Life chaos, social support, and subsequent emergency department use among people who inject drugs in Québec, Canada


Author: Stine Bordier Høj, Nanor Minoyan, Sarah Larney, Julie Bruneau

Theme: Epidemiology & Public Health Research Year: 2023

Background:
Chaotic life circumstances and inadequate social support may impede navigation of health services and foster reliance on permeable settings such as emergency departments (ED). We examined i) life chaos (LC) and ii) social support (SS) as potential predictors of subsequent ED use among people who inject drugs (PWID) in Québec, Canada. 

Methods:
Data were from the Virtual Cascade of Care Cohort (VCCC). Eligible participants (aged ≥18, lifetime injection drug use, past six-month illicit drug/heavy alcohol use) were recruited via community-based harm reduction organisations in three cities (04.2018–01.2019). Interviewer-administered questionnaires captured psychometrically validated measures of LC and SS alongside sociodemographic, social stability, and substance use indicators. Data on ED use in the six months following questionnaire completion, including major diagnostic codes, were obtained via administrative data linkage. 

Multivariable Poisson regression models estimated associations between i) LC and ii) SS (both median-dichotomized) and ED visit count, adjusting for age, gender, and factors univariably associated with the outcome (p<0.10). We additionally compared proportions of participants with ED visits for different diagnostic categories using chi-square analyses. LC (C ronbach alpha=0.64) and SS (Cronbach alpha=0.85) were moderately correlated and analysed separately. 

Results:
183/196 (93%) interviewed participants were successfully linked (74% men, median age 44 (IQR:34–53), 56% past-month injection). High LC was not associated with ED visit count, but did predict higher prevalence of visitation for mental health problems (21% vs. 8%, p=0.01). High SS predicted fewer ED visits overall (aIRR=0.44, 95%CI: 0.34-0.57) and lower visitation for mental health (7% cf. 20%, p=0.02) and musculoskeletal problems (0% vs. 9%, p<0.01).

Conclusion:
LC and SS were differentially associated with ED visit count, but both predicted visitation for mental health problems. Findings suggest interventions to enhance community support and address social challenges may reduce ED reliance, and highlight a need for accessible primary mental health care.

Download abstract Download poster