3.3 Risk of bias assessment
As shown in Figure 2e , 81.5% of the publications reported at least two diagnostic tools to detect NTS-positive samples and thus were judges as not having risk of bias. Among the 94 publications that serotyped the NTS isolates, 87.2% were judges as not having risk of bias because the studies serotyped all the isolates. For these same publications, 74.5% reported the complete diversity of serovars found and thus were free of risk of bias. A total of 84.8% of the 46 publications that assessed the antimicrobial resistance profile of the isolates were free of risk of bias because they included all the NTS-positive samples. However, 54.3% of these 46 publications had risk of bias because they did not report the complete antimicrobial resistance profiles for all the assessed isolates.