基于列线图建立女性不孕预测模型

Establishment of female infertility prediction model based on nomogram

目的探讨女性不孕症的危险因素,构建不孕女性个体化风险预测模型。方法选取2018年5月至2021年10月重庆医科大学附属妇女儿童医院及陆军军医大学西南医院收治的自愿参与本研究的辅助生殖助孕人群为观察组,选取同期在重庆医科大学附属妇女儿童医院产科建档的已自然妊娠<12周的孕妇为对照组,最终纳入观察组4276例,对照组3240例。采用问卷填写方式收集数据,采用Logistic回归分析数据并建立列线图预测模型,通过Bootstrap法对模型内部进行验证,采用一致性指数(C-index)及校准曲线评价列线图模型性能。结果年龄、孕前体质量指数、初潮年龄、毒物接触史、烟草摄入、节食、代餐饮食、毒品摄入、近2年痛经、初次性生活年龄、文化程度、家庭人均月收入、近2年月经量均是女性不孕症的影响因素(P<0.05)。建立的预测不孕风险列线图模型具有较好的预测效能[C-index为0.812(95%CI:0.802~0.822),校正后C-index为0.810(95%CI:0.800~0.819)]。结论列线图预测女性不孕症效果良好,提供了预测女性不孕风险可视化工具,具有一定的临床价值。

Objective To explore the risk factors of female infertility,and to build an individual risk prediction model for infertile women based on independent risk factors.Methods From May 2018 to October 2021,the assisted reproductive and assisted pregnancy population admitted to Chongqing Health Center For Women and Children and the Southwest Hospital of Army Medical University,who voluntarily participated in this study,were recruited as the observation group,and the population with natural pregnancy<12 weeks,who were registered in the Department of Obstetrics of Women and Children′s Hospital of Chongqing Medical University during the same period,was selected as the control group.According to the inclusion and exclusion criteria,4276 cases in the observation group and 3240 cases in the control group were finally included.The data was collected by questionnaire survey,the prediction model based on nomogram was established by Logistic regression analysis,the model was internally verified by bootstrap method,and the performance of nomograph model was evaluated by the consistency index(C-index)and calibration curve.Results Female age,progestational body mass index,age of menarche,history of toxic exposure,tobacco intake,diet,substitute diet,drug intake,painful menstruation in recent two years,age of first sexual life,educational level,monthly household income per capita,menstrual blood volume in recent two years were associated with female infertility.The nomogram model had good prediction efficiency[C-index:0.812(95%CI:0.802-0.822),corrected C-index was 0.810(95%CI:0.800-0.819)].Conclusion The nomogram has a good effect in predicting female infertility,and provides a visual measurement tool for individualized risk prediction of female infertility,which has certain application value.