重庆市万州区已婚女性HPV感染危险因素和预测模型构建及验证

Risk factors for HPV infection in married women of Wanzhou District,Chongqing,establishment of prediction model and its validation

目的探讨重庆市万州区已婚女性人乳头瘤病毒(HPV)感染危险因素及预测模型构建。方法选取2018年2月-2020年3月重庆市万州区行宫颈癌筛查的3226例已婚女性进行基线调查,采用单因素和多因素Logistic回归分析HPV感染的危险因素,构建预测模型并验证。结果共3226例已婚女性进行了HPV检测,年龄22~45岁,平均(34.65±4.23)岁;HPV的总感染率为21.02%(678/3226),其中单一感染占比为67.85%(460/678),多重感染占比为32.15%(218/678);HPV单一感染中,HPV16型最为常见占比为23.70%。多因素Logistic分析结果,年龄≤30岁、配偶吸烟史、初次性交年龄≤23岁、每月性生活次数>5次、妊娠次数>2次、未使用输卵管结扎避孕、有流产史、有宫颈阴道炎症史为HPV感染的独立危险因素(P<0.05)。根据多因素Logistic回归分析结果,得到已婚女性HPV感染的预测模型,利用受试者工作特征(ROC)曲线对模型进行内部验证,结果显示,已婚女性HPV感染的风险预测模型的曲线下面积(AUC)为0.950,95%CI为0.941~0.959,表明该预测模型具有较好的效能。结论多种因素与已婚女性HPV感染相关,临床可基于高危因素构建预测模型,制定针对性预防措施。

OBJECTIVE To explore the risk factors for human papilloma virus(HPV)infection in married women of Wanzhou District,Chongqing and build up the prediction model.METHODS A total of 3226 married women who received cervical cancer screening in Wanzhou District,Chongqing from Feb 2018 to Mar 2020 were enrolled in the study,and the baseline survey was conducted.Univariate analysis and multivariate logistic regression analysis were performed for risk factors for the HPV infection,the prediction model was built up and validated.RESULTS The HPV test was carried out for the 3226 married women who were aged between 22 and 45 years old,with the average age(34.65±4.23)years old.The total incidence rate of HPV infection was 21.02%(678/3226);the married women with single infection accounted for 67.85%(460/678),and the married women with multiple infections accounted for 32.15%(218/678).Among the married women with single HPV infection,23.70%had HPV16 infection.The result of multivariate logistic analysis showed that the no more than 30 years of age,smoking history of spouse,age of initial sexual intercourse no more than 23 years old,more than 5 times of sexual intercourse per month,more than 2 times of pregnancy,contraception without use of tubal ligation,abortion history and history of cervical vaginitis were the independent risk factors for the HPV infection(P<0.05).The prediction model for female HPV infection was established based on the result of multivariate logistic regression analysis,the internal validation was carried out for the model by using receiver operating characteristic(ROC)curves,and the result showed that the areas under curves(AUC)of the risk prediction model was 0.950 in prediction of HPV infection in the married women,with 95%CI 0.941-0.959,indicating that the prediction model had favorable efficiency.CONCLUSION The HPV infection in the married women is associated with multiple factors.The hospital may build up the prediction model based on the high-risk factors and put forward targeted prevention measures.