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

ABSTRACT

BACKGROUND: In Europe, North America, and Australia, mortality due to drug-related (DR) causes amongst people who inject drugs (PWID) is a major issue. Our objective was to characterise temporal trends in DR mortality rates in a large cohort of PWID in Scotland over the past decade, all of whom had been diagnosed with hepatitis C virus (HCV) infection, and to investigate factors associated with DR mortality. METHODS: Retrospective longitudinal cohort study linking Scotland's national HCV Diagnosis Database and deaths registry. The study cohort consisted of all individuals with likely injection drug use-related route of HCV acquisition, who had been diagnosed with HCV between 1991 and 2018, and were alive and aged under 65 years on 1 January 2009. We used Lexis expansion to adjust for ageing cohort effects and calculated the mortality rate from an underlying/contributing DR cause over the period 2009-2018. We fitted Poisson regression models to estimate the temporal trend adjusting for attained age, sex, referral setting, region, and viraemic status at baseline. RESULTS: Amongst the study population (n = 35,065; 236,914 person-years), a total of 1900 DR deaths occurred; the DR mortality rate increased from 5.6/1000 [101 deaths] in 2009 to 12.4/1000 [342] person-years in 2018. Increasing trends were observed for all age-groups except 55-64 years. The overall DR mortality rate was highest for referrals for HCV testing from prison (11.0/1000) and hospital settings (10.0/1000). Mortality increased with calendar time period, with significantly raised adjusted rate ratios (RRs) from 2015 (RR=1.40, 95% CI1.16-1.69) to 2018 (RR=2.23, 95% CI1.88-2.64), compared with 2011-2012, for older age (35-44 RR=1.37, 95% CI1.20-1.56; 45-54 RR=1.32, CI1.14-1.53) compared with <35 years, for persons diagnosed with HCV since 2009 (RR=1.34, 95% CI1.21-1.49), and for prison and hospital referrals (RRs of 1.30, 1.37) compared with GP referrals. CONCLUSION: Increasing DR mortality rates in Scotland over the past decade are not just due to an ageing cohort. Harm reduction services will likely need to expand and adapt to reverse the recent upward trends in DR mortality in PWID.