Objectives To assess whether pooled sample testing with nucleic acid amplification tests was a potential alternative to three single-site sample testing to screen for Chlamydia trachomatis and Neisseria gonorrhoeae infections in asymptomatic men who have sex with men.

Methods We prospectively compared pooled sample testing with single-site sample testing in asymptomatic MSM. Self-obtained paired rectal samples, one gargle sample and one first-void urine sample were collected from participants to generate two sets of samples: one for pooled sample testing and the other for single-site testing. We used modified pooled sampling, which is defined as the use of gargle samples, instead of swabs, for the pooled sample to test for pharyngeal infection.

Results This study included 513 MSM. The positive rates of C. trachomatis and N. gonorrhoeae were 20.3% and 11.7%, respectively, for single-site sample testing. Compared with the sensitivity of single-site testing as the gold standard, the sensitivities of pooled sample testing for C. trachomatis and N. gonorrhoeae were 94.2% (95% CI 88.0% to 97.3%) and 98.3% (95% CI 90.9% to 99.9%), respectively. The concordance rate and kappa coefficient were 98.3% (95% CI 96.7% to 99.2%) and 0.945 (95% CI 0.859 to 1.000), respectively, for C. trachomatis and 98.8% (95% CI 90.1% to 100%) and 0.943 (95% CI 0.857 to 1.000), respectively, for N. gonorrhoeae.

Conclusions The modified pooled sampling had a comparably high consistency with single-site sample testing. The results strongly suggest that the gargle sample is suitable as a part of pooled sample for STI screening of C. trachomatis and N. gonorrhoeae.