评分量表在全弓置换术后急性肾功能损伤中的预测效果评估

Comparison of clinical risk score systems in predicting acute kidney injury after total arch replacement

目的评估几种量化评分标准在全弓置换术后并发急性肾功能损伤(AKI)中的预测效能,寻找预测能力最优的评分标准,为患者的早期诊断提供参考。方法回顾性分析258例行全弓置换患者资料,其中术后并发AKI 123例纳入AKI组,剩余患者纳入非AKI组(n=135),参考简易临床风险评分(SCRS)、心脏术后急性肾损伤评分(AKSCI)、简易肾脏指数评分(SRI)、STS、Cleveland Clinic五项评分量表,分析各项评分在AKI预测中的ROC曲线,并比较各评分量表的曲线下面积(AUC)。结果两组患者基线期资料差异无统计学意义(P>0.05),在单因素分析中,所有评分量表在两组比较中均差异有统计学意义(P<0.05)。在ROC曲线分析中,AUC计算如下:SCRS=0.771、STS=0.730、Cleveland Clinic=0.769、AKSCI=0.728、SRI=0.655,SRI评分要明显低于其他评分且差异有统计学意义(P<0.05),AKSCI与Cleveland Clinic相比差异也存在统计学意义(P<0.05),其余比较中差异无统计学意义(P>0.05)。结论评分量表操作简单,为全弓置换术后早期诊断AKI提供参考,其中SRI及AKSCI评分准确性较差,不适用于全弓置换术后AKI的判断,而其余3项评分中,SCRS的预测效能可能要略优于STS、Cleveland Clinic评分,可以应用于全弓置换术后AKI的早期诊断。

Objective To evaluate the efficacy of several clinical risk scores in predicting acute kidney injury （AKI） after total arch replacement and to find the most optimal predictive score for early diagnosis in postoperative AKI. Methods Data of 258 patients underwent total arch replacement in Guangzhou Military Hospital from March 2007 to March 2017 were retrospectively analyzed. 123 of them had postoperative AKI and were assigned into the AKI group, and the rest were assigned into the non- AKI group （n = 135）. Five clinical risk scores systems： SCRS, AKSCI, SRI, STS, and Cleveland Clinic were analyzed using ROC curves and the area under the curve of each score were compared. Results There were no significant differences in baseline and intraoperative data between AKI group and non - AKI group. There was significant difference among these 5 scores （ P 〈 0. 05 ）. In the ROC curve analysis, the areas under the curves of SCRS, STS, Cleveland Clinic, AKSCI and SRI were 0. 771, 0. 730, 0. 769, 0. 728 and 0. 655, respectively. SRI score was significantly lower than the other scores. The comparison between AKSCI score and Cleveland Clinic score was also statistically significant. Conclusion Clinical risk score systems are simple to use and provide a reference for the early di- agnosis of postoperative AKI after total arch replacement. SRI and AKSCI scores are lack of accuracy and may not be ap- propriate in predicting postoperative AKI. And among the remaining three scores, SCRS＇s predictive efficacy may be slightly better than the other ones.