Drug use behavioral trajectories and HCV incidence among people who are incarcerated

Author: Joanne Carson, Behzad Hajarizadeh, Marianne Martinello, Marianne Byrne, Lise Lafferty, Carla Treloar, Jason Grebely, Andrew Lloyd, Gregory Dore

Theme: Clinical Research Year: 2023

Background: There has been limited exploration of longitudinal trends in drug-use behaviours among incarcerated people. This analysis modelled behaviours following enrolment in the Surveillance and Treatment of Prisoners with hepatitis C (SToP-C) study.

Methods: Participants enrolled at four Australian prisons were followed at 3-6 monthly intervals (minimum three study visits) for up to 4-years to assess drug-use behaviours and HCV incidence. Population averaged behavioural changes were estimated using generalized estimating equations. As population averaged behaviours may mask distinct behavioural trajectories within a population, group-based trajectory modelling was used to identify groups if individuals with similar longitudinal risk behaviour profiles. HCV incidence rates were calculated using person-years (PY) of observation.

Results: Of 985 participants (median age 33 years; 84% male; 60% maximum security prisons 23% injecting drug use [IDU] in the past month; 16% current opioid agonist treatment [OAT]), 18% had previous HCV exposure and 26% had current HCV infection at enrolment. Among those with recent IDU, 83% were injecting opioids and 87% reported receptive needle/syringe sharing. Population averaged increases in the likelihood of opioid IDU (AOR 1.09; 95%CI 1.04, 1.14) at each visit were observed. However, in trajectory modelling, four distinct trajectories of increasing (10%), decreasing (10%), stable high (17%) and low (63%) probabilities of opioid IDU were observed (Figure). Population averaged increases in OAT (AOR 1.06; 95%CI 1.02, 1.10) were also evident in trajectory modelling. Population averaged declines in needle/syringe sharing were observed (AOR 0.90; 95%CI 0.81, 0.99), although in trajectory modelling 70% of those with recent IDU retained a high probability of needle/syringe sharing. HCV incidence was greatest for those with high probability of opioid IDU (22.2/100 PY; 95%CI 16.3, 30.3) or daily IDU (20.1/100 PY; 95%CI 15.2, 26.7).

Conclusions: These findings identify longitudinal risk profiles for targeted harm reduction and regular HCV surveillance and (re)treatment within prisons.

Download abstract Watch video