Objectives: To conduct a systematic review of published and unpublished data from research and public health information systems on the prevalence of male-to-male sex in the total male population; as well as among men who have sex with men (MSM), data on prevalence of heterosexual activity and heterosexual unions; prevalence of condom use with male and female partners; and prevalence of HIV infection and other sexually transmitted infections (STIs).
Methods: Key indicators were defined (a) among men in the general population: prevalence of sex with a man ever and last year; (b) among MSM: prevalence of heterosexual experiences ever and last year; proportion of male-female transgenders; proportion of sex workers; prevalence of HIV and other STIs, condom use in last sexual encounter; consistent condom use with men last year; never used a condom with a man. With help from key informants, study searches were conducted in Pubmed, LILLACS, institutional databases, conference records and other sources. Methodology and quality of information were assessed, and the best data available for 2003-7 were selected. Indicator estimates from each study were used to propose regional estimate ranges.
Results: A total of 83 new entries were entered into the database in addition to the previous 561, totalling 644. Of these, 107 showing 2003-7 data were selected. Many new studies came from sub-Saharan Africa, portraying hidden HIV epidemics among MSM. The most frequently reported estimate was HIV infection, with high estimate ranges in most of the regions, except for Middle East and North Africa and Eastern Europe. The next most frequently reported was lifetime frequency of heterosexual sex, showing that roughly 50% of MSM ever had sex with a woman. The small number of newer studies reporting prevalence of "sex with a man in last 12 months" between 2003 and 2007, did not warrant enough new evidence to revise our 2005 size estimates for MSM populations.
Conclusions: A considerable number of new studies with estimates of relevance to understanding sexual behaviour and HIV among MSM were identified, with an encouraging amount of new data coming from sub-Saharan Africa. However, limitations in the quality, utility and comparability of available information persist. At least three measures could be promoted for use in surveillance and academic studies: standardised indicators for MSM studies; standardised operational definitions of, and instructions to describe, variables; and standardised research designs and data gathering strategies. A prerequisite for this all is intense advocacy to ensure a social climate in which research into such matters is prioritised, resources are made available as needed and the human rights of MSM are respected.