| Sumario: | Background: Treatment interruptions (TIs) may lead to adverse health outcomes for people living with HIV. This study explores individual characteristics as well as proximal factors associated with TIs among people living with HIV in British Columbia (BC), Canada. Methods: Between January 2016 and September 2018, eligible people living with HIV aged ≥19 years across BC completed three surveys, 18 months apart, assessing demographics, health behaviours and healthcare engagement. TIs (≥60 days late for ART refills) were identified through the BC HIV Drug Treatment Program database from enrolment until December 2020. We used Cox proportional hazards models to analyse baseline factors associated with time‐to‐first TI and generalized linear mixed models to examine time‐updated factors related to TIs within 3 months of a study visit. Results: Among 639 participants receiving antiretroviral therapy (ART) at enrolment, 21.3% were women, 59% were men who have sex with men and 24.1% had ≥1 TI over the study period. Factors associated with time‐to‐first TI included history of incarceration (adjusted hazard ratio [AHR] = 1.97; p < 0.001), experiencing violence (AHR = 1.94; p = 0.019) and age ≥ 60 years (AHR = 0.39; p = 0.007); 50–59 years (AHR = 0.63; p = 0.038) and 40–49 years (AHR = 0.61; p = 0.036) versus <40 years. TIs within 3 months of a study visit were associated with being single (adjusted odds ratio [AOR] = 3.63; p = 0.005), recent homelessness (AOR = 2.29; p = 0.049) and recent injection drug use (AOR = 3.55; p <0.001) measured at the time of the visit. Conclusion: Analysing TIs using a time‐to‐event approach identified non‐modifiable risk markers (history of incarceration, experiencing violence and age), whereas our time‐updated analysis found modifiable factors (being single, recent homelessness and recent injection drug use). The latter should be examined further as potential targets for therapeutic interventions.
|
|---|