A review of drug-related deaths in seven Northern European countries

Transcription

1 A review of drug-related deaths in seven Northern European countries Tim Millar a & Andy McAuley b a) Division of Psychology & Mental Health, School of Health Sciences, The University of Manchester b) BBV/STI/VPD Group, Health Protection Scotland & Glasgow Caledonian University, School of Health and Life Sciences

2 Background/purpose: Examine the triggers and dynamics of Drug-Related Deaths (DRD) in seven EU countries (Sweden, Norway, Scotland, Finland, Denmark, Estonia, and Ireland) EMCDDA selected countries that: o may have high, and in some cases increasing, numbers of recorded DRD o general population rate for DRD above the EU average (EMCDDA, 2016) Country profiles, developed with national experts, relevant, available EMCDDDA indicators All seven countries: opioids implicated in 80%-90% of DRD; therefore, opioid-related DRD were the primary focus Considered: o Differences (and trends) in the number of drug users at risk of DRD o Differences (and trends) in factors that may influence the risk of DRD (among those at such risk) o Differences in /changes to mechanisms to record DRD (Sweden) Scotland Sweden Estonia Ireland Denm ark Finland Norway

3 DRD Rates in European Countries: (European Drug Report 2016): NB: these are general population DRD rates. and do not account for variation in the size (and rate) of the populations at risk of DRD.

4 Trends in the number of (recorded) DRDs (opioids): : Denmark Estonia Finland Ireland Norway Scotland Sweden NB: 2014 data were the latest available at the time of the work.

5 Proportional increase in (recorded) DRDs (opioids): anchored on Sweden Norway Ireland Finland Denmark Scotland ST NB: 2014 data were the latest available at the time of the work.

6 The extent of DRD is the product of: Prevalence: the size of the population specifically at risk of DRD Available (albeit flawed) estimates indicate a 30x difference in prevalence rates across EU countries for the main at risk group (EMCDDA, 2016): so, we should expect to observe differences in DRD general population rates! Risk: the level of risk experienced by the at risk group Are users in one place or time more or less likely to suffer a fatal overdose? A much more interesting question. and much more difficult to answer A wide range of (interacting) factors drive prevalence and risk

7 Drivers of DRD and interactions thereof: Enforcement (& situation in producer/transit countries - climate change, war, political instability, etc) Behaviour Demographics Naloxone, treatment, OST, needle exchange, health interventions Availability Prevalence X Risk = Deaths

8 Which denominator, population size or prevalence? The DRD general population rate is informative re comparison of the mortality burden attributable to DRD... but is not informative re comparison of levels of DRD risk.

9 Mean number of (opioid) DRD ( ) vs. best (gu)estimates of problem opioid prevalence (or proxy): Scotland Mean n deaths Estonia Norway Finland Denmark Ireland Sweden Prevalence (n)

10 Available HRDU/PDU prevalence trend estimates

11 Example: available prevalence estimates for HRDU/PDU in Sweden: Problematic abuse: diagnosis according to ICD codes (F11-16; F18-19; O35.5; P04.4; T40; T43.6; Z50.3; Z71.5) Injecting or daily use of narcotics Context: 5x increase in (recorded) opioid DRD

12 Problem opioid user studies (capturerecapture), Finland, Context: 50% increase in opioid DRD

13 Example: trend estimates for IDU prevalence in Estonia: Context: 50% increase in opioid DRD , falling back to 2010 baseline in 2013

14 Alternative approach: Cohort studies: 23 drug user mortality studies identified for the 7 countries 16 excluded - did not report a DRD rate Additional 3 excluded lack of case definition comprising active drug use during observation 4 remaining studies, 2 countries, based on 2 cohorts Scotland (opiate users, observation ): DRD rate during & post-treatment 4.4 (95% CI: ) per 1,000 PY (Merrall et al., 2012) Norway (opiate users, observation ): DRD rate during-treatment 4 (95% C.I. 0-8), post-treatment 21 (17 25), circa 6.7 (derived) per 1,000 PY combined (Clausen et al., 2008): note wide C.I.

15 Drivers of risk: Demographic Behavioural Contextual/environmental setting Set of (non-exhaustive) hypotheses about potential drivers - focus on drivers where (trend) data may be available We are looking at a moving target Upward and downward drivers will co-occur (and may operate simultaneously with changing prevalence) Likely complex set of interactions between some drivers, e.g. age & Hep C No simple answers

16 Demographic risk: trend in mean age at DRD (all DRD): Sweden Norway Ireland Finland Denmark Estonia Scotland (median) 5 0

17 Demographic risk: mean age of heroin users presenting for treatment (TDI) N.B. Just 20 heroin users presenting for treatment in Finland, Sweden Ireland Denmark Finland Data from: Gregorio Barrio, María J Bravo, Luis de la Fuente, José Pulido and Ernesta Sánchez, 2011

18 Variation in behavioural risk? Do the at risk populations vary wrt injecting, type of opioid use, poly drug use, etc? Injecting: there is substantial variation in rate of injecting. Type of opioid: there is variation; fentanyl in Estonia likely to put users at higher risk; buprenorphine (with alcohol & benzos) in Finland; Scotland (& Ireland) unusual re dominance of heroin. Poly use: toxicology suggests that poly use is common.

19 Proportion of (opioid) clients entering treatment by type of opioid used, Buprenorphine (+alcohol/prescription drugs) Fentanyl Heroin Methadone Other opioids Scotland: heroin dominates 0.00 Finland Estonia Denmark Sweden Ireland

20 Behavioural risk: proportion of opioid users presenting for treatment by type of opioid used: Sweden Sweden Heroin Sweden Methadone Sweden Other opioids TDI113

21 Behavioural risk: proportion of opioid users presenting for treatment by type of opioid used: Denmark Denmark Heroin Denmark Methadone Denmark Other opioids TDI113

22 Behavioural risk: percentage injecting among opiate users presenting for treatment Sweden Norway Ireland Finland Denmark Estonia

23 Contextual risk: prevalence of BBV among PWID ( ):

24 Contextual risk: number of persons in OST during 2007 (latest available): Rough prevalence (gu)estimates: Scotland: 60-63k (good POU proxy) Ireland: k (POU) Denmark: 20-24k (HRDU minus cannabis) Norway: k (POU derived) Sweden: 29.5k (PDU) Finland: 13-15k (POU) Estonia: 4-10k (IDU incl stimulants) Scotland Ireland Denmark Norway Sweden Finland Estonia

25 Contextual risk: proportional change in number receiving OST, anchored on Sweden Norway Ireland Finland Denmark Scotland Estonia (Scotland: - Health and social responses > Opioid substitution treatment > Clients > All clients ( )

26 Contextual risk: availability of Harm Reduction Interventions:

27 Summary: DRD increase between /9 (some countries) Prevalence of opioid use a key driver of differences between countries Evidence of increasing demographic risk (age) Evidence of behavioural differences (opioid type & injecting) Some evidence of variation in Tx coverage Some evidence of variation in BBV prevalence Insufficient evidence of substantial variation in harm reduction

28 Recommendations: Enhance specificity re opioid DRD reporting Avoid DRD population rates as an indicator of risk Improve coverage of POU estimates Investigate scope for (comparable) cohort studies, to enable risk comparison