A meta-analysis of the test results of five indicators for hepatitis B virus of college freshmen in current mainland China

Abstract

Objective: Using Meta analysis to evaluate the test results of five indicators and main infection patterns for hepatitis B virus (HBV) of college freshmen in mainland China. Methods: In this paper, "college students", "college freshmen", "hepatitis b virus" and "HBV" were used to search systematically for cross-sectional data of the results of five indicators for HBV of entrants from 2006-2015 through CNKI, Wanfang Data, VIP as well as Pubmed and Web of Science. Then Combie's standard for cross-sectional study was used to assess the qualities of the articles included, and Stata 14.0 software and Meta analysis method (including performing heterogeneity test, calculating combined detection rates and carrying out sensitivity analysis and publication bias test) were used to analyze and evaluate the epidemiological significance of the results above. Results: Sixteen studies containing 82 144 cases were included. The result of heterogeneity test showed that there was significant heterogeneity (P<0.001), hence random effects model was adopted. The Meta analysis resulted in total positive rate for surface antigen (HBsAg) of 7.0% (95%CI: 6.0%-8.0%), surface antibody positive rate (HBsAb) of 46.0% (95%CI: 42.0%-49.0%), and positive rate for HBsAg, core antibody (HBcAb) combined with E antigen (HBeAg) or E antibody (HBeAb) of 3.0% (95%CI: 2.0%-3.0%) and 3.0% (95%CI: 2.0%-3.0%), respectively; whereas the negative rate of all indicators was 41.0% (95%CI: 35.0%-46.0%). Sensitivity analysis showed that the results were highly heterogeneous (P<0.001), while the result of Egger's regression indicated that there were publication bias of all indicators (P<0.05), except for HBsAb (P=0.436). Conclusion: The positive rates of HBsAg and HBsAb of college students are lower than those of general population, while the positive rates of HBsAg, HBcAb combined with HBeAg or HBeAb and the all-negative rate of HBV count for a considerable proportion among entrants.