腹膜透析患者首发腹膜炎的危险因素及风险评估模型构建

Risk factors and risk assessment model of first-episode peritonitis in peritoneal dialysis patients

目的探讨腹膜透析(PD)患者首发腹膜透析相关性腹膜炎(PDAP)的危险因素,构建风险评估模型,为临床早期预防和制定干预措施提供参考。方法选取2020年1月-2022年12月南京鼓楼医院肾内科收治的189例住院PD患者为研究对象,其中163例未发生PDAP患者为对照组,26例首发PDAP患者为病例组,收集两组患者一般情况、临床资料进行分析,探讨患者首发PDAP的危险因素,建立风险评估模型。结果189例患者中,26例患者首次发生PDAP。细菌培养共检出病原菌26株,其中,革兰阳性菌16株,占61.54%;革兰阴性菌8株,占30.77%;真菌2株,占7.69%。年龄、性别、居住地、白蛋白是患者首次发生PDAP的危险因素。构建风险评分模型:“女性或农村”赋值为2分,年龄值22~<43岁、43~<55岁、55~<66岁、66~90岁依次为0、1、2、3分,白蛋白值25.30~<33.56 g/L、33.56~<36.71 g/L、36.71~<40.16 g/L、40.16~51.10 g/L依次为3、2、1、0分,构建的首发PDAP预测模型的受试者工作特征(ROC)曲线下面积(AUC)为0.659(95%CI:0.550~0.768,P=0.009),最佳截断值为6.5分,灵敏度、特异度分别为42.3%、84.7%,约登指数0.27。结论构建的风险评估模型对预测首发PDAP具有一定准确性,同时模型使用的计分规则简便易行,临床使用相对容易,可为临床医务人员发现潜在PDAP提供依据。

Objective To explore the risk factors of first-episode peritoneal dialysis associated peritonitis(PDAP)in patients with peritoneal dialysis(PD),and construct a risk assessment model,so as to provide reference for early prevention and intervention.Methods From January 2020 to December 2022,189 hospitalized PD patients in the Department of Nephrology of our hospital were selected as the study objects,163 of whom did not have PDAP as the control group,26 of whom had PDAP for the first time as the case group.Results Among 189 patients,26 patients developed PDAP for the first time.26 strains of pathogenic bacteria were detected in bacterial culture;of which,16 were Gram-positive bacteria(61.54%),8 were Gram-negative bacteria(30.77%)and 2 were fungi(7.69%).Age,gender,place of residence,and albumin were risk factors for the first occurrence of PDAP in patients.Risk score model indicated that“Female or rural”was assigned a value of 2 points;age of 22-<43 years old,43-<55 years old,55-<66 years old,and 66-90 years old were assigned a value of 0,1,2,and 3,respectively;the values of albumin concentration of 25.30-<33.56 g/L,33.56-<36.71 g/L,36.71-<40.16 g/L,40.16-51.10 g/L were 3,2,1 and 0 points,respectively.The area under the receiver operating characteristic(ROC)curve(AUC)of the constructed first-episode PDAP prediction model was 0.659(95%CI:0.550-0.768,P=0.009),the optimal cut-off value was 6.5 points,and the sensitivity and specificity were 42.3%and 84.7%,respectively;the Youden index was 0.27.Conclusion The risk assessment model had a certain accuracy to predict the initial PDAP,and the scoring rule of the model was simple and easy to use in clinical practice,which provided the basis for clinical medical staff to find potential PDAP.