Social network predictors of recent and sustained injection drug use cessation: findings from a longitudinal cohort study

Abstract

Background and aims: The US opioid crisis has led to increases in overdose fatalities and the incidence of HIV, hepatitis C and other infections. This analysis examines social network predictors of recent (self-report injection followed by non-injection) and sustained (self-report non-injection at two consecutive visits among those who previously injected) injection cessation in Appalachian Kentucky.

Design: Data were collected through bi-annual longitudinal assessments for Social Networks among Appalachian People (SNAP; 2008-17). Using logistic regression with generalized estimating equations that clustered on individuals, we regressed non-injection status on the number of social network members who (a) did not inject and (b) recently stopped injecting and tested for interactions between each social network exposure and prior non-injection status. Social network exposures were self-reported.

Setting: Rural eastern Kentucky, USA.

Participants: Participants entered the analysis only after reporting recent injection and had to have had at least two consecutive study visits (n = 326).

Measurements: Interviewer-administered surveys collected individual-level socio-demographics, recent (past 6 months) drug use behaviors and the names of recent social support, sex and drug-use partners.

Findings: After adjusting for confounders, the number of non-injecting social network members was positively associated with recent/sustained injection cessation (adjusted odds ratio = 1.27; 95% confidence interval = 1.13-1.42) and having more social network members was associated with reduced odds of recent/sustained injection cessation. The number of previously injecting social network members who had recently stopped injecting was not statistically significantly associated with injection cessation. Neither of the interactions we tested for was statistically significant, suggesting that the relationships may be similar for those who recently stopped injecting versus had not injected for at least 1 year.

Conclusions: For each additional network member who did not inject drugs, there was an increased odds of recent and sustained injection cessation among people with a history of injection drug use in Appalachian Kentucky.

Keywords: Injection drug use cessation; longitudinal cohort study; opioids; rural; social influence; social norms.