基于倾向性评分匹配的达格列净相关性酮症危险因素分析

Analysis of Risk Factors of Dapagliflozin-Associated Ketosis Based on Propensity Score Matching

目的为降低达格列净相关性酮症的发生风险提供参考。方法选取医院2021年1月至12月服用达格列净并进行酮体监测的住院患者716例,按是否发生酮症分为酮症组(50例)和非酮症组(666例)。采用倾向性评分匹配法进行1∶1匹配,匹配成功后,记录并比较患者的药物使用情况、疾病诊断情况、实验室检查指标等。在单因素分析结果的基础上,采用二元回归模型分析达格列净相关性酮症的危险因素。结果共成功匹配46对患者。单因素分析结果显示,两组患者的2型糖尿病、低空腹C肽水平、高糖化血红蛋白水平和肾功能异常占比的差异均显著(P<0.05)。二元回归分析结果显示,达格列净相关性酮症的危险因素为2型糖尿病[OR=13.718,95%CI(1.586,118.647),P=0.017]和低空腹C肽水平[OR=2.718,95%CI(1.024,7.213),P=0.045]。结论2型糖尿病和低空腹C肽水平均为达格列净相关性酮症的危险因素,对于合并2种因素的高危患者,在服用达格列净时,要密切关注酮症的发生风险。

Objective To provide a reference for decreasing the risk of dapagliflozin-associated ketosis.Methods A total of 716 inpatients taking dapagliflozin and receiving ketone monitoring in the hospital from January 2021 to December 2021 were selected and divided into the ketosis group(50 cases)and non-ketosis group(666 cases)according to whether ketosis occurred.The propensity score matching(PSM)method was used for matching(1∶1).After successful matching,the patients'information on drug use,disease diagnosis and laboratory examination indicators was recorded and compared.Based on the univariate analysis,the binary regression model was used to analyze the risk factors of dapagliflozin-associated ketosis.Results A total of 46 pairs of cases were successfully matched.The univariate analysis showed that the proportions of diabetes mellitus type 2(T2DM),low level of fasting C-peptide,high level of glycosylated hemoglobin and abnormal renal function between the two groups were significantly different(P<0.05).The results of binary regression analysis showed that the risk factors of dapagliflozin-associated ketosis were T2DM[OR=13.718,95%CI(1.586,118.647),P=0.017]and low level of fasting C-peptide[OR=2.718,95%CI(1.024,7.213),P=0.045].Conclusion T2DM and low level of fasting C-peptide are risk factors of dapagliflozin-associated ketosis.For patients with the above two risk factors,we should closely monitor the risk of ketosis when they taking dapagliflozin.