Quality assessment of 22 included studies (rigor score1).

1One point score for meeting each of the following items (if data were not available in the articles for any item, 0.5 was recorded): (a) is a prospective cohort, (b) use a control or comparison group, (c) collect pre/post intervention data, (d) use random assignment of participants to study arms, (e) do random selection of subjects for assessment, (f) have comparison groups equivalent at baseline on outcome measures, sample size more than 100, (g) have a follow-up rate of 80% or more, (h) have comparison groups equivalent on socio-demographic measures, including age, education, race, employment, income, marital status and others [If more than half of variables were shown equivalent between groups, ‘1’ should be marked; otherwise “0”], and (i) have comparison groups equivalent at baseline on unprotected anal intercourse. If pre-post designs without comparison groups, “0” was recorded for item (h) and (i).