Objective: To evaluate the association between fluid balance within 48 hours after initiating continuous renal replacement therapy (CRRT) and 90-day mortality of patients with cardiac surgery-associated acute kidney injury (CSA-AKI). Methods: Adult patients who received CRRT for CSA-AKI for more than 24 hours between January 2016 and December 2021 in the First Affiliated Hospital of Nanjing Medical University were prospectively included. Exposures were fluid input, fluid output, fluid balance and percentage of fluid overload change (defined as weight-standardized fluid input/output difference) within 48 hours after CRRT initiation. The primary outcome was 90-day mortality. The relationship between the exposures and 90-day mortality was determined by restricted cubic spline and segmented Cox regression. Results: The study enrolled 262 patients, and 90-day mortality was 60.3% (158/262). There were 171 males and 91 females, with a median age of 64 (54, 71) years. The dead patients (n=158) had more fluid input [6.43 (5.62, 7.34) L vs 6.07 (5.09, 6.90) L, P=0.013], greater fluid balance [-0.01 (-1.36, 1.34) L vs -0.65 (-1.71, 0.42) L, P=0.005] and higher percentage of fluid overload change [-0.01% (-2.17%, 2.01%) vs -0.85% (-2.49%, 0.52%), P=0.013] within 48 hours after CRRT initiation than the survived patients (n=104). The cumulative fluid balance and fluid overload changes have a "J" curve relationship with 90-day mortality. Higher positive fluid balance (adjusted HR per 1 L increase above 0.5 L=1.33, 95%CI: 1.17-1.50) and greater proportion of fluid overload changes (adjusted HR per 1% increase above 0.7%=1.20, 90%CI: 1.11-1.30) were associated with an increased risk of death. Cumulative fluid input was linearly related to the 90-day mortality. Higher input (adjusted HR per 1 L increase=1.13, 95%CI: 1.03-1.24) was associated with an increased risk of death. The cumulative fluid output showed a U-shaped relationship with risk of death. Both lower output (adjusted HR per 1 L increase below 6.7 L=0.73, 95%CI: 0.58-0.90) and higher output (adjusted HR per 1 L increase above 6.7 L=1.24, 95%CI: 1.01-1.53) were associated with increased risk of death. Conclusion: In patients with CSA-AKI, excessive fluid input, insufficient or excessive output and a positive fluid balance or percentage of fluid overload change within the first 48 hours after CRRT initiation are associated with increased risk of 90-day mortality.
目的： 评估心脏手术相关急性肾损伤（CSA-AKI）患者启动连续性肾脏替代治疗（CRRT）后48 h液体平衡与90 d死亡率之间的关系。 方法： 前瞻性纳入2016年1月至2021年12月在南京医科大学第一附属医院因CSA-AKI接受CRRT 超过 24 h的成年患者。暴露变量为CRRT启动后48 h液体入量、液体出量、液体平衡和液体超负荷改变比例，主要终点为90 d死亡率。使用限制性立方样条曲线和分段Cox回归模型分析暴露变量与主要终点之间的潜在非线性关系。 结果： 共纳入262例患者，男171例，女91例，年龄M（Q1，Q3）为64（54，71）岁，90 d死亡率60.3%（158/262）。与90 d生存组（104例）相比，90 d死亡组（158例）启动CRRT后48 h液体入量更多［6.43（5.62，7.34）L比6.07（5.09，6.90）L，P=0.013］，液体平衡更大［-0.01（-1.36，1.34）L比-0.65（-1.71，0.42）L，P=0.005］、液体超负荷改变比例更大［-0.01%（-2.17%，2.01%）比-0.85%（-2.49%，0.52%），P=0.013］。累计液体平衡和液体超负荷改变与死亡风险呈“J”形曲线关系，较大的液体正平衡（0.5 L以上每增加1 L的校正HR=1.33，95%CI：1.17~1.50）和液体超负荷改变（0.7%以上每增加1%的校正HR=1.20，90%CI：1.11~1.30）与死亡风险升高相关。累计液体入量与死亡风险呈线性关系，较高的入量与死亡风险升高相关（每增加1 L的校正HR=1.13，95%CI：1.03~1.24）。累计液体出量与死亡风险呈“U”形曲线关系，较低的出量（6.7 L以下每增加1 L的校正HR=0.73，95%CI：0.58~0.90）与较高的出量（6.7 L以上每增加1L的HR=1.24，95%CI：1.01~1.53）均与死亡风险升高相关。 结论： 在CSA-AKI患者中，CRRT启动后48 h内较多液体入量、不足或过多的液体出量、过大液体平衡和过大液体超负荷改变均与90 d死亡风险升高相关。.