We collected information on HIV testing rates among MSM from 2001 to 2011. Linear regression Epacadostat chemical structure was performed to estimate the change in HIV testing rates over time, with 95% confidence intervals (CIs), using information obtained from the available studies. Spearman’s rank correlation was performed to investigate the relationship between testing rates and the average age of surveyed MSM (P-value < 0.05 represents statistical significance). All analyses were performed in stata 10 (version 10.0, College Station, Texas, USA). We identified 1878 articles using the initial keywords (1872 articles were obtained from eight electronic databases and six relevant articles were

identified from the reference lists of these articles). After screening the titles of the 1878 articles, 1574 articles were excluded because of duplication or because they were unrelated to the topic. The abstracts of the remaining 304 articles were screened, and 97 articles were further excluded because they were not related to the topic. There were 207 articles eligible for full-text screening, of which 152 articles were subsequently excluded (143 articles did not report the level of HIV testing; five were duplicated in the databases;

two reported HIV testing rates among male sex workers, and two did not report the study period). Finally, we identified 55 eligible articles (44 in Chinese and 11 in English) that reported the HIV testing rate among Chinese MSM, Hydroxychloroquine which in total provided 37 testing rate estimates for individuals who had ever been tested for HIV during 2002–2009 and 29 testing rate estimates for individuals who had been tested in the past 12 months during 2003–2009 (Table 1). The selection process is illustrated in Figure S1. Among the 55 studies, eight studies reported the HIV testing rate in multiple years [25-32]. Eight of the 55 studies

did not report the recruitment method, while 24 studies recruited MSM participants from MSM venues, six recruited from Internet sites, five recruited from VCT clinics and one recruited Demeclocycline from MSM community settings; 12 studies used multiple recruitment methods (Table 1). The sample sizes of the studies ranged from 20 to 5454 [median 402; interquartile range (IQR) 202–558]. Our trend analysis across all available studies suggested that the percentage of MSM who had ever been tested for HIV increased from ∼10.8% (95% CI −2.8–24.4%) in 2002 to ∼51.2% (95% CI 39.0–63.4%) in 2009, with an average annual growth rate of ∼5.8% per year (95% CI 2.4–9.1%) (P = 0.0013) (Fig. 1a). The percentage of Chinese MSM who reported testing for HIV in the past 12 months also increased significantly, from ∼11.0% (95% CI −4.2–26.2%) in 2003 to ∼43.7% (95% CI 37.1–50.2%) in 2009, with an average increase of approximately 4.9% per year (95% CI 1.8–8.1%) (P = 0.0034) (Fig. 1b). Four of the 55 reported that approximately 82–97% of tested MSM were also notified about their HIV status after confirmation tests [25, 33-35].