高频振荡通气治疗急性呼吸窘迫综合征的荟萃分析和试验序贯分析

The effects of high frequency oscillation ventilation on patients with acute respiratory distress syndrome：a Meta-analysis and sequential analysis of trial

目的：系统评价高频振荡通气（HFOV）与传统机械通气（CV）对急性呼吸窘迫综合征（ARDS）患者治疗及预后的影响。方法检索PubMed、EMBase、Cochrane central register of controlled trials及中国知网、万方数据库，收集2014年5月之前有关对比HFOV与CV对ARDS成人患者预后影响的随机对照试验，并采用STATA 12.0软件就HFOV与CV对病死率及生理参数的影响进行荟萃分析，用TSA 0.9软件对病死率进行试验序贯分析，用Excel 2003软件对生理参数绘制折线图。结果共纳入7篇文献、1731例患者。7篇研究均提供了患者的生理参数数据，6篇研究（1705例患者）提供了病死率。结果显示：与CV相比，HFOV并不能降低ARDS患者院内或30 d病死率〔相对危险度（RR）＝0.93，95%可信区间（95%CI）＝0.70～1.24，P＝0.63〕，不能提高脱离呼吸机生存率（RR＝1.05，95%CI＝0.72～1.54，P＝0.80）、依赖呼吸机生存率（RR＝1.23，95%CI＝0.65～2.35，P＝0.52），亦不能降低治疗失败率（RR＝0.89，95%CI＝0.50～1.56，P＝0.67）；不良事件如低血压（RR＝0.89，95%CI＝0.07～10.99，P＝0.93）、酸中毒（RR＝1.05，95%CI＝0.43～2.56，P＝0.91）、呼吸机漏气（RR＝0.74，95%CI＝0.31～1.80，P＝0.51）的发生率差异也无统计学意义。但是HFOV能更快地改善患者生理指标及降低呼吸机参数，包括氧合指数、呼气末正压、潮气量、平均气道压、动脉血pH值、动脉二氧化碳分压、吸入氧浓度、动脉血氧分压与吸入氧浓度比值。采用试验序贯分析病死率，通过校正随机误差、计算界值曲线的方法发现，能得出最终结论所需样本量应为3874例；试验序贯分析结果显示：累计Z值未穿过传统界值（即传统意义上的P＝0.05），同时也未跨过序贯分析界值，表明在尚未达到期望样本量（3874例）之前， HFOV与CV在病死率上无差异，目前的证据尚无法提前得出最终决定性结论。结论与CV相比，HFOV并不能降低ARDS患者的病死率，但能改善患者生理指标；最终结论仍需要更大样本的随机对照研究验证。

ObjectiveTo evaluate the effect of high frequency oscillation ventilation （HFOV） vs. conventional mechanical ventilation （CV） on the treatment and prognosis of adult patients with acute respiratory distress syndrome （ARDS）.Methods Published articles concerning randomized controlled trials （RCTs） about the effect of HFOV vs. CV on the prognosis of adult patients with ARDS published before May 2014 were retrieved from PubMed, EMBase, Cochrane central registry of controlled trials, CNKI and Wanfang Data. The mortality and data of physiological parameters were analyzed with STATA 12.0, and the mortality rate was also analyzed by trial sequential analysis with TSA 0.9, and the line chart was drawn with Microsoft Office Excel 2003.Results Seven trials with 1 731 patients met the criteria, all of them recorded the physiological parameters data, and mortality rate was mentioned in 6 trials （1 705 patients）. Compared with CV, HFOV did not show any statistically significant beneficial effects on mortality [relative risk （RR） = 0.93, 95% confidence interval （95%CI） = 0.70-1.24,P = 0.63], and other clinical outcomes, including survival without mechanical ventilation （RR = 1.05, 95%CI = 0.72-1.54,P = 0.80）, survival on mechanical ventilation （RR = 1.23, 95%CI = 0.65-2.35,P = 0.52）, or treatment failure （RR = 0.89, 95%CI = 0.50-1.56,P = 0.67）. The risk factors of adverse events including hypotension （RR = 0.89, 95%CI = 0.07-10.99,P =0.93）, acidosis （RR = 1.05, 95%CI = 0.43-2.56,P = 0.91）, and air leakage from ventilator （RR = 0.74, 95%CI = 0.31-1.80,P =0.51） were similar. But the physiologic parameters of patients and parameters of ventilator in HFOV group, including oxygenation index, positive end-expiratory pressure, tidal volume, mean airway pressure, arterial pH, partial pressure of arterial carbon dioxide, fraction of inspired oxygen, ratio of partial pressure of arterial oxygen to fraction of inspired oxygen, were better than those in the CV group. Methods adapted from formal interim monitoring boundaries applied to cumulative Meta-analysis showed that the evidence failed by a considerable degree to meet the standards for forgoing studies, and the necessary sample was 3 874 patients. Trial sequential analysis also showed that the accumulatedZ-score did not cross the traditional boundary （P = 0.05） and interim monitoring boundaries. This result indicated that there was no significant difference between CV and HFOV on mortality before the number of needed sample reached （3 874 cases）. We could not get a definitive conclusion with current evidences.ConclusionsCompared with CV, the use of HFOV in ARDS was not associated with a significant reduction in mortality. But the physiologic parameters of patients in HFOV group were better than those in the CV group. More RCTs are needed to draw a definitive conclusion.