부산대학교 도서관

Background: The objective of this study is, to describe patterns (or "classes") used latent class analysis (LCA) to examine patterns of drug use profiles among people who inject drugs (PWIDs) then to examine factors associated with each identified classes membranes in PWIDs in Iran. Methods: In a cross-sectional survey, using snow-ball sampling, we recruited 500 PWIDs in Tehran. Clustering of the behaviors was investigated using exploratory LCA. After identification of the latent classes and optimal number of latent classes, we used multi-nominal regression to identify factors associated with class membership. Results: The mean, standard deviation (SD) and median durations of injection drug use were 6.0 ± 4.6 and 3.2 (IQR, 3.6–11.1) years the adjusted odds ratios (AOR) represented that the class 1 members had higher odds of being homeless (AOR = 2.6, 95% CI: 1.4–5.7); in past 12 months. Unemployed status predicted membership in the class. Odds of class 1 membership was higher in PWIDs who were Unemployed (AOR = 1.9, 95% CI: 0.52–2.63), and reported HIV-positive (AOR = 3.2, 95% CI: 1.74–6.52) Conclusion: PWIDs who reported being the human immunodeficiency viruses (HIV) positive were significantly more likely to belong to class 1, being primarily methamphetamine users who initiated injection at age 22 years or younger.