Lost to follow-up and its influencing factors among men who have sex with men in a cohort in Xinjiang Uygur Autonomous Region

  Objective  To analyze the rate and associates of lost to follow-up in a cohort of men who have sex with men (MSM) initially negative for human immunodeficiency virus (HIV) in Xinjiang Uygur Autonomous Region (Xinjiang).

  Methods  We sequentially recruited 1 041 MSM tested negative for HIV at the time of inclusion in 3 municipality regions of Xinjiang from June 2010 to March 2015 for a prospective cohort study. The participants of the study were followed-up every 3-month and were surveyed with a questionnaire interview on sex behavior and HIV antibody test every 6-month.

  Results  For all the participants, the number of being successfully interviewed was 900 and 884 at the time of 6th month and 12th month follow-up, with the ratios of 13.54% and 15.08% for being lost to follow-up, respectively. In univariate analysis, there were significant differences between the participants being followed-up and lost to followed-up in age, education, registered residence, monthly income, knowledge about HIV infection, current residential region, motivation for participating in the survey, setting of seeking sexual partners, frequency of condom use, number of sexual partners during previous 6 months, and hepatitis C virus infection (P < 0.05 for all). Multivariate logistic regression analysis revealed following significant impact factors for lost to followed-up: education (high vs. low: odds ratio OR = 0.089, 95% confidence interval 95% CI: 0.060 – 0.133; P < 0.001), current residential region (metropolis vs. small and medium cities: OR = 1.800, 95% CI: 1.231 – 2.632; P = 0.002), motivation for participating in the survey (to have HIV detection vs. others: OR = 0.663, 95% CI: 0.465 – 0.946; P = 0.023), and the number of sexual partners during previous 6 months (high vs. low: OR = 4.070, 95% CI: 2.002 – 8.278; P < 0.001).

  Conclusion  The rate of lost to followed-up is high among MSM involved in a prospective cohort in Xinjiang and the follow up among the MSM is mainly associated with education, current residential region, motivation for the participation and the number of sexual partners.