【摘要】 目的 探讨心搏骤停（CA）患者体外心肺复苏（ECPR）失败的影响因素，并构建预测ECPR失败的列线图。方法 回顾性分析2021年2月—2024年3月淮安市第二人民医院救治的149例CA患者的临床资料，根据ECPR结局分为成功组和失败组，对两组临床特征进行比较。通过Logistic回归模型筛查ECPR失败的危险因素，构建预测ECPR失败的模型，并验证其效能。结果 149例患者中，有113例ECPR失败，失败率为75.84%。失败组院外CA比例（69.03%vs. 38.89%）、CA至开始常规CPR时间[（7.35±2.48）min vs.(9.56±3.02)min]、常规CPR至体外膜肺氧合（ECMO）时间[（41.83±8.56）min vs.(32.26±6.49)min]、去甲肾上腺素用量[（6.79±1.74）μg/（kg·min）vs.(4.61±1.36)μg/（kg·min）]、急性生理及慢性健康状况评分Ⅱ（APACHEⅡ）评分[（16.29±4.25）分vs.(14.47±3.67)分]及序贯器官衰竭评估（SOFA）评分[（11.43±3.20）分vs.(9.03±2.58)分]和成功组比较差异均有统计学意义（均P<0.05）。Logistic回归分析结果显示，院外CA(OR=3.478,95%CI:1.242～9.738)、常规CPR至ECMO时间长（OR=1.163,95%CI:1.076～1.258）、去甲肾上腺素用量大（OR=2.525,95%CI:1.712～3.722）、SOFA评分高（OR=1.287,95%CI:1.072～1.545）为CA患者ECPR失败的危险因素。利用上述4项指标构建预测ECPR失败的列线图模型，受试者操作特征（ROC）曲线下面积为0.848(95%CI:0.786～0.911)，灵敏度、特异度分别为83.30%、83.10%，校准曲线预测ECPR失败的概率接近实际概率，拟合优度HL检验，χ~2=8.301,P=0.405。结论 根据CA地点、常规CPR至ECMO时间、去甲肾上腺素及SOFA评分构建的列线图模型可有效预测ECPR失败的风险。更多还原

【Abstract】 Objective To investigate the influencing factors of extracorporeal cardiopulmonary resuscitation(ECPR)failure in patients with respiratory cardiac arrest(CA), and to construct a column chart for predicting ECPR failure.Methods A retrospective analysis was conducted on the clinical data of 149 patients with CA treated in our hospital from February 2021 to March 2024. According to the ECPR outcome, they were grouped into a successful group and a failed group, and the clinical characteristics of the two groups were compared. The Logistic regression model was used to screen for risk factors for ECPR failure, and a model for predicting ECPR failure was constructed and its effectiveness was validated.Results Among 149 patients, 113 cases of ECPR failed, with a failure rate of 75.84%.There were statistically significant differences between the failed group and the successful group in terms of the proportion of out-of-hospital CA(69.03% vs. 38.89%), time from CA to initiation of CPR [(7.35±2.48) min vs.(9.56±3.02) min], time from routine CPR to extracorporeal membrane oxygenation(ECMO)[(41.83±8.56) min vs.(32.26±6.49)min], norepinephrine dosage [(6.79±1.74) μg/(kg·min) vs.(4.61±1.36) μg/(kg·min)], APACHE Ⅱ score [(16.29±4.25)points vs.(14.47 ± 3.67) points], and SOFA score [(11.43 ± 3.20) points vs.(9.03 ± 3.20) points]. Logistic regression analysis showed that out of hospital CA(OR=3.478, 95%CI: 1.242-9.738), long time from routine CPR to ECMO(OR=1.163, 95%CI: 1.076-1.258), high dose of norepinephrine(OR=2.525, 95%CI: 1.712-3.722), and high SOFA score(OR=1.287, 95%CI: 1.072-1.545) were risk factors for ECPR failure in patients with CA. A column chart model for predicting ECPR failure was constructed using the above four indicators. The area under the ROC curve was 0.848(95%CI: 0.786-0.911), with sensitivity and specificity of 83.30% and 83.10%, respectively. The probability of the calibration curve in predicting ECPR failure was close to the actual probability, and the goodness of fit HL test showedχ~2=8.301, P=0.405.Conclusion A column chart model constructed based on CA location, conventional CPR to ECMO time, norepinephrine, and SOFA score can effectively predict the risk of ECPR failure.更多还原