这是用户在 2025-7-25 19:08 为 https://app.immersivetranslate.com/pdf-pro/b76e8f4f-bec8-45b0-8c68-63a532b37131/ 保存的双语快照页面，由 沉浸式翻译 提供双语支持。了解如何保存？

Early physiologic changes after awake prone positioning predict clinical outcomes in patients with acute hypoxemic respiratory failure
清醒俯卧位后的早期生理变化可预测急性低氧性呼吸衰竭患者的临床结果

Matias Olmos 1 1 ^(1){ }^{1}, Mariano Esperatti 1 , 2 1 , 2 ^(1,2){ }^{1,2}, Nora Fuentes 1 , 2 1 , 2 ^(1,2){ }^{1,2}, Anabel Miranda Tirado 1 1 ^(1){ }^{1}, María Eugenia Gonzalez 1 , 2 1 , 2 ^(1,2){ }^{1,2}, Hiromi Kakisu 1 1 ^(1){ }^{1}, Juan Suarez 1 1 ^(1){ }^{1}, Manuel Tisminetzky 3 3 ^(3){ }^{3}, Veronica Barbaresi 1 1 ^(1){ }^{1}, Ignacio Santomil 1 1 ^(1){ }^{1}, Alejandro Bruhn Cruz 4 4 ^(4){ }^{4}, Domenico Luca Grieco 5 , 6 5 , 6 ^(5,6){ }^{5,6} and Bruno L. Ferreyro 3 ∗ 3 ∗ ^(3**){ }^{3 *} ( ©
马蒂亚斯·奥尔莫斯 1 1 ^(1){ }^{1} 、马里亚诺·埃斯佩拉蒂 1 , 2 1 , 2 ^(1,2){ }^{1,2} 、诺拉·富恩特斯 1 , 2 1 , 2 ^(1,2){ }^{1,2} 、安娜贝尔·米兰达·蒂拉多 1 1 ^(1){ }^{1} 、玛丽亚·尤金妮亚·冈萨雷斯 1 , 2 1 , 2 ^(1,2){ }^{1,2} 、柿须博美 1 1 ^(1){ }^{1} 、胡安·苏亚雷斯 1 1 ^(1){ }^{1} 、曼努埃尔·蒂斯米内茨基 3 3 ^(3){ }^{3} 、维罗妮卡·巴巴雷西 1 1 ^(1){ }^{1} 、伊格纳西奥·桑托米尔 1 1 ^(1){ }^{1} 、亚历杭德罗·布鲁恩·克鲁兹 4 4 ^(4){ }^{4} 、多梅尼科·卢卡·格里科 5 , 6 5 , 6 ^(5,6){ }^{5,6} 和布鲁诺·费雷罗 3 ∗ 3 ∗ ^(3**){ }^{3 *} （ ©

© 2024 Springer-Verlag GmbH Germany, part of Springer Nature
© 2024 Springer-Verlag GmbH 德国，隶属于 Springer Nature

Abstract  抽象

Purpose: The optimal physiologic parameters to monitor after a session of awake prone positioning in patients with acute respiratory failure are not well understood. This study aimed to identify which early physiologic changes after the first session of awake prone positioning are linked to the need for invasive mechanical ventilation or death in patients with acute respiratory failure. Methods: We performed a secondary analysis of a prospective cohort study of adult patients with acute respiratory failure related to coronavirus disease 2019 (COVID-19) treated with awake prone positioning. We assessed the association between relative changes in physiological variables (oxygenation, respiratory rate, pCO 2 pCO 2 pCO_(2)\mathrm{pCO}_{2} and respiratory rate-oxygenation [ ROX ] [ ROX ] [ROX][\mathrm{ROX}] index) within the first 6 h of the first awake prone positioning session with treatment failure, defined as endotracheal intubation and/or death within 7 days.
目的：急性呼吸衰竭患者清醒俯卧位后监测的最佳生理参数尚不清楚。本研究旨在确定第一次清醒俯卧位后的哪些早期生理变化与急性呼吸衰竭患者需要有创机械通气或死亡有关。方法：我们对接受清醒俯卧位治疗的与 2019 年冠状病毒病 （COVID-19） 相关的急性呼吸衰竭成年患者的前瞻性队列研究进行了二次分析。我们评估了第一次清醒俯卧位的前 6 小时内生理变量（氧合、呼吸频率 pCO 2 pCO 2 pCO_(2)\mathrm{pCO}_{2} 和呼吸频率-氧合 [ ROX ] [ ROX ] [ROX][\mathrm{ROX}] 指数）的相对变化与治疗失败（定义为气管插管和/或 7 天内死亡）之间的关联。

Results: 244 patients [70 female (29%), mean age 60 (standard deviation [SD] 13) years] were included. Seventy-one (29%) patients experienced awake prone positioning failure. ROX index was the main physiologic predictor. Patients with treatment failure had lower mean [SD] ROX index at baseline [5 (1.4) versus 6.6 (2.2), p < 0.0001 p < 0.0001 p < 0.0001p<0.0001 ] and within 6 h of prone positioning [5.6 (1.7) versus 8.7 (2.8), p < 0.0001 p < 0.0001 p < 0.0001p<0.0001 ]. After adjusting for baseline characteristics and severity, a relative increase of the ROX index compared to baseline was associated with lower odds of failure [odds ratio (OR) 0.37; 95% confidence interval (CI) 0.25-0.54 every 25% increase]. Conclusion: Relative changes in the ROX index within 6 h of the first awake prone positioning session along with other known predictive factors are associated with intubation and mortality at day 7.
结果：纳入 244 名患者 [70 名女性 （29%），平均年龄 60 岁（标准差 [SD] 13） 岁]。71 名 （29%） 患者经历了清醒俯卧位失败。ROX 指数是主要的生理预测因子。治疗失败的患者在基线时 [5 （1.4） vs 6.6 （2.2）， p < 0.0001 p < 0.0001 p < 0.0001p<0.0001 ] 和俯卧位 6 小时内 [5.6 （1.7） vs 8.7 （2.8）， p < 0.0001 p < 0.0001 p < 0.0001p<0.0001 ] 的平均 [SD] ROX 指数较低。在调整基线特征和严重程度后，与基线相比，ROX 指数的相对增加与较低的失败几率相关[比值比（OR）0.37;95%置信区间（CI）每增加 25%增加 0.25-0.54]。结论：第一次清醒俯卧位后 6 小时内 ROX 指数的相对变化以及其他已知的预测因素与第 7 天的插管和死亡率相关。

Keywords: Prone positioning, Respiratory failure, High flow nasal cannula, COVID-19, Respiratory distress syndrome
关键词：俯卧位，呼吸衰竭，高流量鼻插管，COVID-19，呼吸窘迫综合征

Introduction  介绍

Prone positioning has been shown to decrease mortality and increase ventilator-free days in patients with moder-ate-severe acute respiratory distress syndrome (ARDS) receiving invasive mechanical ventilation [1, 2]. During the coronavirus disease 2019 (COVID-19) pandemic, awake prone positioning was widely applied despite controversial evidence stemming both from randomized controlled trials and observational studies [3-9]. Recent meta-analyses and guidelines suggest that awake prone positioning may reduce the risk of endotracheal intubation in patients with COVID-19-related acute respiratory failure [10-13]. However, up to 30 % 30 % 30%30 \% of these patients may still receive endotracheal intubation, with an incidence of mortality up to 50 % 50 % 50%50 \% [3, 7, 11]. Importantly, delaying intubation with high-flow nasal cannula may itself detrimentally affect outcome [14], highlighting the need to early identify patients at high risk of failure despite this intervention.
俯卧位已被证明可以降低接受有创机械通气的中度重度急性呼吸窘迫综合征（ARDS）患者的死亡率并增加无呼吸机天数[1,2]。在 COVID-19 大流行期间，尽管随机对照试验和观察性研究的证据存在争议，但清醒俯卧位仍被广泛应用[3-9]。最近的 meta 分析和指南表明，清醒俯卧位可降低 COVID-19 相关急性呼吸衰竭患者气管插管的风险[10-13]。然而，这些患者 30 % 30 % 30%30 \% 中仍有可能接受气管插管，死亡率高达 50 % 50 % 50%50 \% [3,7,11]。重要的是，延迟使用高流量鼻插管插管本身可能会对结果产生不利影响[14]，这凸显了尽管有这种干预，但仍需要及早识别失败高风险的患者。
Previous literature has described how awake prone positioning might improve certain physiologic parameters [15, 16]. However, only a few studies have evaluated which changes in these variables can identify those patients who will ultimately benefit from the intervention [17]. An improvement in oxygenation is often utilized as a metric of response to awake prone positioning [18]. However, relying solely on oxygenation may be limiting, as other physiological factors could also play a significant role [19, 20]. Determining which early changes after prone positioning are associated with a lower risk of endotracheal intubation might help determine the appropriate monitoring tools to guide decision-making [19, 20].
先前的文献描述了清醒俯卧位如何改善某些生理参数 [15， 16]。然而，只有少数研究评估了这些变量的哪些变化可以识别最终将从干预中受益的患者[17]。氧合的改善通常被用作对清醒俯卧位的反应量度[18]。然而，仅依靠氧合可能会受到限制，因为其他生理因素也可能发挥重要作用 [19， 20]。确定俯卧位后的哪些早期变化与较低的气管插管风险相关可能有助于确定适当的监测工具来指导决策[19,20]。
We sought to evaluate the association between physiological parameters (hypoxemia, pCO 2 pCO 2 pCO_(2)\mathrm{pCO}_{2}, respiratory rate and the respiratory rate-oxygenation [ROX] index [21]) at baseline and within the first 6 h of the first session of awake prone positioning and the risk of endotracheal intubation and/or mortality at 7 days in patients with COVID-19-related acute hypoxemic respiratory failure.
我们试图评估基线和第一次清醒俯卧位前 6 小时内的生理参数（低氧血症、 pCO 2 pCO 2 pCO_(2)\mathrm{pCO}_{2} 呼吸频率和呼吸频率-氧合 [ROX] 指数 [21]）与 COVID-19 相关急性低氧性呼吸衰竭患者 7 天时气管插管和/或死亡的风险之间的关联。

Methods  方法

Study design and setting
研究设计和设置

We performed a secondary analysis of a prospective cohort study of patients with COVID-19 related acute respiratory failure within the Argentine Collaborative Group on High Flow and Prone Positioning (NCT05178212) [7]. The study period was between November 2020 and February 2022. The Internal Review Board approved the study (Reference number 2919/2143/2020) and guaranteed a waiver of informed consent. The project was carried out in accordance with
我们对阿根廷高流量和俯卧位协作组（Argentine Collaborative Group on High Flow and Prone Positioning， NCT05178212）中一项针对 COVID-19 相关急性呼吸衰竭患者的前瞻性队列研究进行了二次分析[7]。研究时间为 2020 年 11 月至 2022 年 2 月。内部审查委员会批准了这项研究（参考编号 2919/2143/2020），并保证放弃知情同意。该项目是按照

Take-home message  带回家的信息

In this secondary analysis of a prospective cohort study, we assessed which baseline characteristics and early physiologic changes were associated with awake prone positioning failure (intubation or death within 7 days) among patients with acute respiratory failure.
在一项前瞻性队列研究的二次分析中，我们评估了哪些基线特征和早期生理变化与急性呼吸衰竭患者的清醒俯卧位失败（插管或 7 天内死亡）相关。
A relative increase in the respiratory rate-oxygenation (ROX) index within 6 h of the first awake prone positioning session, along with baseline characteristics (lower severity scores and higher ROX index), were associated with a lower risk of treatment failure.
第一次清醒俯卧位后 6 小时内呼吸频率氧合 （ROX） 指数的相对增加，以及基线特征（较低的严重程度评分和较高的 ROX 指数），与较低的治疗失败风险相关。
the Helsinki Declaration [22]. This study included a subset of the original cohort of patients at Hospital Privado de la Comunidad, where subsequent physiologic measures were systematically collected. Additional details are published elsewhere [7].
赫尔辛基宣言 [22]。这项研究包括 Hospital Privado de la Comunidad 原始患者队列的一部分，该医院系统地收集了随后的生理测量值。其他详见其他专题[7]。

Population  人口

We included patients older than 17 years admitted to the intensive care unit (ICU) with acute respiratory failure secondary to COVID-19 and receiving awake prone positioning. Acute respiratory failure was defined by the requirement of supplemental oxygen greater than 4 L / 4 L / 4L//4 \mathrm{~L} / minute to achieve oxygen saturation ( SpO 2 SpO 2 SpO_(2)\mathrm{SpO}_{2} ) of 92 % 92 % 92%92 \% or higher; increased work of breathing with the use of accessory muscles, and a respiratory rate greater than 30 breaths / min / min //min/ \mathrm{min}. Only patients that tolerated prone positioning for at least 3 h were included, consistent with the time of subsequent physiologic assessments. Patients with decreased level of consciousness (Glasgow Coma Scale [ GCS ] < 9 [ GCS ] < 9 [GCS] < 9[\mathrm{GCS}]<9 ), presence of shock requiring vasopressors (norepinephrine equivalent dose > 0.1 mcg / kg / min > 0.1 mcg / kg / min > 0.1mcg//kg//min>0.1 \mathrm{mcg} / \mathrm{kg} / \mathrm{min} ), pregnancy or use of awake prone positioning prior to ICU admission were excluded. We also excluded patients who had an immediate need for intubation according to the attending physicians’ criteria.
我们纳入了因继发于 COVID-19 的急性呼吸衰竭而入住重症监护病房（ICU）并接受清醒俯卧位的 17 岁以上患者。急性呼吸衰竭的定义是需要补充氧气超过 4 L / 4 L / 4L//4 \mathrm{~L} / 分钟才能达到血氧饱和度 （ SpO 2 SpO 2 SpO_(2)\mathrm{SpO}_{2} ） 或 92 % 92 % 92%92 \% 更高;使用辅助肌肉增加呼吸功，呼吸频率大于 30 次 / min / min //min/ \mathrm{min} 呼吸。仅纳入耐受俯卧位至少 3 小时的患者，与随后的生理评估时间一致。意识水平下降（格拉斯哥昏迷量表 [ GCS ] < 9 [ GCS ] < 9 [GCS] < 9[\mathrm{GCS}]<9 ）、存在需要血管加压药（去甲肾上腺素当量剂量）的休克 > 0.1 mcg / kg / min > 0.1 mcg / kg / min > 0.1mcg//kg//min>0.1 \mathrm{mcg} / \mathrm{kg} / \mathrm{min} 、怀孕或在入住 ICU 前使用清醒俯卧位的患者被排除在外。我们还根据主治医师的标准排除了立即需要插管的患者。

Awake prone positioning protocol and timing of measurements
清醒俯卧位定位协议和测量时间

At ICU admission, high flow nasal oxygen was initiated through a nasal cannula (Optiflow, Fisher and Paykel Healthcare). A first set of physiologic parameters (see below) were recorded in the supine position (time 0 ) and within 6 h of awake prone positioning (time 1). Additional details on the protocol are included in the online electronic supplementary material (ESM) (Section 1 and Fig. S1).
在入住 ICU 时，通过鼻插管（Optiflow、Fisher 和 Paykel Healthcare）启动高流量鼻氧。在仰卧位（时间 0）和清醒俯卧位后 6 小时内（时间 1）记录第一组生理参数（见下文）。有关该协议的其他详细信息包含在在线电子补充材料（ESM）中（第 1 节和图 S1）。

Variables and measurements
变量和测量

Data on patients demographics (age and sex), body mass index, pre-hospitalization comorbidities, severity scores at ICU admission (Acute Physiologic Assessment and
有关患者人口统计（年龄和性别）、体重指数、住院前合并症、入住 ICU 时严重程度评分的数据（急性生理评估和
Chronic Health Evaluation [APACHE] II and Sequential Organ Failure Assessment [SOFA]), type of medical treatment (corticosteroids, tocilizumab), ROX index and receipt of vasopressors at admission were collected. We also collected information on whether patients had a do-not-intubate order, and the duration of study interventions.
收集慢性健康评估 [APACHE] II 和序贯器官衰竭评估 [SOFA]）、药物类型（皮质类固醇、托珠单抗）、ROX 指数和入院时接受血管加压药的情况。我们还收集了有关患者是否有不插管命令以及研究干预持续时间的信息。
The main exposures of interest were the physiologic parameters at baseline (immediately before the first prone positioning) and its relative changes within the first 6 h . The parameters of interest were PaO 2 PaO 2 PaO_(2)\mathrm{PaO}_{2} to FiO 2 FiO 2 FiO_(2)\mathrm{FiO}_{2} ratio, SpO 2 SpO 2 SpO_(2)\mathrm{SpO}_{2} to FiO 2 FiO 2 FiO_(2)\mathrm{FiO}_{2} ratio, respiratory rate, ROX index, pCO 2 pCO 2 pCO_(2)\mathrm{pCO}_{2}, and heart rate. A relative change was computed for each parameter as follows: value at time 1 (prone)value at time 0 (supine) / value at time 0 . For instance, a 25 % 25 % 25%25 \% relative change in PaO 2 PaO 2 PaO_(2)\mathrm{PaO}_{2} to FiO 2 FiO 2 FiO_(2)\mathrm{FiO}_{2} ratio implies a corresponding improvement in oxygenation [23]. Based on preliminary findings, we focused on relative changes in the ROX index as the key parameter.
感兴趣的主要暴露是基线时（第一次俯卧位之前）的生理参数及其在前 6 小时内的相对变化。感兴趣的参数是 PaO 2 PaO 2 PaO_(2)\mathrm{PaO}_{2} 比率 FiO 2 FiO 2 FiO_(2)\mathrm{FiO}_{2} 、 SpO 2 SpO 2 SpO_(2)\mathrm{SpO}_{2} 比率、呼吸 FiO 2 FiO 2 FiO_(2)\mathrm{FiO}_{2} 频率、ROX 指数 pCO 2 pCO 2 pCO_(2)\mathrm{pCO}_{2} 和心率。计算每个参数的相对变化如下：时间 1 的值（俯卧）时间 0（仰卧）的值/时间 0 的值。例如，比 FiO 2 FiO 2 FiO_(2)\mathrm{FiO}_{2} 值的 25 % 25 % 25%25 \% PaO 2 PaO 2 PaO_(2)\mathrm{PaO}_{2} 相对变化意味着氧合作用的相应改善[23]。基于初步研究结果，我们将 ROX 指数的相对变化作为关键参数。
The primary composite outcome was treatment failure, defined as receiving endotracheal intubation or death within 7 days of the first prone positioning session. We used a composite outcome given the plausibility that certain patients might die without receiving endotracheal intubation (i.e.: do-not-intubate order at baseline or during follow-up). Ignoring that pathway would result in incorrectly labelling those cases as having a successful treatment response. The decision to intubate was made by the treating team, who was unaware of the study aims. Treatment failure at 28-days was the secondary outcome.
主要综合结果是治疗失败，定义为在第一次俯卧位后 7 天内接受气管插管或死亡。考虑到某些患者可能在未接受气管插管的情况下死亡（即：基线或随访期间的不插管顺序），我们使用了复合结局。忽视该途径将导致错误地将这些病例标记为具有成功的治疗反应。插管的决定是由治疗团队做出的，他们不知道研究目的。28 天时的治疗失败是次要结果。

Statistical analysis  统计分析

We used descriptive statistics to characterize the baseline features of study participants. Baseline characteristics between patients with treatment failure and success (extubated and alive at 7 days) were compared with Fisher’s exact test (for categorical variables) and student’s T T TT test or Wilcoxon rank sum test, according to the observed distribution (for continuous variables). The incidence of the primary outcome was described as frequencies with proportions and its timing of occurrence with cumulative density plots.
我们使用描述性统计来表征研究参与者的基线特征。根据观察到的分布（对于连续变量），将治疗失败和成功（拔管并存活 7 天）的患者之间的基线特征与 Fisher 精确检验（对于分类变量）和学生检 T T TT 验或 Wilcoxon 秩和检验进行比较。主要结果的发生率被描述为频率及其发生时间与累积密度图的比例。
We fitted a multivariable logistic regression model to evaluate the association between patients’ baseline characteristics (age, sex, disease severity [SOFA and APACHE scores] and baseline ROX index) and relative changes in the ROX index with the primary outcome of treatment failure. The ROX was highlighted among other relative physiologic changes based on the bivariate analysis’ results, its reproducibility and acceptability, ease of use (i.e.: no need for arterial blood gas sample) and lack of missing data. Other baseline characteristics were included in the model to estimate the independent association of the relative changes in ROX
我们拟合了多变量逻辑回归模型来评估患者的基线特征（年龄、性别、疾病严重程度 [SOFA 和 APACHE 评分] 和基线 ROX 指数）与 ROX 指数的相对变化与治疗失败的主要结局之间的关联。基于双变量分析的结果、其可重复性和可接受性、易用性（即：不需要动脉血气样本）和缺乏缺失数据，ROX 在其他相对生理变化中得到了强调。模型中包括其他基线特征，以估计 ROX 相对变化的独立关联
index while adjusting for potential confounders, selected by subject matter and previous literature [17]. Each estimate of association was expressed as odds ratios (OR) with 95% confidence interval (CI). The discrimination performance of the model was reported using the area under the receiver operating characteristic (AU-ROC) curve. Additionally, we fitted an identical model without the inclusion of the relative changes in the ROX index (i.e. only baseline variables) to assess whether the model’s performance improved with the addition of relative changes in this parameter. Models were compared with the likelihood ratio test (nested models) and the difference in the AU-ROC curve. The rationale was to assess whether awaiting to observe the initial physiologic changes after prone positioning would enhance prediction.
指数，同时调整潜在的混杂因素，按主题和以前的文献选择[17]。每个关联估计值都表示为比值比 （OR） 和 95% 置信区间 （CI）。使用受试者工作特征下面积（AU-ROC）曲线报告了模型的辨别性能。此外，我们拟合了一个相同的模型，而不包含 ROX 指数的相对变化（即仅基线变量），以评估模型的性能是否随着该参数的相对变化的增加而提高。将模型与似然比检验（嵌套模型）和 AU-ROC 曲线的差异进行比较。其基本原理是评估等待观察俯卧位后的初始生理变化是否会增强预测。

Secondary and sensitivity analyses
二次分析和敏感性分析

We conducted a post hoc secondary analysis to estimate the incidence of treatment failure among the following subgroups: patients with a ROX index lower than 4.88 (both during supine and prone positioning) and based on their relative change in the ROX index within 6 h (no change or worsening ROX, relative change equal or lower than 25 % 25 % 25%25 \% and relative change equal or lower than 50 % 50 % 50%50 \% ). We estimated the sensitivity, specificity, negative and positive predictive values for each of these cut-off points to identify treatment failure.
我们进行了事后二次分析，以估计以下亚组治疗失败的发生率：ROX 指数低于 4.88 的患者（仰卧位和俯卧位期间）并基于他们在 6 小时内 ROX 指数的相对变化（ROX 无变化或恶化，相对变化等于或低于 25 % 25 % 25%25 \% ，相对变化等于或低于 50 % 50 % 50%50 \% ).我们估计了每个临界点的敏感性、特异性、阴性和阳性预测值，以确定治疗失败。
We also performed a series of sensitivity analyses: changing the time frame for defining treatment failure from 7 to 28 days; excluding patients with a do-not-intubate order at baseline; and assessing a simplified model including only relative changes on the SpO 2 SpO 2 SpO_(2)\mathrm{SpO}_{2} to FiO 2 FiO 2 FiO_(2)\mathrm{FiO}_{2} ratio or respiratory rate, instead of the ROX index.
我们还进行了一系列敏感性分析：将定义治疗失败的时间范围从 7 天更改为 28 天;排除基线时有不插管命令的患者;并评估一个简化的模型，仅包括 SpO 2 SpO 2 SpO_(2)\mathrm{SpO}_{2} 比率 FiO 2 FiO 2 FiO_(2)\mathrm{FiO}_{2} 或呼吸频率的相对变化，而不是 ROX 指数。

Results  结果

Study population  研究人群

Between November 2020 and February 2022, 343 patients were admitted to the ICU with COVID-19 related acute respiratory failure, of which 244 tolerated prone positioning for at least 3 h and were included (Fig. 1). The main characteristics of the subjects are described in Table 1. Among study participants, the mean (standard deviation, SD) age was 60 (13.2) years and 70 ( 29 % ) 70 ( 29 % ) 70(29%)70(29 \%) were female. Mean (SD) APACHE II and SOFA scores at baseline were 16.3 (5.1) and 3.5 (1.5), respectively. Seventeen patients (7%) had a do-not-intubate order at baseline.
2020 年 11 月至 2022 年 2 月期间，有 343 名患者因 COVID-19 相关急性呼吸衰竭入住 ICU，其中 244 名患者耐受俯卧位至少 3 小时并被纳入（图 1）。受试者的主要特征如表 1 所述。在研究参与者中，平均（标准差，SD）年龄为 60 （13.2） 岁， 70 ( 29 % ) 70 ( 29 % ) 70(29%)70(29 \%) 为女性。基线时的平均 （SD） APACHE II 和 SOFA 分数分别为 16.3 （5.1） 和 3.5 （1.5）。17 名患者 （7%） 在基线时有不插管命令。

Incidence and timing of study outcomes
研究结果的发生率和时间

Among the 244 patients included, 71 ( 29 % ) 71 ( 29 % ) 71(29%)71(29 \%) experienced the primary outcome. Among these patients, 15 (6%) died without being intubated ( 12 of these had a do-notintubate order at baseline). Treatment failure at 28 days occurred in 92 ( 38 % 38 % 38%38 \% ) patients. All-cause death at 28 days occurred in 57 ( 23 % 23 % 23%23 \% ) patients. Supplementary Tables S1,
在纳入的 244 名患者中， 71 ( 29 % ) 71 ( 29 % ) 71(29%)71(29 \%) 经历了主要结果。在这些患者中，15 名 （6%） 在未插管的情况下死亡（其中 12 名在基线时有不插管的命令）。92 （ 38 % 38 % 38%38 \% ） 名患者在 28 天时治疗失败。57 （ 23 % 23 % 23%23 \% ） 名患者在 28 天时发生全因死亡。补充表 S1，
  1. *Correspondence: bruno.ferreyro@utoronto.ca
    *对应：bruno.ferreyro@utoronto.ca
    3 3 ^(3){ }^{3} Department of Medicine, Interdepartmental Division of Critical Care Medicine, Sinai Health System and University Health Network, University of Toronto, Toronto, Canada
    加拿大多伦多大学西奈卫生系统和大学健康网络重症监护医学跨科部医学 3 3 ^(3){ }^{3} 系
    Full author information is available at the end of the article
    完整的作者信息可在文章末尾找到