Language bias between publications identified in Chinese and English bibliographic databases
Table 5 summarizes the differences in pooled HSV-2 seroprevalence between publications identified in Chinese and English bibliographic databases. More publications were collected in Chinese bibliographic databases (371 identified in CNKI and Wanfang vs . 58 identified in PubMed and Embase). HSV-2 seroprevalence in Chinese publications ranged from 0.0% to 85.3%, with a median of 11.3%, and in English publications ranged from 0.0% to 88.0%, with a median of 15.6%. Pooled HSV-2 seroprevalence was higher among studies identified in English than in Chinese (20.5% vs . 13.6%).
For intermediate-risk populations and drug users, we observed a higher synthesized seroprevalence in publications published in Chinese than in English bibliographic databases (22.1% vs . 9.6%, 63.1%vs . 35.5%, respectively). The results of the meta-regression analysis also supported this observation (RR=0.84, 95% CI: 0.76-0.93, RR=1.11, 95% CI: 1.04-1.20, respectively). Within FSWs and MSM/MSWs, HSV-2 seroprevalence was significantly higher in publications published in English compared to those published in Chinese (RR=1.21, 95% CI: 1.10-1.33, RR=1.11, 95% CI: 1.04-1.20, respectively). However, within general populations, STI clinic attendees and symptomatic populations, and HIV-positive individuals and individuals in HIV-discordant couples, there was no significant difference in HSV-2 seroprevalence between publications identified from Chinese and English bibliographic databases.