2003) and, secondly, to address a Scottish paradox about injectors’ DRD rate having apparently increased in 2003–05.
By analysing within age-group (Bird et al., 2003), we show, first, that the above 3-year totals obscure public health success, in terms of very significant reductions in UK's DRDs at younger ages. However, we also reveal the late epidemiological consequence of UK's national epidemics of injecting drug use in the 1980s. Morgan, Vicente, Griffiths, and Hickman (2008) recently, and rightly, called for a better international understanding of the mechanisms that underlie overdose mortality statistics than mere age-standardisation to a common European population allows (Morgan et al., 2006). Age-standardisation masks, rather than reveals, trends which are importantly different by age.
The differences by age that we reveal have their origin in the epidemic nature of injecting drug use. The number of current IDUs was estimated to have increased 6-fold in Scotland between 1980 and 1989 (Hutchinson, Bird, Taylor, & Goldberg, 2006). The corresponding increase in England was at least 4-fold between 1987 and 1996 (De Angelis, Hickman, & Yang, 2004), and so somewhat delayed. These time periods, while not exact and subsuming of sub-national variation, are sufficiently different for us to anticipate that any sharp 1980s-related increase in DRDs in older-age IDUs may be relatively delayed in England compared to Scotland (and possibly also less sharp). The Health Protection Agency (2005) refers also to an earlier (smaller) IDU epidemic prior to 1970 in England and Wales.
The Scottish paradox arose as follows. Because DRDs are mainly opiate-related and occur mostly in IDUs, there is reason to estimate DRDs rates per 100 current IDUs (Bird et al., 2003), a relevant denominator, as recommended by ACMD (2000). Capture–recapture estimates of Scotland's current IDUs in 2000 (see Bird et al., 2003, Hay et al., 2001, King et al., 2005) and again in 2003 (Hay et al., 2005, King et al., 2008) were publicly reported (Scottish Government News Release, 2005) as suggesting that the prevalent number of IDUs had decreased from 25,000 in 2000 to nearer 19,000 by 2003. Since Scotland's DRDs had been 1006 in 2000–02 and 1009 in 2003–05, we were concerned by the apparent paradoxical implication that Scotland's DRD rate per 100 current IDUs should have increased in 2003–05, and this despite enhanced access to methadone maintenance, including in prisons (Information Services Division, 2006, The Scottish Government, 2008a, The Scottish Government, 2008b).
To address this Scottish paradox, we here set Scotland's markedly increased DRDs in 2006 + 2007 against the backdrop of previously published, credibly estimated DRD rates, by sex and age-group, for Scottish injectors in 2003–05 (King et al., 2008). The rates were derived from a Bayesian capture–recapture approach to estimating the number of Scotland's current IDUs in 2003, which had given a higher central estimate of around 26,000 IDUs (King et al., 2008).
Section snippets
Our approach
Scotland's DRDs in 2000–02 and 2003–05 are summarised in eight pre-defined (Bird et al., 2003, King et al., 2005) subgroups (sex by age-group by region) as in our Scottish work hitherto. Thus, prior to Bird et al. (2003), the age-group (15–34 years; 35+ years) and regional subdivisions (Greater Glasgow; elsewhere in Scotland) were specifically selected on the basis of knowledge of Scotland's IDU epidemic. They have been used by us since then to demonstrate that, as anticipated by Scotland's IDU
Does Scotland's historical injector epidemic determine its age-related increase in DRDs?
Table 1 shows that DRDs under 35 years of age decreased very significantly (p = 0.005) between eras: by 15%, from 672 in 2000–02 to 572 in 2003–05. But DRDs at 15–34 years rose again to 466 in 2006 + 2007, significantly up on the expectation of 381.3 based on 2003–05 (p < 0.001). A similar rise, see below, was not evident for the youngest age-group in England and Wales.
Epidemiologically unsurprisingly is that Scotland's DRDs at 35+ years of age have increased dramatically between eras: by 31%, from
Discussion
Major sources of uncertainty hinder a full understanding of injectors’ DRD rates, namely: which DRDs actually pertain to IDUs, toxicology (whether opiate-related or not; and co-presence of other drugs), and how influential the demographic factors of sex, older age and region actually are on DRD rates (Bargagli et al., 2006, Bird et al., 2003, King et al., 2005, Morgan et al., 2008).
For England and Wales, as well as for Scotland, there was good news in 2000–05. Public health measures were
Acknowledgements
We are grateful to Mr. Graham Jackson, formerly at General Register Office for Scotland, and his successor Mr. Frank Dixon for annually providing to us key cross-tabulations of Scotland's drugs-related deaths by age-group, sex and region which enable our estimation of drugs-related deaths by subgroup.
Funding: Medical Research Council programme (WBS number): U.1052.00.002.00001.01. Authors also belong to the Medical Research Council NIQUAD. Cluster for quantitative understanding of addiction
References (27)
- et al.
Drug-related deaths by region, gender and age-group per 100 injectors: Scotland, 2000–2001
Lancet
(2003) - et al.
Estimating the prevalence, incidence and cessation of injecting drug use in Glasgow 1996–2000: Combining expert opinion with capture–recapture prevalence data
International Journal of Drug Policy
(2006) - Advisory Council on the Misuse of Drugs (chairman: Professor Sir Michael Rawlins). (2000). Reducing drug related...
- Advisory Council on the Misuse of Drugs (chairman: Professor Sir Michael Rawlins). (2008). Cannabis: Classification and...
- et al.
Drug-related mortality and its impact on adult mortality in eight European countries
European Journal of Public Health
(2006) - Bird, S. M., Cox, D. R., Farewell, V. T., Goldstein, H., Holt, T., & Smith, P. C. (2005). Performance indicators: Good,...
- et al.
Estimating long-term trends in the incidence and prevalence of opiate use/injecting drug use and the number of former users: Back-calculation methods and opiate overdose deaths
American Journal of Epidemiology
(2004) - Hay, G., Gannon, M., McKeganey, N., Hutchinson, S., & Goldberg, D. (2005). Estimating the National and Local Prevalence...
- Hay, G., McKeganey, N., & Hutchinson, S. (2001). Estimating the National and Local Prevalence of Problem Drug Use in...
- Health Protection Agency Centre for Infections. (2005). Hepatitis C in England: The First Health protection Agency...
Modelling the current and future disease burden of Hepatitis C among injecting drug users in Scotland
Hepatology
Cited by (8)
Non drug-related and opioid-specific causes of 3262 deaths in Scotland's methadone-prescription clients, 2009–2015
2019, Drug and Alcohol DependenceCitation Excerpt :There is strong evidence that opioid users’ DRD-risk increases with age; and that younger females are at much lower DRD-risk than young males, an advantage which diminishes with age. In the UK, evidence of these demographic influences is consistent across a) evidence-syntheses, b) cohort studies and c) official statistics (Bird et al., 2010; Merrall et al. (2012); King et al. (2014, 2013); Pierce et al. (2015a, 2015b); White et al., 2015; Gao et al., 2016; National Records of Scotland (NRS), 2018). More recently, Gao et al. (2016) and Pierce et al. (2018) showed that the risk of methadone-specific DRD (that is: methadone implicated in DRD but neither heroin nor buprenorphine) rises steeply with age, quadrupling at 45+ years of age compared with 25–34 years.
Experimental and epidemic risks: Matters of science and judgement
2013, European ReviewRisk factors for mortality, hospitalisation and imprisonment in substance misuse patients
2018, Clinical PharmacistBereavement following a fatal overdose: The experiences of adults in England and Scotland
2017, Drugs: Education, Prevention and PolicyInjecting drug users in Scotland, 2006: Listing, number, demography, and opiate-related death-rates
2013, Addiction Research and Theory
- 1
Department of Mathematics and Statistics at University of Strathclyde, United Kingdom.