【摘要】 目的 基于倾向性匹配评分研究心型脂肪酸结合蛋白(H-FABP)对脓毒性心肌病的早期诊断价值。方法 回顾性分析2019年1月至2022年6月期间入住陕西省人民医院重症医学科(ICU)的258例脓毒症及脓毒性休克患者临床资料，根据是否发生脓毒性心肌病分为非脓毒性心肌病组和脓毒性心肌病组，行倾向性评分匹配后对两组临床资料进行比较；使用Spearman相关分析H-FABP和急性生理与慢性健康状况评分Ⅱ(APACHEⅡ评分)、肌酸激酶同工酶(CK-MB)、高敏心肌肌钙蛋白T(hs-cTnT)、N末端B型利钠肽前体(NT-proBNP)的相关性；绘制CK-MB、hs-cTnT、NT-proBNP和H-FABP预测脓毒性心肌病的ROC曲线，计算AUC、最佳临界值、敏感度和特异度。结果 倾向性评分匹配前，两组患者年龄(P<0.001),入ICU时心率(P=0.034)、呼吸频率(P=0.018)、平均动脉压(P=0.008)、WBC(P=0.032)、乳酸(P<0.001),APACHEⅡ评分(P=0.001),SOFA评分(P=0.023),住院天数(P=0.034),住院病死率(P<0.001)差异有统计学意义，其余指标差异无统计学意义(P>0.05);倾向性评分匹配后，两组患者基线资料平衡，差异无统计学意义(P>0.05),匹配后两组CK-MB(P<0.001)、hs-cTnT(P<0.001)、NT-proBNP(P<0.001)、H-FABP(P<0.001)和住院病死率(P=0.038)差异有统计学意义。相关性分析显示，H-FABP和APACHEⅡ评分(r=0.251,P=0.010)、CK-MB(r=0.311,P=0.001)、hs-cTnT(r=0.479,P<0.001)、NT-proBNP(r=0.608,P<0.001)均呈正相关。CK-MB、hs-cTnT、NT-proBNP、H-FABP预测脓毒性心肌病的曲线下面积(AUC)分别为0.742(95%CI 0.647～0.838,P<0.001)、0.841(95%CI 0.763～0.919,P<0.001)、0.736(95%CI 0.641～0.832,P<0.001)和0.926(95%CI 0.872～0.979,P<0.001)。CK-MB预测脓毒性心肌病发生的最佳截断值为21.45 ng/mL,敏感度为0.846,特异度为0.596;hs-cTnT最佳截断值为0.125 ng/mL,敏感度为0.827,特异度为0.788;NT-proBNP最佳截断值为5569.50 pg/mL,敏感度为0.846,特异度为0.577;H-FABP最佳截断值为26.81 ng/mL,敏感度为1.000,特异度为0.769。结论 H-FABP早期预测脓毒性心肌病的诊断效能优于CK-MB、hs-cTnT、NT-proBNP等传统心肌损伤指标。更多还原

【Abstract】 Objective To study the value of heart-type fatty acid-binding protein(H-FABP) in early diagnosis of septic cardiomyopathy based on propensity score matching. Methods 258 patients with sepsis and septic shock who admitted to ICU of Shanxi Provincial People′s Hospital from January 2019 to June 2022 were retrospectively analyzed and were divided into non-septic cardiomyopathy group and septic cardiomyopathy group depend on whether they had septic cardiomyo-pathy. The clinical date of the patients were collected and compared between two groups after propensity score matching. Spearman correlation was used to analyze the correlation of H-FABP with acute physiology and chronic health evaluation Ⅱ(APACHE Ⅱ), creatine kinase-MB(CK-MB), high sensitivity cardiac troponin T(hs-cTnT), N-terminal B-type natriuretic peptide precursor(NT-proBNP). The ROC curves of CK-MB, hs-cTnT, NT-proBNP and H-FABP for predicting septic cardiomyopathy were drawn and the AUC, optimal threshold, sensitivity and specificity were calculated. Results Before propensity score matching, there were significant differences in age(P<0.001), heart rate(P=0.034), respiratory rate(P=0.018), mean arterial pressure(P=0.008), WBC(P=0.032), lactic acid(P<0.001), APACHEⅡ score(P=0.001), SOFA score(P=0.023), the days of hospital stay(P=0.034) and inpatient mortality(P<0.001) between the two groups. The difference of other indexes was not statistically significant(P>0.05). The baseline data in the two groups balanced after propensity score matching and the difference was not statistically significant(P>0.05). After propensity score matching, there were statistically significant differences in CK-MB(P<0.001), hs-cTnT(P<0.001), NT-proBNP(P<0.001), H-FABP(P<0.001) and inpatient mortality(P=0.038). Correlation analysis showed that H-FABP was positively correlated with APACHEⅡ score(r=0.251, P=0.010), CK-MB(r=0.311, P=0.001), hs-cTnT(r=0.479, P<0.001) and NT-proBNP(r=0.608, P<0.001). The areas under the curve of CK-MB, hs-cTnT, NT-proBNP and H-FABP in predicting septic cardiomyopathy were 0.742(95%CI 0.647-0.838, P<0.001), 0.841(95%CI 0.763-0.919, P<0.001), 0.736(95%CI 0.641-0.832, P<0.001) and 0.926(95%CI 0.872-0.979, P<0.001) respectively. The best cut-off value of CK-MB for diagnosis of septic cardiomyopathy was 21.45 ng/mL, the sensitivity was 0.846, the specificity was 0.596. The best cut-off value of hs-cTnT was 0.125 ng/mL, the sensitivity was 0.827, the specificity was 0.788. The best cut-off value of NT-proBNP was 5569.50 pg/mL, the sensitivity was 0.846, the specificity was 0.577. The best cut-off value of H-FABP was 26.81 ng/mL, and the sensitivity was 1.000, the specificity was 0.769. Conclusions H-FABP is superior to CK-MB, hs-cTnT, NT-proBNP in early diagnosis of septic cardiomyopathy.更多还原