Carbamazepine cutaneous adverse reactions and HLA gene variation in the Chinese population: a systematic review and meta-analysis

This study was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement. The PRISMA checklist is described in Supplementary Table 1. The literature search strategy is shown in Supplementary Material II.The quality range of all included studies, evaluated by Newcastle–Ottawa scoring, is six to nine (Supplementary Table 2). Egger’s test shows that there is no publication bias (Egger’s test, t = 3.32, p = 0.004, Supplementary Figure 4). The publication bias verification is performed using the clipping method, and the result of clipping method indicates that there were no indications of publication with the Duval’s trim and fill method (no new studies added); the results are robust (Supplementary Figure 4). Sensitivity analysis was conducted on 16 meta-results of HLA-B*15:02 and CBZ-SJS/TEN in Figure 3. The combined effect of 16 meta-results did not change significantly after being excluded from the study one by one (Supplementary Figure 5). The distribution of the funnel diagram was relatively symmetrical (Supplementary Figure 6)

本研究遵循系统评价与荟萃分析优先报告项目（PRISMA）声明进行。PRISMA清单详述于补充表1。文献检索策略展示于补充材料II。所有纳入研究的质量范围，经纽卡斯尔-渥太华评分评估，介于六至九（补充表2）。Egger检验表明不存在发表偏倚（Egger检验，t = 3.32，p = 0.004，补充图4）。发表偏倚的验证通过裁剪法进行，裁剪法的结果显示，未出现Duval裁剪和填充法中的发表迹象（未添加新研究）；结果稳健（补充图4）。对HLA-B*15:02和CBZ-SJS/TEN的16项荟萃结果进行了敏感性分析，如图3所示。在逐一排除研究后，16项荟萃结果的合并效应未发生显著变化（补充图5）。漏斗图的分布相对对称（补充图6）