带状疱疹病人并发带状疱疹后神经痛的风险预测模型构建

Construction of risk prediction model for postherpetic neuralgia of patients with herpes zoster

目的:分析带状疱疹病人并发带状疱疹后神经痛(PHN)的危险因素,并构建风险预测模型。方法:选取2019年9月—2021年8月我院皮肤科收治的425例带状疱疹病人作为研究对象,根据是否并发PHN将其分为PHN组(128例)和无PHN组(297例),分析带状疱疹病人临床资料,采用单因素和Logistic回归分析筛选带状疱疹病人并发PHN的危险因素,根据独立危险因素构建带状疱疹病人并发PHN风险预测Nomgram模型并进行拟合优度检验。结果:带状疱疹病人PHN发生率为30.12%,Logistic回归分析结果表明,年龄>60岁、患有糖尿病、吸烟、皮疹面积>5%、急性期重度疼痛、CD_(4)^(+)/CD_(8)^(+)比值异常是带状疱疹病人并发PHN的独立危险因素(P<0.05)。基于6项独立危险因素构建的带状疱疹病人并发PHN风险预测Nomgram模型一致性指数为0.824[95%CI(0.796,0.852)],校正曲线趋近于理想曲线,受试者工作特征曲线下面积为0.812[95%CI(0.787,0.837)]。结论:带状疱疹病人并发PHN风险预测Nomgram模型具有较好的预测效能,有利于临床及早筛查高风险病人和进一步改进治疗计划。

Objective:To analyze the risk factors of postherpetic neuralgia(PHN)of patients with herpes zoster and to construct risk prediction model.Methods:425 patients with herpes zoster admitted to department of dermatology of our hospital were selected as study subjects from September 2019 to August 2021.According to whether they were complicated with PHN or not,they were divided into PHN group(128 cases)and without PHN group(297 cases).The clinical data of patients with herpes zoster were analyzed,and risk factors of PHN of patients with herpes zoster were screened by univariate and Logistic regression analysis.Nomgram model for predicting risk of PHN of patients with herpes zoster was constructed according to independent risk factors and goodness of fit test was performed.Results:Incidence of PHN of patients with herpes zoster was 30.12%.Logistic regression analysis showed that age>60 years old,with diabetes,smoking,rash area>5%,acute severe pain and abnormal CD_(4)^(+)/CD_(8)^(+) ratio were independent risk factors for PHN of patients with herpes zoster(P<0.05).The C-index of Nomgram model of risk prediction model for PHN of patients with herpes zoster based on the six independent risk factors was 0.824[95%CI(0.796,0.852)].Correction curve close to ideal curve.The area under the receiver operating characteristic curve was 0.812[95%CI(0.787,0.837)].Conclusions:Risk prediction Nomgram model for PHN of patients with herpes zoster has a good prediction efficiency,which is helpful for early screening of high-risk patients and further improving treatment plan.