Department of Pediatrics, Northwestern University Feinberg School of Medicine, Chicago, IL.
Stanley Manne Children's Research Institute, Ann & Robert H. Lurie Children's Hospital of Chicago, Chicago, IL.
Pediatr Crit Care Med. 2024 Jan 1;25(1):24-36. doi: 10.1097/PCC.0000000000003320. Epub 2023 Jul 18.
In this systematic review and meta-analysis we asked: Do predictors of fluid responsiveness in children perform comparably: 1) in the PICU as in non-PICU settings? 2) in shock states compared with nonshock states? Additionally, 3) is there an association between preload responsiveness and clinical response?
Ovid Medline, PubMed, and Embase databases were searched from inception through May 2022.
Included studies reported physiological response to IV fluid administration in humans less than 18 years. Only studies reporting an area under the receiver operating characteristic curve (AUROC) were included for descriptive analysis. Only studies for which a se could be estimated were included for meta-analysis.
Title, abstract, full text screening, and extraction were completed by two authors (S.B.W., J.M.W.). Variables extracted included predictors ("tools") and outcome measures ("reference tests") of fluid responsiveness, demographic, and clinical variables.
We identified 62 articles containing 204 AUROCs for 55 tools, primarily describing mechanically ventilated children in an operating room or PICU. Meta-analysis across all tools showed poor predictive performance (AUROC, 0.66; 95% CI, 0.63-0.69), although individual performance varied greatly (range, 0.49-0.87). After controlling for PICU setting and shock state, PICU setting was associated with decreased predictive performance (coefficient, -0.56; p = 0.0007), while shock state was associated with increased performance (0.54; p = 0.0006). Effect of PICU setting and shock state on each tool was not statistically significant but analysis was limited by sample size. The association between preload responsiveness and clinical response was rarely studied but results did not suggest an association. Ultrasound measurements were prone to inherent test review and incorporation biases.
We suggest three opportunities for further research in fluid responsiveness in children: 1) assessing predictive performance of tools during resuscitation in shock states; 2) separating predictive tool from reference test when using ultrasound techniques; and 3) targeting decreasing time in a shock state, rather than just increase in preload.
在本系统评价和荟萃分析中，我们提出了以下问题：预测儿童液体反应性的指标在以下情况下表现是否相当：1）在儿科重症监护病房（PICU）和非 PICU 环境中？2）在休克状态与非休克状态之间？此外，3）前负荷反应性与临床反应之间是否存在关联？
通过 Ovid Medline、PubMed 和 Embase 数据库检索，检索时间从建库至 2022 年 5 月。
纳入的研究报告了 18 岁以下人群中静脉输液后生理反应。仅对报告了受试者工作特征曲线（ROC）曲线下面积（AUROC）的研究进行描述性分析。仅对可以估计出 SE 的研究进行荟萃分析。
标题、摘要、全文筛选和提取由两名作者（S.B.W.、J.M.W.）完成。提取的变量包括液体反应性的预测指标（“工具”）和结果指标（“参考测试”）、人口统计学和临床变量。
我们共识别出 62 篇包含 204 个 AUROC 的文章，涉及 55 种工具，主要描述了在手术室或 PICU 中机械通气的儿童。所有工具的荟萃分析显示，预测性能较差（AUROC，0.66；95%CI，0.63-0.69），尽管个体表现差异很大（范围，0.49-0.87）。在控制 PICU 环境和休克状态后，PICU 环境与预测性能下降相关（系数，-0.56；p=0.0007），而休克状态与性能提高相关（0.54；p=0.0006）。但工具的 PICU 环境和休克状态效应无统计学意义，分析受到样本量限制。前负荷反应性与临床反应之间的关联很少被研究，但结果并未表明存在关联。超声测量容易受到内在测试审查和纳入偏倚的影响。
我们建议在儿童液体反应性研究方面开展三项进一步研究：1）评估休克状态复苏过程中工具的预测性能；2）在使用超声技术时，将预测工具与参考测试分开；3）以减少休克状态的持续时间为目标，而不仅仅是增加前负荷。
