Comparison of condom use at last sex and HIV prevalence by receipt of any interventions in the past 12 months among MSM surveyed in 16 Chinese cities from 2006 to 2009.

*Adjusted for age, marriage, education, ethnicity, residence, occupation, recruitment venue, age at first sex, and 6-month drug use.**Coverage rate presents the percentage of participants in each study year who stated they received any of six interventions from the China GF-5 Project in the past 12 months, e.g., 59% said they received interventions in 2006 while 41% did not.***There is strong evidence that the association between receipt of interventions and condom use at last sex is different over time (interaction, p = 0.002). There is strong evidence that the association between receipt of interventions and HIV prevalence is different over time (interaction, p#Intervention score ranged from 0–6, and the larger score presents receipt of more interventions in the past 12 months.