Boutin Amélie, Chassé Michaël, Shemilt Michèle, Lauzier François, Moore Lynne, Zarychanski Ryan, Lacroix Jacques, Fergusson Dean A, Desjardins Philippe, Turgeon Alexis F
Centre Hospitalier Universitaire (CHU) de Québec Research Center, Population Health and Optimal Health Practices Research Unit, Traumatology-Emergency-Critical Care Medicine, Université Laval, Québec, QC, Canada.
Syst Rev. 2014 Jun 18;3:66. doi: 10.1186/2046-4053-3-66.
Anemia is a prevalent condition in critically ill patients and red blood cell transfusions are frequent. Although transfusions at low hemoglobin levels have been shown to be associated with equivalent or better outcomes than higher hemoglobin thresholds, clinical equipoise persists in patients with traumatic brain injury considering their susceptibility to secondary cerebral insults such as those from hypoxemia.
Our objectives are to estimate the frequency of red blood cell transfusion in patients with traumatic brain injury and to evaluate transfusion thresholds, determinants and outcomes associated with transfusion strategies.We will conduct a systematic review of cohort studies and randomized controlled trials of patients with traumatic brain injury. We will search MEDLINE, Embase, BIOSIS and the Cochrane Library for eligible studies. Two independent reviewers will screen all identified references. Studies including adult patients with traumatic brain injury reporting data on red blood cell transfusions will be eligible. We will collect data on baseline demographics, trauma characteristics, hemoglobin thresholds, blood transfusions and clinical outcomes (mortality, length of stay, complications, and so on). Two independent reviewers will extract data using a standardized form. We will pool cumulative incidences using DerSimonian and Lair random-effect models after a Freeman-Tukey transformation to stabilize variances. We will pool risk ratios or mean differences with random-effect models and Mantel-Haenszel or inverse variance methods in order to evaluate the association between red blood cell transfusion and potential determinants or outcomes. Sensitivity and subgroup analysis according to timing of red blood cell transfusion, traumatic brain injury severity, year of conduction of the study, risk of bias, notably, are planned.
We expect to observe high heterogeneity in the proportion of transfused patients across studies and that the global proportion will be similar to the frequency observed in the general medical critically ill population. Our systematic review will allow us to better describe and understand current transfusion practices in patients with traumatic brain injury, a clinical population in which liberal transfusions are still advocated in the absence of evidence-based data.
CRD42014007402.
贫血在重症患者中普遍存在，红细胞输注频繁。尽管已表明低血红蛋白水平时输血与高于较高血红蛋白阈值时的结局相当或更好，但考虑到创伤性脑损伤患者易受继发性脑损伤（如低氧血症所致）影响，临床 equipoise 仍然存在。
我们的目标是估计创伤性脑损伤患者红细胞输注的频率，并评估输血阈值、决定因素以及与输血策略相关的结局。我们将对创伤性脑损伤患者的队列研究和随机对照试验进行系统评价。我们将在 MEDLINE、Embase、BIOSIS 和考克兰图书馆中检索符合条件的研究。两名独立的评审员将筛选所有识别出的参考文献。纳入报告红细胞输注数据的成年创伤性脑损伤患者的研究将符合条件。我们将收集基线人口统计学、创伤特征、血红蛋白阈值、输血情况和临床结局（死亡率、住院时间、并发症等）的数据。两名独立的评审员将使用标准化表格提取数据。在进行 Freeman-Tukey 变换以稳定方差后，我们将使用 DerSimonian 和 Lair 随机效应模型汇总累积发病率。我们将使用随机效应模型以及 Mantel-Haenszel 或逆方差方法汇总风险比或均值差异，以评估红细胞输注与潜在决定因素或结局之间的关联。计划根据红细胞输注时间、创伤性脑损伤严重程度、研究开展年份、偏倚风险等进行敏感性和亚组分析。
我们预计各研究中输血患者比例存在高度异质性，且总体比例将与普通内科重症患者中观察到的频率相似。我们的系统评价将使我们能够更好地描述和理解创伤性脑损伤患者目前的输血实践，这是一个在缺乏循证数据的情况下仍提倡宽松输血的临床群体。
系统评价注册 PROSPERO：CRD42014007402 。
