牙周炎患者发生种植体周围炎风险预测模型的构建  

Establishment of risk prediction model for peri⁃implantitis after dental implants in patients with periodontitis

作　　者：唐金鑫 汤春波[2] 宋鑫[1] 芮娜[1] 薛昌敖[1] TANG Jinxin;TANG Chunbo;SONG Xin;RUI Na;XUE Chang′ao(Department of Stomatology,Nanjing First Hospital,Nanjing Medical University,Nanjing 210006,China)

机构地区：[1]南京医科大学附属南京医院(南京市第一医院)口腔科,江苏南京210006 [2]南京医科大学附属口腔医院种植科,江苏省口腔疾病研究重点实验室,江苏省口腔转化医学工程研究中心,江苏南京210029

出　　处：《口腔医学》2023年第8期706-710,共5页Stomatology

基　　金：国家自然科学基金(82170993)。

摘　　要：目的分析牙周炎患者牙种植术后发生种植体周围炎的危险因素,并建立列线图预测模型。方法共203例患者纳入研究,并采用随机抽样法将其分为建模组(n=142)与验证组(n=61)。基于牙周炎患者发生种植体周围炎的独立危险因素,建立Logistic风险预测模型。采用Bootstrap法对模型进行内部验证,外部验证通过验证组完成。利用Hosmer-Lemeshow检验(H-L检验)和受试者工作特征(receiver operating characteristic,ROC)曲线下面积(area under the curve,AUC)对预测模型进行评价。结果吸烟史、糖尿病史、未定期行牙周治疗、种植体周围黏膜厚度<2 mm、前牙区种植和粘接固位修复等都是牙周炎患者种植后发生种植体周围炎的独立危险因素(P<0.05)。根据以上危险因素,建立预测牙周炎患者发生种植体周围炎的列线图模型。H-L检验的显著性水平为0.536。校准曲线显示,建模组和验证组的预测值与实测值基本一致。建模组AUC为0.906,表明此模型具有较好的判别效果。结论以吸烟史、糖尿病史、未能定期行牙周治疗、种植体周围黏膜厚度<2 mm、前牙区种植和粘接固位修复为预测因子构建的列线图模型可有效预测牙周炎患者发生种植体周围炎的概率。Objective To analyze the risk factors of peri⁃implantitis after dental implants in patients with periodontitis,and construct a nomogram prediction model.Methods A total of 203 patients were included into the study,and divided into modeling group(n=142)and validation group(n=61)by random sampling.Based on the independent risk factors of peri⁃implantitis in patients with peri⁃odontitis,the Logistic risk prediction model was established.The model was internally verified by Bootstrap method and the external verification was completed by the verification group.Hosmer⁃Lemeshow test(H⁃L test)and the area under the receiver operating char⁃acteristic(ROC)curve was used to evaluate the prediction model.Resultsq100 Smoking history,diabetes history,irregular periodontal treatment,thickness of peri⁃implant mucosa<2 mm,anterior teeth implantation and adhesive restoration were the independent risk fac⁃tors for peri⁃implantitis after dental implantation in patients with periodontitis(P<0.05).According to the above risk factors,a nomo⁃gram model was constructed to predict the occurrence of peri⁃implantitis in patients with periodontitis.H⁃L test had a significant level of 0.536.The calibration curve showed that the predicted values of the modeling group and the verification group were basically consistent with the measured values.The area under the curve(AUC)of the model group was 0.906,indicating that the model had good prediction accuracy.Conclusion The risk prediction model,consisting of smoking history,diabetes history,irregular periodontal treatment,thickness of peri⁃implant mucosa<2 mm,anterior teeth implantation and adhesive restoration,may effectively predict the incidence of⁃peri⁃implantitis after dental implants in patients with periodontitis.

关 键 词：牙周炎 牙种植体 种植体周围炎 列线图预测模型 

分 类 号：R783.4[医药卫生—口腔医学]