Abstract: Objective: To systematically evaluate the risk prediction models of postherpetic neuralgia. Methods: Therelevant literatures on postherpetic neuralgia risk prediction models published in PubMed, Cochrane Library, Embase, Web of Science, Sinomed, CNKI, VIP and Wanfang data knowledge service platform were searched from the inception of the database to April 1, 2024. PROBAST was used to evaluate the quality of the model, and RevMan 5.4 was used to conduct meta-analysis of the predictive factors in the model. Results: A total of 21 literatures were included, including 29 models, and the area under the curve (AUC) of the included models ranged from 0.710 to 0.990. Model calibration was performed on 16 models and validation was performed on 20 models. The results of PROBAST evaluation showed that 1 of the 21 included articles had a low risk of bias, and the other 20 articles had a high risk of bias. In terms of the applicability of the model, 2 articles are unclear, and the rest are highly applicable. The results of meta-analysis showed that age, gender, degree of pain in the acute phase, size of lesions, duration of initial treatment, presence of underlying disease, presence of prodromal pain, and presence of immunosuppressive status were factors influencing the risk of developing postherpetic neuralgia (P<0.05). Conclusion: The risk prediction models of postherpetic neuralgia still have shortcomings, and the overall research quality needs to be further strengthened.

Key words: Postherpetic neuralgia, Risk assessment, Forecast, Model, Systematic review; 

摘要： 目的：系统评价带状疱疹后遗神经痛风险预测模型，为临床遴选良好的风险评估工具及后续模型优化提供借鉴。方法：检索 PubMed、Cochrane Library、Embase、Web of Science、Sinomed、中国知网、维普及万方数据知识服务平台发表的关于带状疱疹后遗神经痛风险预测模型的相关文献，检索期限自建库至 2024 年 4 月 1 日。采用 PROBAST 对模型进行质量评价，采用 RevMan 5.4 对模型中预测因子进行 Meta 分析。结果：共纳入 21 篇文献，包含 29 个模型，纳入模型的曲线下面积（AUC）为 0.710~0.990。16 个模型进行了模型校准，20 个模型进行了验证。PROBAST 评估结果显示，纳入的 21 篇文献中有 1 篇为低偏倚风险，其余 20 篇均为高偏倚风险；模型适用性方面，2 篇为不清楚，其余为适用性高。Meta 分析结果显示，年龄、性别、急性期疼痛程度、皮损面积、初始治疗时间、是否有基础病、是否有前驱疼痛、是否存在免疫抑制状态是带状疱疹后遗神经痛发病风险的影响因素（P<0.05）。结论：带状 疱疹后遗神经痛风险预测模型仍存在不足，总体研究质量还需进一步加强。

关键词: 带状疱疹后遗神经痛, 危险性评估, 预测, 模型, 系统评价