Sun Shiren, Ma Feng, Li Qiaoneng, Bai Ming, Li Yangping, Yu Yan, Huang Chen, Wang Hanmin, Ning Xiaoxuan
Department of Nephrology, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China.
Department of Geriatrics, Xijing Hospital, Fourth Military Medical University, Xi'an, Shaanxi, China.
Interact Cardiovasc Thorac Surg. 2017 Oct 1;25(4):548-554. doi: 10.1093/icvts/ivx210.
Acute kidney injury (AKI) is a serious complication after cardiac surgery and is associated with increased in-hospital deaths. Renal replacement therapy (RRT) is becoming a routine strategy for severe AKI. Our goal was to evaluate the risk factors for death and RRT dependence in patients with AKI after cardiac surgery.
We included 190 eligible adult patients who had AKI following cardiac surgery and who required RRT at our centre from November 2010 to March 2015. We collected preoperative, intraoperative, postoperative and RRT data for all patients.
In this cohort, 87 patients had successful RRT in the hospital, whereas 103 patients had RRT that failed (70 deaths and 33 cases of RRT dependence). The multivariable logistic analysis identified old age [odds ratio (OR): 1.042, 95% confidence interval (CI): 1.012-1.074; P = 0.011], serum uric acid (OR: 1.015, 95% CI: 1.003-1.031; P = 0.024), intraoperative concentrated red blood cell transfusions (OR: 1.144, 95% CI: 1.006-1.312; P = 0.041), postoperative low cardiac output syndrome (OR: 3.107, 95% CI: 1.179-8.190; P = 0.022) and multiple organ failure (OR: 5.786, 95% CI: 2.115-15.832; P = 0.001) as factors associated with a higher risk for RRT failure. The prediction model (-4.3 + 0.002 × preuric acid + 0.10 × concentrated red blood cells + 0.04 × age + 1.12 × [low cardiac output syndrome = 1] + 1.67 × [multiple organ failure = 1]) based on the multivariate analysis had statistically significant different incriminatory power with an area under the curve of 0.786.
The prediction model may serve as a simple, accurate tool for predicting in-hospital RRT failure for patients with AKI following cardiac surgery.
急性肾损伤（AKI）是心脏手术后的一种严重并发症，与住院死亡率增加相关。肾脏替代治疗（RRT）正成为治疗严重AKI的常规策略。我们的目标是评估心脏手术后AKI患者死亡和依赖RRT的危险因素。
我们纳入了190例符合条件的成年患者，这些患者在2010年11月至2015年3月期间在我们中心接受心脏手术后发生AKI且需要RRT。我们收集了所有患者的术前、术中和术后以及RRT数据。
在这个队列中，87例患者在医院内成功进行了RRT，而103例患者的RRT失败（70例死亡和33例RRT依赖）。多变量逻辑分析确定高龄[比值比（OR）：1.042，95%置信区间（CI）：1.012 - 1.074；P = 0.011]、血清尿酸（OR：1.015，95% CI：1.003 - 1.031；P = 0.024）、术中浓缩红细胞输注（OR：1.144，95% CI：1.006 - 1.312；P = 0.041）、术后低心排血量综合征（OR：3.107，95% CI：1.179 - 8.190；P = 0.022）和多器官功能衰竭（OR：5.786，95% CI：2.115 - 15.832；P = 0.001）是与RRT失败风险较高相关的因素。基于多变量分析的预测模型（-4.3 + 0.002×术前尿酸 + 0.10×浓缩红细胞 + 0.04×年龄 + 1.12×[低心排血量综合征 = 1] + 1.67×[多器官功能衰竭 = 1]）的曲线下面积为0.786，具有统计学上显著的判别能力。
该预测模型可作为预测心脏手术后AKI患者院内RRT失败的一种简单、准确的工具。
