Over 1200 drugs-related deaths and 190,000 opiate-user-years of follow-up: relative risks by sex and age-group

Heroin users/injectors' risk of drugs-related death by sex and current age is weakly estimated both in individual cohorts of under 1000 clients, 5000 person-years or 50 drugs-related deaths and when using cross-sectional data. A workshop in Cambridge analysed six cohorts who were recruited according to a common European Monitoring Centre for Drugs and Drug Addiction (EMCDDA) protocol from drug treatment agencies in Barcelona, Denmark, Dublin, Lisbon, Rome and Vienna in the 1990s; and, as external reference, opiate-user arrestees in France and hepatitis C diagnosed ever-injectors in Scotland in 1993-2001, both followed by database linkage to December 2001. EMCDDA cohorts recorded approximately equal numbers of drugs-related deaths (864) and deaths from other non-HIV causes (865) during 106,152 person-years of follow-up. External cohorts contributed 376 drugs-related deaths (Scotland 195, France 181) and 418 deaths from non-HIV causes (Scotland 221, France 197) during 86,417 person-years of follow-up (Scotland 22,670, France 63,747). EMCDDA cohorts reported 707 drugs-related deaths in 81,367 man-years {8.7 per 1000 person-years, 95% CI: 8.1 to 9.4} but only 157 in 24,785 person-years for females {6.3 per 1000 person-years, 95% CI: 5.4 to 7.4}. Except in external cohorts, relative risks by current age-group were not particularly strong, and more modest in Poisson regression than in cross-sectional analyses: relative risk was 1.2 (95% CI: 1.0-1.4) for 35-44 year olds compared to 15-24 year 3 olds, but 1.4 for males (95%CI: 1.2-1.6), and dramatically lower at 0.44 after the first year of follow-up (95% CI: 0.37-0.52).