Temporal trend in the survival of injecting and non-injecting heroin-dependent patients admitted to treatment in Barcelona, Spain: a 30-year cohort study.

Survival of Heroin-Use-Disorder (HUD) has improved in past decades due to the generalization of opiate substitution treatment, risk reduction interventions and the effectiveness of antivirals for HIV and/or HCV-infection among other life-threatening illnesses. We aimed to analyze the evolving characteristics and survival of Injecting-Drug-Use-(IDU) and non-IDU HUD patients admitted to treatment for 30 years in 3-hospital-based, addiction units in Barcelona, Spain.

Cohort study of patients admitted between 1989-and-2018. Drug use characteristics and blood samples were obtained at admission. Three periods of 10-years-each were defined and vital status was ascertained from charts and the mortality register until June-30, 2019. The primary outcome was the elapsed time from admission to either death or end-of-follow-up. Using the period-analysis approach, individuals contributed as many records to the analyses as periods in which they were observed at risk. Survival time was treated as late-entry if admission occurred in a previous period and was right-censored if individuals were alive at-the-end of period. Kaplan-Meier methods and log-rank tests were used.

3.100 HUD-patients (80% men) were eligible. Age at admission was 28-years [IQR: 24-33], 87% were IDUs and 31% had co-occurring alcohol-misuse. Prevalence of HIV and HCV-infection was 45% and 72%, respectively. 75% of patients were admitted between 1989-1998, 21% in 1999-2008 and 4% in 2009-2018. There was a significant reduction in prevalence of IDU, HIV and HCV-infection by-period-of admission. Total follow-up was 52.114 person-years and 1.254 (24%) patients died. Overall mortality decreased from 4.3 x 100 p-y in-first-period to 0.7 x 100 p-y in the last (RR 6.1). Survival-of-IDUs admitted during the two-first periods was worse than the non-IDUs (P=0.003 and P=0.029, respectively). However, there were no differences in survival among IDU and non-IDU patients admitted in recent period (P=0.937, log rank test).

Major changes in heroin administration, blood-borne infections and survival are observed in our setting.