Theme: Epidemiology & Public Health Research Year: 2022
Heroin and other opioids account for more than half of deaths due to drug use, and are involved in a
majority of fatal overdoses in Europe. Opioid agonist treatment (OAT) can be highly effective in
reducing mortality of persons with opioid use disorders (PWOD).
Our goal was to calculate all-cause mortality rates (ACMR), and provide a small number of mutually
exclusive, but exhaustive categories of common causes of death (CCD) in such a way that it helps to
describe time trends. From clinical perspective, the knowledge of shifts between clusters of specific
causes of death during and out of OAT may supply valuable information on areas where to improve
medical treatments and preventive measures.
In this retrospective treatment cohort study with individual level linkage of different data sources,
we investigated PWOD who entered OAT with methadone, buprenorphine, or diacetylmorphine in
Zurich (Switzerland) between 1992 and 2013, and observed death rates and causes of death during
and out of treatment over this 22 year-period.
Using a parametric approach (extended exponential survival analysis), we developed a statistical
model which allowed simultaneously estimating ACMR and the fractions of CCD categories together
Overall ACMR was 15.0 per 1’000 person-years. We found ACMR increasing with age (rate ratio per
age decade, 95%-CI = 1.81, 1.64 to 1.97), but no effect of sex, treatment-status, -duration, and
calendar year. Age controlled ACMR was stable during OAT, sharply increased in year of
discontinuation, non-linearly diminishing over following years. Proportions of CCD-categories were
only related to calendar year. Drug-induced and HIV related causes declined, while liver-related and
other internal causes increased during the observed time.
OAT seeking PWOD’s common causes of death can shift over time independently of individuals’
aging and treatment status. Treatment and prevention measures should be adjusted to such trends.