High versus low positive end-expiratory pressure (PEEP) levels for mechanically ventilated adult patients with acute lung injury and acute respiratory distress syndrome

  • Review
  • Intervention

Authors


Abstract

Background

Mortality in patients with acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) remains high. These patients require mechanical ventilation, but this modality has been associated with ventilator-induced lung injury. High levels of positive end-expiratory pressure (PEEP) could reduce this condition and improve patient survival.

Objectives

To assess the benefits and harms of high versus low levels of PEEP in patients with ALI and ARDS.

Search methods

We searched the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, 2013, Issue 4), MEDLINE (1950 to May 2013), EMBASE (1982 to May 2013), LILACS (1982 to May 2013) and SCI (Science Citation Index). We used the Science Citation Index to find references that have cited the identified trials. We did not specifically conduct manual searches of abstracts of conference proceedings for this review. We also searched for ongoing trials (www.trialscentral.org; www.clinicaltrial.gov and www.controlled-trials.com).    

Selection criteria

We included randomized controlled trials that compared the effects of two levels of PEEP in ALI and ARDS participants who were intubated and mechanically ventilated in intensive care for at least 24 hours.

Data collection and analysis

Two review authors assessed the trial quality and extracted data independently. We contacted investigators to identify additional published and unpublished studies. 

Main results

We included seven studies that compared high versus low levels of PEEP (2565 participants). In five of the studies (2417 participants), a comparison was made between high and low levels of PEEP with the same tidal volume in both groups, but in the remaining two studies (148 participants), the tidal volume was different between high- and low-level groups. We saw evidence of risk of bias in three studies, and the remaining studies fulfilled all criteria for adequate trial quality.

In the main analysis, we assessed mortality occurring before hospital discharge only in those studies that compared high versus low PEEP with the same tidal volume in both groups. With the three studies that were included, the meta-analysis revealed no statistically significant differences between the two groups (relative risk (RR) 0.90, 95% confidence interval (CI) 0.81 to 1.01), nor was any statistically significant difference seen in the risk of barotrauma (RR 0.97, 95% CI 0.66 to 1.42). Oxygenation was improved in the high-PEEP group, although data derived from the studies showed a considerable degree of statistical heterogeneity. The number of ventilator-free days showed no significant difference between the two groups. Available data were insufficient to allow pooling of length of stay in the intensive care unit (ICU). The subgroup of participants with ARDS showed decreased mortality in the ICU, although it must be noted that in two of the three included studies, the authors used a protective ventilatory strategy involving a low tidal volume and high levels of PEEP.

Authors' conclusions

Available evidence indicates that high levels of PEEP, as compared with low levels, did not reduce mortality before hospital discharge. The data also show that high levels of PEEP produced no significant difference in the risk of barotrauma, but rather improved participants' oxygenation to the first, third, and seventh days. This review indicates that the included studies were characterized by clinical heterogeneity.  

Résumé scientifique

Niveaux élevés versus bas de pression expiratoire positive (PEP) pour les patients adultes sous ventilation mécanique présentant une lésion pulmonaire aiguë et un syndrome de détresse respiratoire aiguë

Contexte

La mortalité des patients atteints de lésion pulmonaire aigüe (LPA) et du syndrome de détresse respiratoire aigüe (SDRA) demeure élevée. Ces patients nécessitent une ventilation mécanique, mais cette modalité est associée à des lésions pulmonaires induites par la ventilation elle-même. Des niveaux élevés de pression expiratoire positive (PEP) pourraient réduire cette affection et améliorer la survie des patients.

Objectifs

Évaluer les avantages et les inconvénients des niveaux élevés versus bas de PEP chez les patients atteints de LPA et du SDRA.

Stratégie de recherche documentaire

Nous avons effectué une recherche dans le registre Cochrane des essais contrôlés (CENTRAL) (The Cochrane Library 2013, numéro 4), MEDLINE (de 1950 à mai 2013), EMBASE (de 1982 à mai 2013), LILACS (de 1982 à mai 2013) et SCI (Science Citation Index). Nous avons utilisé le Science Citation Index pour trouver des références ayant cité les essais identifiés. Nous n'avons pas effectué spécifiquement pour cette revue de recherches manuelles de résumés d'actes de conférence. Nous avons également recherché des essais en cours (www.trialscentral.org, www.clinicaltrial.gov et www.controlled-trials.com).    

Critères de sélection

Nous avons inclus des essais contrôlés randomisés ayant comparé les effets de deux niveaux de PEP chez des participants souffrant de PLA et du SDRA qui avaient été intubés et mis sous ventilation mécanique en soins intensifs pendant au moins 24 heures.

Recueil et analyse des données

Deux auteurs de la revue ont évalué la qualité des essais et extrait les données de manière indépendante. Nous avons contacté des chercheurs pour identifier des études supplémentaires, publiées ou non publiées.

Résultats principaux

Nous avons inclus sept études ayant comparé des niveaux élevés et bas de PEP (2565 participants). Cinq des études (2417 participants) avaient comparé des niveaux élevé et bas de PEP avec le même volume courant dans les deux groupes, mais dans les deux autres études (148 participants) le volume courant était différent entre les groupes à niveaux élevé et bas. Nous avons identifié des risques de biais dans trois études, et les autres études remplissaient tous les critères de qualité méthodologique nécessaires.

Dans l'analyse principale, nous avons évalué la mortalité survenant avant la sortie de l'hôpital en n'utilisant que les études ayant comparé PEP élevée et faible avec le même volume courant dans les deux groupes. La méta-analyse des trois études incluses n'a révélé aucune différence statistiquement significative entre les deux groupes (risque relatif (RR) 0,90 ; intervalle de confiance (IC) à 95 % 0,81 à 1,01), et aucune différence statistiquement significative n'a été observée pour le risque de barotraumatisme (RR 0,97 ; IC 95 % 0,66 à 1,42). L'oxygénation avait été améliorée dans le groupe à PEP élevée, bien que les données issues des études se soient avérées présenter un degré considérable d'hétérogénéité statistique. Il n'y avait pas de différence significative entre les deux groupes quant au nombre de jours sans assistance respiratoire. Les données disponibles étaient insuffisantes pour permettre un regroupement par durée du séjour en unité de soins intensifs (USI). Il y avait une baisse de la mortalité en réanimation dans le sous-groupe de participants atteints du SDRA, mais il faut noter que dans deux des trois études incluses les auteurs avaient utilisé une stratégie de ventilation protective reposant sur un volume courant faible et des niveaux élevés de PEP.

Conclusions des auteurs

Les données disponibles indiquent que les niveaux élevés de PEP n'avaient pas réduit la mortalité avant la sortie de l'hôpital, par rapport aux niveaux bas. Les données montrent également que les niveaux élevés de PEP n'avaient pas entrainé de différence significative dans le risque de barotraumatisme, mais avaient plutôt amélioré l'oxygénation des participants aux premier, troisième et septième jours,. Cette revue indique que les études incluses étaient caractérisées par une hétérogénéité clinique.  

Resumo

Pressão expiratória final positiva (PEEP) alta versus baixa em pacientes adultos sob ventilação mecânica com lesão pulmonar aguda e síndrome da angústia respiratória aguda

Introdução

A mortalidade em pacientes com lesão pulmonar aguda e síndrome do desconforto respiratório agudo (SDRA) permanece elevada. Esses pacientes necessitam de ventilação mecânica, mas essa modalidade tem sido associada a lesão pulmonar induzida por ventilador. Altos níveis de pressão expiratória final positiva (PEEP) poderiam reduzir esta condição e melhorar a sobrevida do paciente.

Objetivos

Avaliar os benefícios e os riscos do PPEP com níveis alto versus baixo em pacientes com lesão pulmonar aguda e SARA sob ventilação mecânica.

Métodos de busca

Pesquisamos na Base de Dados Central de Ensaios Controlados da Cochrane (CENTRAL)(The Cochrane Library, 2013, Issue 4), MEDLINE (1950 a maio de 2013), EMBASE (1982 a maio de 2013), LILACS (1982 a maio de 2013) e SCI (Science Citation Index). Utilizamos o Science Citation Index para encontrar referências que citaram os ensaios clínicos identificados. Não realizamos uma busca manual de resumos contidos em anais de conferências para esta revisão. Buscamos também ensaios clínicos em andamento(nas bases www.trialscentral.org; www.clinicaltrial.gov e www. controlled-trials.com).

Critério de seleção

Incluímos ensaios clínicos randomizados que compararam os efeitos de dois níveis de PEEP em pacientes entubados com SDRA e lesão pulmonar aguda e sob ventilação mecânica em unidade de terapia intensiva por pelo menos 24 horas.

Coleta dos dados e análises

Dois revisores diferentes realizaram a análise da qualidade dos estudos e extraíramos dados independentemente. Contatamos os autores dos estudos para obter informações adicionais publicadas ou não. 

Principais resultados

Incluímos sete estudos que compararam os níveis altos versus baixos de PEEP (2.565 participantes). Em cinco dos sete estudos (2.417 participantes), foi feita comparação entre altos e baixos níveis de PEEP com o mesmo volume corrente em ambos os grupos, mas nos dois estudos restantes (148 participantes), o volume corrente foi diferente entre os grupos. Havia evidência de risco de viés em três estudos; os outros estudos preencheram todos os critérios de qualidade.

Na análise principal, avaliamos a mortalidade que ocorre antes da alta hospitalar apenas nos estudos que compararam PEEP de alto versus baixo nível com o mesmo volume corrente em ambos os grupos. Esta meta-análise, que incluiu três estudos, não demonstrou diferenças estatisticamente significativas entre os dois grupos (risco relativo, RR, 0,90, intervalo de confiança de 95%, 95% CI, 0,81-1,01), e não houve qualquer diferença estatisticamente significativa quanto ao risco de barotrauma (RR 0,97, 95% CI 0,66-1,42). Houve melhora da oxigenação no grupo de PEEP de alto nível, embora os dados obtidos a partir dos estudos tenham mostrado grau considerável de heterogeneidade estatística. Não houve diferença significativa entre os dois grupos quanto ao número de dias sem ventilado. Devido à falta de datos, não foi possível fazer uma meta-análise do tempo de permanência na unidade de terapia intensiva (UTI). O subgrupo de participantes com SARA teve diminuição da mortalidade na UTI, embora deva ser notado que, em dois dos três estudos incluídos, os autores utilizaram uma estratégia ventilatória protetora que envolve baixo volume corrente e PEEP de alto nível.

Conclusão dos autores

A evidência disponível indica que PEEP de altos níveis, em comparação com os níveis baixos, não reduziu a mortalidade antes da alta hospitalar. Os dados também mostraram que PEEP de alto nível não produziu nenhuma diferença significativa no risco de barotrauma, mas houve melhora da oxigenação dos participantes para o primeiro, terceiro e sétimo dias. Esta revisão indica que os estudos incluídostiveram grande heterogeneidade clínica.  

Notas de tradução

Tradução do Centro Cochrane do Brasil (João Luís Caldeira Breijão)

摘要

使用機械通氣且有急性肺損傷與急性呼吸窘迫症候群的成人患者之高與低吐氣末正壓(PEEP)比較

背景

急性肺損傷(ALI)與急性呼吸窘迫症候群(ARDS)患者的死亡率仍高。這些患者需要機械通氣,但這種方式與呼吸器導致的肺損傷有關;高吐氣末正壓(PEEP)可能減少這種狀況並改善患者的存活率。

目的

評估高與低PEEP在ALI與ARDS患者上的益處及危害比較。

搜尋策略

我們搜尋了Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, 2013年,Issue 4)、MEDLINE (1950年至2013年5月)、EMBASE (1982年至2013年5月)、LILACS (1982年至2013年5月)以及SCI (Science Citation Index)等資料庫。我們採用Science Citation Index來尋找引用已確認試驗的參考資料。我們沒有特別對會議論文的摘要進行人工搜尋。我們也搜尋了正在進行的試驗(www.trialscentral.org;www.clinicaltrial.govwww.controlled-trials.com)。

選擇標準

我們收錄了比較兩種PEEP對於在加護病房中待了至少24小時、已插管且使用機械通氣的ALI與ARDS患者影響的隨機對照試驗。

資料收集與分析

兩位作者獨立地評估試驗品質並摘錄資料。我們連絡調查員以確認額外已發表及未發表的研究。

主要結果

我們收錄了7份比較高與低PEEP的研究(2,565位受試者)。其中5份研究(2,417位受試者)在兩個組別中,以相同的潮氣量在高與低PEEP之間做比較,但其他2份研究(148位受試者)高與低PEEP組中的潮氣量則不同。我們在3份研究中看見偏誤風險的證據,而其他的研究則符合適當試驗品質的所有標準。

在主要分析中,我們只在兩個組別皆以相同潮氣量來比較高與低PEEP的研究裡評估出院前的死亡率。收錄的3份研究,統合分析顯示兩個組別間沒有統計上的顯著差異(相對風險(RR) 0.90, 95% 信賴區間(CI) 0.81至1.01),氣壓性傷害的風險也看不出統計上的顯著差異(RR 0.97, 95% CI 0.66至1.42)。雖然由研究取得的資料顯示有相當高的統計異質性,氧合作用在高PEEP組中仍有改善。在不需使用呼吸器的天數上,兩個組別間沒有顯著的差異。可取得的資料不夠充分,無法將在加護病房(ICU)中的停留期合併歸類。雖然3份收錄的研究中有2份作者使用包含低潮氣量與高PEEP的保護通氣策略,ARDS患者次群組在ICU中的死亡率顯示為降低。

作者結論

可取得的證據指出,高PEEP相較於低PEEP並沒有降低出院前的死亡率。資料也顯示高PEEP在氣壓性傷害的部分沒有產生顯著差異,反而是在第1、3與7天時改善患者的氧合作用。本文獻指出收錄研究的特點是臨床異質性。 

譯註

翻譯者:臺北醫學大學考科藍臺灣研究中心(Cochrane Taiwan)

本翻譯計畫由臺北醫學大學考科藍臺灣研究中心(Cochrane Taiwan)、台灣實證醫學學會及東亞考科藍聯盟(EACA)統籌執行
聯絡E-mail:cochranetaiwan@tmu.edu.tw

Plain language summary

High versus low levels of positive end-expiratory pressure in patients with acute lung injury and acute respiratory distress syndrome

Acute lung injury (ALI) and acute respiratory distress syndrome (ARDS) are acute and severe conditions affecting the structure and function of the lungs that are caused by increased permeability of the alveolar-capillary barrier leading to an inflammatory injury. The mortality rate of ALI and ARDS has decreased over time and is currently reported at 43%. Patients with ALI and ARDS require mechanical ventilation, but this intervention can cause ventilator-induced lung injury. For this reason, the therapeutic target for these patients is based on lung-protective ventilation. The use of high levels of positive end-expiratory pressure (PEEP) is part of the strategy aimed at reducing ventilator-induced lung injury. PEEP is a mechanical manoeuvre that exerts a positive pressure in the lung and is used primarily to correct the hypoxaemia caused by alveolar hypoventilation. In this Cochrane review, we assess the benefits and harms of high versus low levels of PEEP in patients with ALI and ARDS. The undertaking of this review was both relevant and necessary because the optimal level of PEEP in these patients is still controversial, and available evidence indicates no difference in mortality. We included seven studies involving a total of 2565 participants and found that high levels of PEEP, as compared with low levels, did not produce a reduction in hospital mortality, although we did see a trend towards decreased mortality. We also found evidence of clinical heterogeneity among the included studies (clinical heterogeneity concerns differences in participants, interventions, and outcomes that might have an impact on results from the use of PEEP). The studies included were of moderate to good quality. We did not find a significant difference with respect to barotrauma—defined as the presence of pneumothorax on chest radiography or a chest tube insertion for known or suspected pneumothorax. We furthermore ascertained that high levels of PEEP improved participants' oxygenation up to the first, third, and seventh days. The number of ventilator-free days showed no significant difference between the two groups (the term ventilator-free days refers to the number of days between successful weaning from mechanical ventilation and day 28 after study enrolment), and available data were insufficient to allow pooling of lengths of stay in the intensive care unit. Additional trials will be required to determine which patients should receive high PEEP levels and the best means of applying this intervention. 

Résumé simplifié

Niveaux élevés versus bas de pression expiratoire positive chez les patients présentant une lésion pulmonaire aiguë et un syndrome de détresse respiratoire aiguë

La lésion pulmonaire aiguë (LPA) et le syndrome de détresse respiratoire aiguë (SDRA) sont des affections aiguës et graves affectant la structure et la fonction pulmonaires qui sont causées par une perméabilité accrue de la barrière alvéolo-capillaire menant à une lésion inflammatoire. Le taux de mortalité des LPA et SDRA a diminué au fil du temps et est considéré aujourd'hui comme étant de 43 %. Les patients atteints de LPA et du SDRA nécessitent une ventilation mécanique, mais cette ventilation peut elle-même induire des lésions pulmonaires. Pour cette raison, l'objectif thérapeutique pour ces patients repose sur une ventilation protectrice des poumons. L'utilisation de niveaux élevés de pression expiratoire positive (PEP) fait partie de la stratégie visant à réduire les lésions pulmonaires induites par la ventilation. La PEP est une manœuvre mécanique qui exerce une pression positive dans les poumons et est principalement utilisée pour corriger l'hypoxémie causée par une hypoventilation alvéolaire. Dans cette revue Cochrane, nous évaluons les avantages et les inconvénients de niveaux élevés versus bas de PEP chez les patients souffrant de LPA et du SDRA. La réalisation de cette étude était à la fois pertinente et nécessaire parce que le niveau optimal de PEP chez ces patients est encore controversé et que les données disponibles ne montrent aucune différence de mortalité. Nous avons inclus sept études impliquant un total de 2565 participants et avons constaté que les niveaux élevés de PEP, par rapport aux niveaux bas, n'avaient pas entrainé une réduction de la mortalité à l'hôpital, bien que nous ayons pu observer une tendance à la diminution de la mortalité. Nous avons également trouvé des preuves d'une hétérogénéité clinique entre les études incluses (l'hétérogénéité clinique concerne les différences entre les participants, les interventions et les critères de résultat qui sont susceptibles d'avoir une incidence sur les résultats de l'utilisation de la PEP). Les études incluses étaient de qualité moyenne à bonne. Nous n'avons pas trouvé de différence significative quant au barotraumatisme - défini comme la présence d'un pneumothorax sur la radiographie du thorax ou l'insertion d'un drain thoracique pour pneumothorax avéré ou soupçonné. Nous avons en outre constaté que des niveaux élevés de PEP avaient amélioré l'oxygénation des participants jusqu'aux premier, troisième et septième jours. Il n'y avait pas de différence significative entre les deux groupes concernant le nombre de jours sans assistance respiratoire (le terme jours sans assistance respiratoire fait référence au nombre de jours entre le sevrage réussi de la ventilation mécanique et le 28ème jour suivant l'enrôlement dans l'étude), et les données disponibles étaient insuffisantes pour permettre le regroupement par durée de séjour en unité de soins intensifs. Des essais supplémentaires seront nécessaires pour déterminer quels patients devraient recevoir des niveaux de PEP élevés et quels sont les meilleurs moyens d'appliquer cette intervention.

Notes de traduction

Traduit par: French Cochrane Centre 16th July, 2013
Traduction financée par: Pour la France : Minist�re de la Sant�. Pour le Canada : Instituts de recherche en sant� du Canada, minist�re de la Sant� du Qu�bec, Fonds de recherche de Qu�bec-Sant� et Institut national d'excellence en sant� et en services sociaux.

Resumo para leigos

Níveis altos versus baixos de pressão expiratória final positiva em pacientes com lesão pulmonar aguda e síndrome da angústia respiratória aguda

A lesão pulmonar aguda (LPA) e a síndrome do desconforto respiratório do adulto (SARA) são condições graves e agudas que afetam a estrutura e função dos pulmões. Elas são causadas por um aumento da permeabilidade da barreira alvéolo-capilar, causando lesão inflamatória. A taxa de mortalidade por LPA e SARA tem diminuído com o tempo e atualmente está em torno de 43% dos pacientes. Doentes com LPA e SARA necessitam de ventilação mecânica, porém essa intervenção pode causar lesões induzidas pelo ventilador. Por essa razão, a estratégia de tratamento nesses pacientes tem sido a ventilação protetora do pulmão. O uso de altos níveis de pressão expiratória final positiva (PEEP) faz parte dessa estratégia, que objetiva reduzir a lesão induzida pelo ventilador mecânico. A PEEP é uma manobra mecânica que exerce uma pressão positiva no pulmão e é usada principalmente para corrigir a hipoxemia causada por hipoventilação alveolar. Nesta revisão da Cochrane, nós avaliamos os benefícios e os malefícios da PEEP em altos versus baixos níveis em pacientes com LPA e SARA. A realização desta revisão foi relevante e necessária, porque o nível ideal de PEEP nesses pacientes ainda é controverso, e as evidências disponíveis indicam que não há diferença quanto à mortalidade. Incluímos sete estudos com um total de 2.565 participantes, e encontramos que altos níveis de PEEP, em comparação com níveis baixos, não produziram redução na mortalidade hospitalar, apesar de se observar uma tendência para a diminuição da mortalidade. Nós também encontramos evidências de heterogeneidade clínica entre os estudos incluídos (heterogeneidade clínica diz respeito a diferenças entre os participantes, intervenções e resultados que possam ter um impacto sobre os resultados do uso de PEEP). Os estudos incluídos foram de qualidade moderada a boa. Não encontramos diferença significativa em relação ao barotrauma, definido como a presença de pneumotórax na radiografia de tórax ou necessidade de introduzir um tubo torácico para drenar casos de pneumotórax confirmados ou suspeitos. Encontramos também que altos níveis de PEEP provocaram uma melhora da oxigenação dos participantes até o primeiro, terceiro e sétimo dias. O número de dias sem ventilação mecânica (ou seja, o número de dias entre o desmame do ventilador até o dia 28 após a inclusão no estudo) não foi significativamente diferente entre os grupos e os dados disponíveis eram insuficientes para permitir o agrupamento dos pacientes por tempo de permanência na unidade de terapia intensiva. Estudos adicionais serão necessários para determinar quais pacientes devem receber altos níveis de PEEP e as melhores formas de usar essa intervenção.

Notas de tradução

Tradução do Centro Cochrane do Brasil (João Luís Caldeira Breijão)

淺顯易懂的口語結論

急性肺損傷與急性呼吸窘迫症候群患者之高與低吐氣末正壓(PEEP)比較

急性肺損傷(ALI)與急性呼吸窘迫症候群(ARDS)為急性且嚴重的狀況,是由肺泡-微血管屏障的滲透性提高而導致的發炎損傷,影響肺部的結構與功能。ALI與ARDS的死亡率已隨著時間降低,目前的記錄為43%。ALI與ARDS患者需要機械通氣,但這個干預措施可能造成呼吸器相關的肺損傷。因為這個理由,這些患者的治療目標是以保護肺部的換氣法為基礎。採用高吐氣末正壓(PEEP),是以減少呼吸器相關肺損傷為目標的策略的一部分。PEEP是一種在肺部使用正壓的機械策略,主要矯正因肺泡換氣不足而造成的低血氧。在本Cochrane文獻中,我們評估了高與低PEEP在ALI與ARDS患者上的比較。本文獻任務工作為切題且必要的,因為這些患者中的最佳化PEEP依然具有爭議,而且可取得的證據指出死亡率沒有差異。我們收錄了7份包含2,565位受試者的研究,發現雖然我們的確看見死亡率有降低的趨勢,但高PEEP相較於低PEEP並沒有降低醫院的死亡率。我們也發現在收錄研究中的臨床異質性證據(臨床異質性是考量受試者、干預策略與結果的差異,可能對使用PEEP的效果造成影響)。收錄的研究品質為中等至良好。我們沒有發現關於氣壓性傷害(定義為在胸腔X光中出現氣胸或因已確診或疑似的氣胸而插入胸管)的顯著差異。此外,我們確定高PEEP在第1、3及7天時,改善患者的氧合作用。在不需使用呼吸器的天數上,兩個組別間沒有顯示顯著的差異 (「不需使用呼吸器的天數」指的是成功脫離呼吸器與研究登記後28天之間的天數),並且可取得的資料不夠充分,無法將在加護病房(ICU)的停留期合併歸類。為了判定哪些患者需要接受高PEEP治療以及進行這項干預措施最好的方式,將需要做額外的試驗。

譯註

翻譯者:臺北醫學大學考科藍臺灣研究中心(Cochrane Taiwan)

本翻譯計畫由臺北醫學大學考科藍臺灣研究中心(Cochrane Taiwan)、台灣實證醫學學會及東亞考科藍聯盟(EACA)統籌執行
聯絡E-mail:cochranetaiwan@tmu.edu.tw

Summary of findings(Explanation)

Summary of findings for the main comparison. High versus low levels of PEEP for mechanically ventilated adult patients with acute lung injury and acute respiratory distress syndrome
  1. 1 The baseline risk is low (the risk of presenting the outcome in participants who did not receive the intervention).
    2 There is clinical heterogeneity in relation to subgroups of participants who receive high levels of PEEP (participants with ALI and ARDS); how to apply those high PEEP levels.
    3 The quality of evidence was downgraded because of a minimal overlap among the studies. There the P value for heterogeneity was < 0,00001 and I2 was 86%.
    4 The quality of evidence was downgraded because of a minimal overlap among the studies. There the P value for heterogeneity was 0,0001 and the I2 was 82%.
    5 The quality of evidence was downgraded because of a minimal overlap among the studies. There the P value for heterogeneity was 0,0002 and the I2 was 85%.
    6 The baseline risk is low (the risk of presenting the outcome in participants who did not receive the intervention).
    7 The quality of evidence was downgraded because of a minimal overlap among the studies. There the P value for heterogeneity was 0,005 and the I2 was 87%.

High versus low levels of PEEP for mechanically ventilated adult patients with acute lung injury and acute respiratory distress syndrome
Patient or population: participants mechanically ventilated, adult participants with acute lung injury and acute respiratory distress syndrome.
Settings:
Intervention: high versus low levels of PEEP.
OutcomesIllustrative comparative risks* (95% CI)Relative effect
(95% CI)
No of participants
(studies)
Quality of the evidence
(GRADE)
Comments
Assumed riskCorresponding risk
Control High versus low levels of PEEP
Mortality before hospital discharge Study population1 RR 0.9
(0.81 to 1.01)2
2299
(3 studies)
⊕⊕⊕⊕
high
 
369 per 1000 332 per 1000
(299 to 373)
Low1
0 per 1000 0 per 1000
(0 to 0)
Moderate1
390 per 1000 351 per 1000
(316 to 394)
Oxygen efficiency (PaO2/FIO2). Day 1 The mean oxygen efficiency (PaO2/FIO2) day 1 in the intervention groups was
41.31 higher
(24.11 to 58.52 higher)
 2337
(5 studies)
⊕⊕⊕⊝
moderate 3
 
Oxygen efficiency (PaO2/FIO2). Day 3 The mean oxygen efficiency (PaO2/FIO2) day 3 in the intervention groups was
42.51 higher
(25 to 60.02 higher)
 2059
(5 studies)
⊕⊕⊕⊝
moderate 4
 
Oxygen efficiency (PaO2/FIO2). Day 7 The mean oxygen efficiency (PaO2/FIO2) day 7 in the intervention groups was
24.77 higher
(0.92 lower to 50.45 higher)
 1379
(4 studies)
⊕⊕⊕⊝
moderate 5
 
Barotrauma Study population6 RR 0.97
(0.66 to 1.42)
2504
(6 studies)
⊕⊕⊕⊕
high
 
90 per 1000 87 per 1000
(59 to 127)
Low6
0 per 1000 0 per 1000
(0 to 0)
Ventilator-free days (only means) The number of mean ventilator-free days (only means) in the intervention groups was
1.89 higher
(3.58 lower to 7.36 higher)
 644
(2 studies)
⊕⊕⊕⊝
moderate 7
 
*The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI).
CI: Confidence interval; RR: Risk ratio.
GRADE Working Group grades of evidence:
High quality: Further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: We are very uncertain about the estimate.

Background

Description of the condition

Acute lung injury (ALI) is caused by increased permeability of the alveolar-capillary barrier leading to an inflammatory injury to the lung with accumulation of a protein-rich pulmonary oedema, haemorrhage, a procoagulant tendency and invasion of neutrophils, macrophages and cytokines. These inflammatory damages lead to diffuse alveolar damage-the histopathologic correlate of ALI. This early, exudative phase is followed by a proliferative phase and may proceed to a fibrotic phase.

ALI is defined by the clinical features of hypoxaemia (arterial oxygen tension/fractional inspired oxygen (PaO2/FIO2) ≤ 300) regardless of the level of positive end-expiratory pressure (PEEP), bilateral pulmonary infiltrates and lack of evidence of left heart failure. Two distinct causes of ALI are known. Primary ALI can be caused by direct injury to the lung (e.g. pneumonia). Secondary ALI is caused by an indirect lung injury within the setting of a systemic process (e.g. sepsis). A more serious form of ALI is acute respiratory distress syndrome (ARDS), which has the same clinical characteristics as ALI, except that the PaO2/FIO2 in ARDS is ≤ 200 (Bernard 1994).

A new definition of ARDS was proposed last year by a panel of experts (Ranieri 2012). This new definition contains certain variations with respect to the one previously prescribed. Furthermore the authors of this recent definition found a slight improvement in predictive validity compared with the previous definition (Bernard 1994).  

The incidence of ALI and ARDS is variable among studies, ranging from 5 to 86 cases per 100,000 person-years (Linko 2009; Rubenfeld 2005). The mortality rate of ALI and ARDS has decreased over time and is currently reported at 43% with high variability (Zambon 2008).                                                           

                                                                  

Description of the intervention

Nearly all hospitalized patients with ALI and ARDS require mechanical ventilation (MV). Among ALI and ARDS patients receiving MV, the application of a supra-atmospheric pressure at end-expiration is referred to as positive end-expiratory pressure (PEEP) (Imberger 2010). PEEP is an easily implemented intervention that is used primarily to prevent atelectasis and to correct the hypoxaemia caused by alveolar hypoventilation.

Four mechanisms have been proposed to explain the improved pulmonary function and gas exchange achieved with PEEP in MV ALI and ARDS patients. These include the following:

  • An increase in functional residual capacity (FRC).

  • Alveolar recruitment.

  • Redistribution of extravascular lung water.

  • Improved ventilation-perfusion matching (Villar 2005).

The risk-benefit profile of PEEP is unclear because this therapy may produce side effects. It may increase the physiologic dead space (Coffey 1983), decrease cardiac output (Dorinsky 1983), worsen tissue perfusion (Jedlinska 2000), promote bacterial translocation (Lachmann 2007) and increase the risk of barotrauma (Eisner 2002).

How the intervention might work

MV may induce lung injury in patients. This is usually referred to as ventilator-induced lung injury (VILI). Two mechanical factors may contribute to the development of VILI: volutrauma, generated by an overdistention of the aerated lung regions; and atelectrauma, that is, large shear forces produced by repetitive alveolar recruitment and derecruitment (collapse) (Mead 1970). The use of low tidal volumes and an optimal level of PEEP are important in preventing VILI. Two randomized clinical trials that used small ventilatory volumes and low plateau pressures have demonstrated reduced mortality (Amato 1998; ARDSnet 2000).

The beneficial effects of PEEP on ARDS include alveolar recruitment, which avoids cyclic airway collapse and reopening, protects lung surfactant and improves ventilation homogeneity, thus reducing VILI. Gattinoni et al demonstrated that in participants with ARDS, sequential levels of PEEP measured by computed tomographic section prevented cyclic airway collapse (Gattinoni 1993). Richard et al. found that in participants with ALI, the combination of small tidal volume ventilation and high PEEP could induce alveolar recruitment and improve oxygenation (Richard 2003). Similarly, Ranieri et al ventilated participants with low tidal volume and high PEEP and showed that this strategy reduced bronchoalveolar lavage and systemic levels of cytokines when compared with ventilation with high tidal volume and low PEEP (Ranieri 1999). Borges et al. showed that a recruitment manoeuvre with PEEP with subsequent maintenance of high levels of PEEP reversed the collapse of alveoli and improved oxygenation (Borges 2006).

Why it is important to do this review

Evidence indicates that high levels of PEEP reduce VILI in patients with ALI and ARDS. It is important to conduct this systematic review because available studies testing the effects of higher levels of PEEP were not powered sufficiently to show differences in mortality (Brower 2004; Meade 2008; Mercat 2008), and the optimal level of PEEP in patients with ALI and ARDS is still controversial. Other published Cochrane reviews have focused on this topic (Imberger 2010; Petrucci 2007), but no review has compared high versus low levels of PEEP.  

A full list of terminologies used in this review will be found in Appendix 1.

Objectives

We assessed the benefits and harms of high levels of PEEP compared with low levels of PEEP in adults with ALI and ARDS. We further aimed to investigate the primary outcome of mortality before hospital discharge, as well as the secondary outcomes of oxygen efficiency (PaO2/FIO2), barotrauma, ventilator-free days (VFDs) and length of stay (LOS) in the intensive care unit (ICU). We also conducted subgroup and sensitivity analyses.

Methods

Criteria for considering studies for this review

Types of studies

We included randomized controlled trials (RCTs) that compared the effects of two levels of PEEP in participants with ALI and ARDS who were intubated and mechanically ventilated in intensive care for at least 24 hours.

We included studies irrespective of language and publication status.

We excluded studies that used noninvasive ventilation (NIV) as an intervention.

We excluded studies that used zero PEEP as an intervention for participants with ALI and ARDS. 

We excluded prospective cohort studies, cross-over studies and quasi-randomized studies.

Types of participants

We included adults (16 years of age or older) with ALI and ARDS who were intubated and received MV using PEEP for at least 24 hours.

Types of interventions

We compared two levels of PEEP in participants with ALI and ARDS receiving MV and PEEP with or without other interventions.

Participants with higher levels of PEEP constituted the intervention group.

Types of outcome measures

Primary outcomes
  • We included all trials reporting mortality before hospital discharge. If that information was absent, we considered mortality within 28 days of randomization or mortality in the ICU.

Secondary outcomes
  • Oxygen efficiency (PaO2/FIO2): first, third, and seventh days.

  • Barotrauma: defined as the presence of pneumothorax on chest radiograph or chest tube insertions for known or suspected spontaneous pneumothorax.

  • VFDs: defined as the number of days between successful weaning from mechanical ventilation and day 28 after study enrolment (Schoenfeld 2002).

  • LOS in ICU.

Search methods for identification of studies

We used the optimally sensitive search strategy developed for The Cochrane Collaboration to identify all relevant published and unpublished RCTs (Higgins 2011). We did not impose language restrictions.

Electronic searches

We searched the following databases: The Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library, Issue 4, 2013) (see Appendix 2), MEDLINE via Ovid SP (from 1950 to May 2013) (see Appendix 3), EMBASE via Ovid SP (from 1982 to May 2013) (see Appendix 4) and LILACS via the BIREME interface (from 1982 to May 2013) (see Appendix 5).

Searching other resources

We screened the reference lists of all available review articles and primary studies. We used the Science Citation Index to find references that have cited the identified trials. We contacted investigators to identify additional published and unpublished studies. We did not specifically conduct manual searches of abstracts of conference proceedings for this review.

We searched for ongoing trials at the following Websites:

Data collection and analysis

Selection of studies

Two review authors (RSC and RN) independently screened all studies for eligibility on the basis of their titles and abstracts. We documented the reasons for excluding any studies. We resolved disagreements by consulting with a third review author (RH), who decided whether a trial would be included in our review. When the published information was insufficient, RSC contacted the first author of the relevant trial before making a decision about inclusion of the study.

We inserted a PRISMA flow chart in our review to reflect this process (Liberati 2009; Moher 2009) (Figure 1).

Figure 1.

Flow diagram of the selection of trials included in the meta-analysis.

Data extraction and management

Two review authors independently (RSC and RN) extracted and collected data from the included studies on a standardized form. We resolved any discrepancies in the data by discussion. We extracted data on study design, inclusion and exclusion criteria, participant characteristics, intervention characteristics, outcomes and complications associated with intervention. One review author (RSC) entered data into Review Manager. When additional information was needed, we contacted the first author of the relevant trial. 

Assessment of risk of bias in included studies

Two review authors (RSC and RN) independently judged trial quality according to the criteria outlined in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011) and evaluated several types of biases:

  • Selection bias through evaluation of the randomization procedure and allocation concealment.

  • Performance bias through evaluation of the blinding of participants and individuals administering the treatment.

  • Attrition bias through evaluation of the number of participants withdrawn from the studies, reported for each group and through analysis by intention-to-treat (ITT).

  • Detection bias through evaluation of the blinding of outcome assessment.

  • Reporting bias through evaluation of the differences between reported and unreported findings.

  • Any other sources of bias present in relevant studies.

We (RSC and RN) assessed the risk of bias. We resolved any disagreements through consultation with a third review author (RH).

We displayed the results by creating a 'Risk of bias' summary (Figure 2) and a 'Risk of bias' graph (Figure 3) using RevMan 5.1 software. We presented the risk of bias in the 'Results' section. We provided a summary assessment of the risk of bias for each outcome within and across studies.

Figure 2.

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Figure 3.

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Measures of treatment effect

We presented dichotomous data as risk ratios (RRs) for relative measures and risk differences (RDs) for absolute measures. We reported continuous data as mean differences (MDs). The goals were to obtain numerical estimates of these summary statistics from each trial and then to perform a stratified analysis to combine the results.

Unit of analysis issues

We did not include studies with nonstandard designs, such as cluster-randomized trials, and studies with multiple treatment groups.

Dealing with missing data

We contacted the first authors and investigators of the studies to inquire about missing data essential for the analysis of outcomes.

Assessment of heterogeneity

We assessed statistical heterogeneity with the I2 statistic, thereby estimating the percentage of total variance across studies that was attributable to heterogeneity rather than to chance (Higgins 2003). We considered a value of greater than 50% as definitely considerable if it was also significant. We searched for possible sources of heterogeneity including differences among participants (e.g. participants with ALI or ARDS, APACHE II (Acute Physiology and Chronic Health Evaluation II) adjusted risk of death, age, lung injury score, sepsis, number of organ failures) and the interventions used (different levels of PEEP with or without other interventions).

Clinical heterogeneity

We used clinical heterogeneity to describe differences among participants, interventions and outcomes that might have an impact on the effects of high levels of PEEP (as the study of Hodgson 2009).

Assessment of reporting biases

We examined funnel plots (a graphical display) of the size of the treatment effect for the primary outcome against trial precision (1/standard error) for asymmetry, to evaluate whether evidence suggested a publication bias.

Data synthesis

In the absence of significant heterogeneity (I2< 20%), we used the fixed-effect model. At moderate levels of heterogeneity, we applied a random-effects model (I2 statistic 20% to 50%). We interpreted an I2 > 50% as indicating high levels of heterogeneity and then investigated its causes as follows:

  • We undertook subgroup analyses, whenever possible, considering the described potential source of heterogeneity. When heterogeneity persisted, we presented the results separately and reported the reasons for heterogeneity.

  • We investigated diversity in clinical and methodological aspects of the included trials.

  • We performed sensitivity analyses to address the methodological quality of the trials, including and excluding trials with moderate and high risk of bias, respectively.

We used The Cochrane Collaboration's software Review Manager 5.1 (RevMan 5.1) for data organization and analysis and entered data so that the area to the left of the line of no effect indicated a favourable outcome for high positive-pressure PEEP.

Subgroup analysis and investigation of heterogeneity

We performed subgroup analyses for the following categories.

Participants
  • Participants with ALI.

  • Participants with ARDS.

  • APACHE II–adjusted risk of death, age, lung injury score, sepsis, number of organ failures.

Intervention
Different ways of applying PEEP:
  • PEEP according to the mechanical characteristics of the lung.

  • PEEP according to FIO2 and PaO2.

PEEP along with other interventions:
  • High PEEP and low tidal volume versus low PEEP and high tidal volume.

     

Sensitivity analysis

We performed sensitivity analyses. We excluded trials with a high risk of bias. We compared random-effects and fixed-effect model estimates for each outcome variable. We excluded any study that appears to have a large effect size to assess its impact on the meta-analysis. We also performed sensitivity analyses because of large variations in the event rate of the control group and excluded studies with the widest variations.

Summary of findings table                    

We used the principles of the GRADE system (Guyatt 2008) to assess the quality of the body of evidence associated with specific outcomes in our review and used the GRADE software to construct a 'Summary of findings' (SoF) table. The GRADE approach appraises the quality of a body of evidence within a study by considering the risk of bias (methodological quality), the directness of the evidence, the heterogeneity of the data, precision effect estimates, and the risk of publication bias.

Specific outcomes included the following:

  • Mortality.

  • Oxygen efficiency, PaO2/FIO2 above baseline levels during the first, third, and seventh days of treatment.

  • Barotrauma.

  • VFDs.

  • LOS in ICU.

Results

Description of studies

See: Characteristics of included studies; Characteristics of excluded studies; Characteristics of ongoing studies.

Results of the search

The electronic search resulted in 3601 studies. We excluded 3582 studies, which were clearly irrelevant or duplicates. We retrieved 19 studies for further assessment. From these 19 studies, we excluded a further 12 trials. We included only seven studies in the final analysis (Figure 1).

Included studies

We analysed the seven included studies, consisting of 2565 participants, all with ALI and ARDS. One study used lung injury severity scores (LISS) in the definition of ALI and ARDS (Amato 1998), five studies (Brower 2004; Huh 2009; Meade 2008; Mercat 2008; Talmor 2008) employed the American–European Consensus Conference (AECC) definition and the remaining study (Villar 2006) examined participants with ARDS (likewise as defined by the AECC) who had been ventilated at the standard setting for 24 hours and persisted at a PaO2/FIO2 of ≤ 200.

The number of participants in each study ranged from 53 (Amato 1998) to 983 (Meade 2008).

The average age at randomization was 52 years. One study (Amato 1998) included younger participants (average age 34.5 years).

One study (Meade 2008) exhibited certain differences in baseline characteristics: Participants in the control group were 2.4 years older than those in the experimental group, and their rate of sepsis at baseline was 3.7% higher. We wanted to know whether those differences were statistically significant and accordingly asked the author. Dr Meade replied that the associated P value (with a Bonferroni correction) for age was 0.03 and for sepsis was 0.24, but that these differences were minimal after the data were pooled. Furthermore, upon conducting a study in which they pooled all of the data derived from a meta-analysis of the data from individual participants, investigators found no differences between high- and low-PEEP groups with respect to age or sepsis (Briel 2010).

In one study (Talmor 2008), the data on allocation of interventions to participants, random sequence generation, and measurements of ventilator function during the first seven days of treatment were not published. The author, when contacted, answered that investigators had used a block-randomization scheme with blocks of eight. These blocks were kept in sealed envelopes that had been prepared before the study was conducted. Also, data were available to investigators for only the first 72 hours of treatment, at which point participants were turned over to their team for the usual care.

One study (Huh 2009) provided no data on allocation of interventions to participants, on exclusion criteria, and on PEEP levels and oxygenation values during the first week of treatment. We contacted the author, who sent us those missing data.

In five studies, participants were randomly assigned to receive high or low levels of PEEP with the same tidal volume in both groups (Brower 2004; Huh 2009; Meade 2008; Mercat 2008, Talmor 2008), and participants in the remaining two studies (Amato 1998; Villar 2006) received either high or low levels of PEEP with a different tidal volume in each group. Four studies included recruitment manoeuvres in the intervention group (Amato 1998; Huh 2009; Meade 2008; Talmor 2008), whereas one study included recruitment manoeuvres for only the first 80 participants (Brower 2004).

Primary and secondary outcomes varied among the included studies (Table 1). Mortality before hospital discharge was measured in five studies (Amato 1998; Brower 2004; Meade 2008; Mercat 2008; Villar 2006), mortality within 28 days in five publications (Amato 1998; Huh 2009; Meade 2008; Mercat 2008; Talmor 2008), and mortality in the ICU in four (Amato 1998; Huh 2009; Meade 2008; Villar 2006).

Table 1. Different primary and secondary outcomes  
Amato 19981. Mortality at day 28.
  • Mortality before hospital discharge.

  • Barotrauma.

  • Weaning rate adjusted for APACHE II score.

  • Mortality in the intensive care unit (ICU).

  • Death after weaning of MV.

  • Nosocomial pneumonia.

  • Use of paralysing agents > 24 hours.

  • Neuropathy after extubation.

  • Dialysis required.

  • Packed red cells infused.

Brower 20041. Mortality before hospital discharge.
  • Ventilator-free days.

  • Days not spent in ICU.

  • Days free without organ failure.

  • Barotrauma.

  • Breathing without assistance by day 28.

Huh 20091. Improvement in oxygenation.
  • Respiratory mechanics (PEEP and dynamic compliance).

  • ICU stay.

  • Duration of sedatives and paralysing agents.

  • Mortality at day 28.

  • Mortality at day 60.

  • Duration of MV.

Meade 20081. Mortality all-cause hospital (mortality before hospital discharge).
  • Mortality during MV.

  • Mortality in ICU.

  • Mortality at day 28.

  • Barotrauma.

  • Refractory hypoxaemia.

  • Refractory acidosis.

  • Refractory barotrauma.

  • Use of rescue therapies in response to refractory hypoxaemia, refractory acidosis or refractory barotrauma.

  • Days of MV.

  • Days of ICU.

  • Days of hospitalization.

Mercat 20081. Mortality at day 28.
  • Mortality at day 60.

  • Mortality before hospital discharge censored on day 60.

  • Ventilator-free days.

  • Days free without organ failure.

  • Barotrauma between day 1 and day 28.

Talmor 20081. Improvement in oxygenation.
  • Indexes of lung mechanics and gas exchange (respiratory system compliance and ratio of physiological dead space to tidal volume).

  • Ventilator-free days.

  • ICU stay.

  • Days not spent in ICU.

  • Mortality at day 28.

  • Mortality at day 180.

  • Days of ventilation among survivors.

Villar 20061. Mortality in the intensive care unit (ICU).
  • Hospital mortality (mortality before hospital discharge).

  • Ventilator-free days.

  • Extrapulmonary organ failure.

  • Barotrauma.

Five studies observed changes in oxygenation (PaO2/FIO2) on the first and third days (Brower 2004; Huh 2009; Meade 2008; Mercat 2008; Villar 2006); four studies observed changes in oxygenation (PaO2/FIO2) on the seventh day (Brower 2004; Huh 2009; Meade 2008; Mercat 2008); six reported barotrauma (Amato 1998; Brower 2004; Huh 2009; Meade 2008; Mercat 2008; Villar 2006); four indicated the number of VFDs (Brower 2004; Mercat 2008; Talmor 2008; Villar 2006); and two estimated LOS in the ICU (Huh 2009; Meade 2008).  

Three studies were stopped prematurely because of a significant difference in survival between groups (Amato 1998; Mercat 2008; Villar 2006); one study was discontinued on the basis of the specified futility stopping rule (Brower 2004).

Excluded studies

We excluded 12 studies (see Characteristics of excluded studies). Six were not RCTs (Badet 2009; Burns 2001; Dellamonica 2011; Grasso 2007; Richard 2003; Toth 2007); one was a review (Kallet 2007); in three the outcomes were physiologic (Carvalho 1997; Grasso 2005; Ranieri 1999); one was the first part of a larger investigation (Amato 1995) and one did not meet our inclusion criteria because both participant groups were administered equal levels of PEEP (Oczenski 2004).

Two ongoing studies (Kacmarek 2007; Pintado 2003) were considered relevant to this review (see Characteristics of ongoing studies).

Risk of bias in included studies

The seven studies included in this review were of moderate to good quality (see Characteristics of included studies). The summary of quality assessment can be found in Figure 2 and Figure 3.

Allocation

In relation to the sequence generation process, one study had insufficient information (Amato 1998), five studies used blocked randomization to form the allocation for the two comparison groups (Brower 2004; Meade 2008; Mercat 2008; Talmor 2008; Villar 2006) and the last study used a random number table in the sequence generation process (Huh 2009).

In relation to concealment of the allocation sequence, in three of the studies a centralized interactive voice system was used to assign eligible participants randomly (Brower 2004; Meade 2008; Mercat 2008).

In three studies, randomization was performed through the use of sealed envelopes (Amato 1998; Villar 2006; Talmor 2008), and the last study provided insufficient information about concealment of the allocation sequence (Huh 2009).

Blinding

In relation to blinding of participants and personnel, because of the nature of the intervention, the investigators could not be blinded, but participants were unaware of their group allocation because they were critically ill and were under deep sedation. Likewise we believe that the risk of bias was low because the primary outcome is objective and all studies had a strict protocol for both treatment groups.

In relation to blinding of outcome assessment, as with blinding of participants and personnel, because of the characteristics of the primary outcome, we believe that the risk of bias was low. Likewise, in two studies the data analysis was conducted in a blinded fashion (Meade 2008; Mercat 2008).

Incomplete outcome data

Four studies performed the analysis on the basis of the ITT principle (Amato 1998; Meade 2008; Mercat 2008; Talmor 2008); one study (Amato 1998) was hampered by minor protocol violations in both groups; one study showed that seven participants who were withdrawn from the study contributed partial data for the secondary analysis (Meade 2008); one study indicated that one of the participants in the experimental group was lost on day 29 of follow-up after discharge (Mercat 2008); and the last study reported that measurements were not performed on one participant in the experimental group because the participant could not be sedated (Talmor 2008).

Three studies had incomplete outcome data (Huh 2009; Mercat 2008; Villar 2006). Mercat 2008 excluded only one participant because the family withdrew consent after randomization. Villar 2006 indicated that eight participants were lost (three in the intervention group and five in the control group) because one of the centres failed to adhere to the randomization methodology. Although no differences in outcomes were reported, these eight participants were not included in the final analysis. We believe that  in these two studies (Mercat 2008; Villar 2006), the risk of bias was low because the reasons for exclusion were reported and were balanced across groups. In the remaining study (Huh 2009), the reasons for excluding participants were not reported; therefore we believe that the risk of bias was not clear.

Selective reporting

A reporting bias occurred in one study (Brower 2004) because the primary outcomes were proposed in the protocol but were performed in the study differently, and some secondary outcomes proposed in the protocol were not assessed in the study.

Other potential sources of bias

Two studies (Brower 2004; Meade 2008) reported differences in baseline characteristics between the two groups, but these differences did not change the main results.

 

Effects of interventions

See: Summary of findings for the main comparison High versus low levels of PEEP for mechanically ventilated adult patients with acute lung injury and acute respiratory distress syndrome

We collected data comparing the effects of high versus low levels of PEEP; five studies did so with the same tidal volume in both groups, and two studies examined high levels of PEEP with a low tidal volume versus low levels of PEEP with a higher tidal volume.

For the main analysis, we assessed mortality before hospital discharge, including those studies that compared high versus low levels of PEEP with no other interventions. We pooled three studies (Brower 2004; Meade 2008; Mercat 2008), used the fixed-effect model because the I2 statistic was 0%, and found no statistically significant differences between the two groups (relative risk (RR) 0.90, 95% confidence interval (CI) 0.81 to 1.01) (Analysis 1.1; Figure 4).

Figure 4.

Forest plot of comparison: 1 High versus low levels of PEEP, outcome: 1.1 Mortality before hospital discharge.

Five studies assessed oxygen efficiency by means of the PaO2/FIO2 ratio on the first and third days (Brower 2004; Huh 2009; Meade 2008; Mercat 2008; Villar 2006). An improvement in oxygenation occurred, but with heterogeneity among the included studies (Analysis 1.2; Analysis 1.3). In the analysis that assessed oxygen efficacy in this way, the random-effects model was used on the first day because the I2 statistic was 86%, and a statistically significant difference between the two groups was found (mean difference 41.31, 95% CI 24.11 to 58.52; Analysis 1.2). On the third day, the random-effects model was used because the I2 statistic was 82%, and a statistically significant difference between the two groups was evident (mean difference 42.51, 95% CI 25 to 60.02; Analysis 1.3). In the assessment of oxygen efficiency by means of the PaO2/FIO2 ratio on the seventh day, only four studies were included (Brower 2004; Huh 2009; Meade 2008; Mercat 2008) because Villar 2006 had no data on the seventh day. In this analysis, the random-effects model was used (with an I2 statistic of 85%), and a statistically significant difference between the two groups was not found (mean difference 24.77, 95% CI -0.92 to 50.45; Analysis 1.4).

Among all possible sources of heterogeneity included in the protocol, we could undertake a subgroup analysis only for the participants with ARDS. Two studies assessed oxygen efficiency, as measured by the PaO2/FIO2 ratio, on the first and third days with these ARDS participants (Huh 2009; Villar 2006). In that analysis, on the first day the fixed-effect model was used because the I2 statistic was 0%, and we saw evidence of a benefit (mean difference 17.79, 95% CI 1.37 to 34.21; Analysis 1.5). In that same analysis on the third day, the fixed-effect model was used as well (with the I2 statistic likewise being 0%), and we saw evidence of a benefit (mean difference 35.30, 95% CI 14.71 to 55.90; Analysis 1.6). On the seventh day (as in Analysis 1.4), we could not perform an analysis, as Villar 2006 provided no data for the seventh day.

In six studies assessing barotrauma (Amato 1998; Brower 2004; Huh 2009; Meade 2008; Mercat 2008; Villar 2006), the random-effects model was used because the I2 statistic was 40%, and a statistically significant difference between the two groups was not found (RR 0.97, 95% CI 0.66 to 1.42) (Analysis 1.7).

We assessed the number of VFDs in four studies (Brower 2004; Mercat 2008; Talmor 2008; Villar 2006). In two of those four studies, the data were expressed as medians (Mercat 2008; Talmor 2008); therefore, we treated mean and median data separately (Analysis 1.8). When we accordingly excluded both of these studies from the analysis and analysed the two studies expressing data as mean values (Brower 2004; Villar 2006), we found no differences in the meta-analyses (RR 1.89, 95% CI -3.58 to 7.36), and we noted heterogeneity among the included studies (I² = 87%) (Analysis 1.9).

We assessed LOS in the ICU in three studies but did not pool the data for analysis for this outcome because the data were expressed differently in terms of mean and median among the three: We thus included only the data and their statistical values with those studies (Analysis 1.10). Huh 2009 used mean values and found no difference between the two groups (P = 0.6); Meade 2008 and Talmor 2008 expressed the data for this parameter as medians and likewise found no significant difference. The P value was 0.98 for the study of Meade 2008 and 0.16 for the study of Talmor 2008.

We also assessed the mortality occurring before hospital discharge, including those studies that compared high versus low levels of PEEP with or without other interventions (Amato 1998; Brower 2004; Meade 2008; Mercat 2008; Villar 2006); we used the fixed-effect model because the I2 statistic was 9% and found a statistically significant difference between the two groups (RR 0.88, 95% CI 0.79 to 0.98) in the analysis (Analysis 1.11).

We pooled studies assessing mortality within 28 days of randomization. Five studies were included (Amato 1998; Huh 2009; Meade 2008; Mercat 2008; Talmor 2008); the random-effects model was used because the I2 statistic was 37% and a statistically significant difference between the two groups was not found (RR 0.83, 95% CI 0.67 to 1.01; Analysis 1.12).

Subgroup analysis

We conducted subgroup analyses assessing hospital mortality. Included studies provided insufficient data for subgroup analyses evaluating the effects of sepsis, organ failure, the lung injury score, or the APACHE II-adjusted risk of death.

In subgroup analyses based on age and comprising older participants (Brower 2004; Meade 2008; Mercat 2008; Villar 2006), the fixed-effect model was used because the I2 statistic was 0%, and the meta-analysis indicated a statistically significant difference between the two groups (RR 0.89, 95% CI 0.80 to 0.99) (Analysis 1.13).

In subgroup analyses based on the use of PEEP according to mechanical characteristics of the lung (Amato 1998; Mercat 2008; Villar 2006), the random-effects model was used because the I2 statistic was 47% and a statistically significant difference between the two groups was not found (RR 0.76, 95% CI 0.57 to 1.01) (Analysis 1.14).

In subgroup analyses based on the use of PEEP according to FIO2 and PaO(Brower 2004; Meade 2008), the fixed-effect model was used because the I2 statistic was 0% and a statistically significant difference between the two groups was not found (RR 0.90, 95% CI 0.79 to 1.04) (Analysis 1.15).

In subgroup analyses based on the use of high PEEP and low tidal volume versus low PEEP and high tidal volume (Amato 1998; Villar 2006), the fixed-effect model was used because the I2 statistic was 0% and the meta-analysis indicated a statistically significant difference between the two groups (RR 0.62, 95% CI 0.44 to 0.87) (Analysis 1.16).

We conducted subgroup analyses to assess mortality in the ICU in participants with ARDS. In the three studies included (Amato 1998; Huh 2009; Villar 2006), the random-effects model was used because the I2 statistic was 24% and the meta-analysis indicated a statistically significant difference between the two groups (RR 0.67, 95% CI 0.48 to 0.95) (Analysis 1.17).

Sensitivity analysis

We evaluated mortality before hospital discharge in the studies of good quality (Meade 2008; Mercat 2008; Villar 2006). The random-effects model was used because the I2 statistic was 21% and a statistically significant difference between the two groups was not found (RR 087, 95% CI 0.76 to 1.01) (Analysis 1.18).

We excluded the study with large effect size (Meade 2008) and used the random-effects model because the I2 statistic was 26% and the meta-analysis indicated a statistically significant difference between the two groups (RR 0.83, 95% CI 0.69 to 0.99) (Analysis 1.19).

We did not perform sensitivity analysis to exclude studies with large variations in the control group event rate because the studies to be included are those of Analysis 1.1.

Publication bias

This assessment could not be estimated because of the small number of studies included in this review.

Discussion

Summary of main results

In this meta-analysis, seven studies met the criteria for inclusion.

Of the studies selected, five assessed mortality before hospital discharge (Amato 1998; Brower 2004; Meade 2008; Mercat 2008; Villar 2006) and were of moderate to good quality.

For the main analysis, we decided to exclude studies that applied different tidal volumes between intervention and control arms (Amato 1998; Villar 2006) because lack of clarity as to whether positive results were attributable to a reduction in tidal volume, to higher levels of PEEP, or to both tactics together made conclusions difficult.

The remaining three studies (Brower 2004; Meade 2008; Mercat 2008) assessed mortality before hospital discharge; thus our primary aim was to determine whether use of higher levels of PEEP improved clinical outcome.

High levels of PEEP compared with low levels showed no statistically significant decrease in mortality before hospital discharge in participants with ALI and ARDS (Analysis 1.1; Figure 4). The forest plot, however, showed a trend towards a mortality benefit in the higher-PEEP group. Furthermore, the existence of clinical heterogeneity should be taken into account in considering that outcome because two of the studies (Brower 2004; Mercat 2008) included participants with a PaO2/FIO2 ≤ 300, and Meade 2008 reported on participants with a PaO2/FIO2 ≤ 250.

This outcome is similar to that obtained by Dasenbrook 2011, who found that higher levels of PEEP were not associated with significantly different 28-day mortality. That meta-analysis included four studies; three were the same as in our analysis (i.e. Brower 2004; Meade 2008; Mercat 2008), and in Talmor 2008, a ventilator strategy was used to adjust high levels of PEEP according to the oesophageal pressures of participants with ALI and ARDS. In this review, according to our analysis, clinical heterogeneity is evidenced among the included studies.  

The benefit of high levels of PEEP is assessed through a reduction of the collapsed lung and is dependent on the extent of lung recruitment. Gattinoni 2006 demonstrated that use of high levels of PEEP was more beneficial in participants with lower PaO2/FIO2 ratios (ARDS rather than ALI participants) and a high percentage of potentially recruitable lung volume. Furthermore, in a later paper, Caironi 2010 analysed data from that study and provided evidence in favour of the application of high levels of PEEP, especially in participants with great lung recruitability. In addition, Hess 2011 showed that modest levels of PEEP may be more appropriate for participants with ALI (at PaO2/FIO2 ratios ≤ 300), whereas higher levels of PEEP should be used for participants with ARDS (at PaO2/FIO2 ratios ≤ 200). Therefore, the strategy of using high levels of PEEP regardless of the type of patient (i.e. those with ALI or ARDS) could be incorrect.

Of the studies involving participants with ARDS that we analysed, three had assessed mortality in the ICU (Amato 1998; Huh 2009; Villar 2006) and indicated significant differences between groups (RR 0.67, 95% CI, 0.48 to 0.95; Analysis 1.17). In assessing this result, however, we need to consider that in two of those studies (Amato 1998; Villar 2006), high levels of PEEP were part of a ventilatory strategy that also included a low tidal volume. Furthermore, in the assessment of the forest plot, these two studies proved to be of greater benefit. In addition, the meta-analysis of individual-patient data (Briel 2010) indicated that treatment effects varied with the presence or absence of ARDS, as defined by a value of PaO2/FIO2 of 200 or less, and that in participants with ARDS, higher levels of PEEP were associated with improved survival.

We need also to consider clinical heterogeneity because the trials in this analysis used different approaches to determine PEEP levels in the intervention group. Two studies (Brower 2004; Meade 2008) included recruitment manoeuvres (before the setting of PEEP) with high levels of PEEP preset according to PaO2 and FIO2 values. In the remaining study (Mercat 2008), the PEEP level was preset to achieve plateau pressures between 28 and 30 cm H2O. Despite the use of different criteria for PEEP selection, the optimal method still remains unclear. In this review, subgroup analysis that included different ways of applying PEEP (PEEP on the basis of mechanical characteristics of the lung and PEEP according to FIO2 and PaO2; Analysis 1.14; Analysis 1.15) failed to evidence any significant difference between these two approaches.

With respect to these two issues (heterogeneous population and different ways of applying high levels of PEEP), the results of an ongoing study (Kacmarek 2007) using high levels of PEEP only in participants with ARDS, and in an individualized way, should prove to be both relevant and informative.

In the subgroup analysis, we pooled studies that assessed lung-protective ventilation (low tidal volumes and high levels of PEEP) versus conventional ventilation (Amato 1998; Villar 2006; Analysis 1.16) and found a decrease in mortality with the use of lung-protective ventilation (RR 0.62; 95% CI 0.44 to 0.87). For this reason, we conclude—and we must emphasize—that lung-protective ventilation was shown to be beneficial in participants with ALI and ARDS.

In participants with ALI and ARDS, the use of PEEP has been recognized as producing an improvement in oxygenation (Borges 2006; Ranieri 1991; Suter 1975). In this review we accordingly observed that high levels of PEEP did indeed improve oxygenation in participants up to the first, third, and seventh days of mechanical ventilation, but with heterogeneity occurring in the different analyses surveyed (Analysis 1.2; Analysis 1.3; Analysis 1.4). For this reason we undertook subgroup analysis of oxygen efficacy through PaO2/FIOon the first and third days in participants with ARDS, and we saw an improvement in oxygenation (Analysis 1.5; Analysis 1.6), consistent with the findings of Gattinoni (Gattinoni 2006) and Grasso (Grasso 2005). Gattinoni showed that in participants with ARDS, the potential for recruitment may be greater, and the earlier physiological study of Grasso applied both lower (9 ± 2 cm H2O) and higher (16 ± 1 cm H2O) PEEP levels and found that participants with a high potential for recruitment had a greater increase in oxygenation (from PaO2/FIO2 of 150 ± 36 to PaO2/FIO2 396 ± 138). By contrast, in the four studies assessing oxygenation on the seventh day (Brower 2004; Huh 2009; Meade 2008; Mercat 2008), PEEP levels did not differ markedly between groups. Furthermore, in participants with late ARDS, improvements in oxygenation and in respiratory mechanics were less well defined, and the potential for alveolar recruitment could have been reduced (Grasso 2002). These observations could justify the marked heterogeneity that can occur in the analysis of oxygen efficiency through PaO2/FIO2 after seven days.

Barotrauma is a complication in mechanically ventilated patients. Two studies have examined the relationship between barotrauma and high levels of PEEP in participants with ARDS (Weg 1998; Eisner 2002). Weg 1998 performed an analysis of data from a prospective trial of aerosolized synthetic surfactant in participants with ARDS and found no relationship between barotrauma and a mean level of PEEP of 12 cm H2O. Eisner 2002 reported that high levels of PEEP may increase the likelihood of early barotrauma, even after control is applied for markers of acute and chronic disease severity. In the present review, six studies assessed barotrauma (Amato 1998; Brower 2004; Huh 2009; Meade 2008; Mercat 2008; Villar 2006); in accordance with Weg 1998, no statistically significant difference was noted between the groups.

In conjunction with statistically significant heterogeneity, high levels of PEEP produced no significant difference between the two groups in terms of the number of VFDs. We included two studies (Brower 2004; Villar 2006) that differed in two ways: First, in terms of population, Brower 2004 used participants with PaO2/FIO2 ≤ 300, and Villar 2006 reported only participants with ARDS ventilated at a standard setting for 24 hours and persisting with a PaO2/FIO2 ≤ 200. Second, with respect to the mode of applying high levels of PEEP, in Brower 2004, PEEP levels were preset according to PaO2 and FIO2 values, whereas in Villar 2006, those levels were preset at 2 cm H2O above the Pflex (upward shift in the slope of the pressure-volume curve). These differences could explain the statistical heterogeneity present in the analysis of VFDs.

Data on LOS in the ICU were insufficient for an examination to be performed.

Summary of findings table

In the main analysis (mortality before hospital discharge), we do not degrade for publication bias owing to the smaller number of included studies. We downgraded the secondary outcomes of oxygen efficiency on the first, third, and seventh days,  and ventilator-free days were likewise downgraded because of an inconsistency resulting from minimal overlap among the studies. Here the P value for heterogeneity was less than 0.05, and the I2 was large. We did not include LOS in the ICU in the summary of findings table because we were not able to pool the data for analysis of this outcome (see Summary of findings for the main comparison).

Overall completeness and applicability of evidence

On the whole, high levels of PEEP showed no statistically significant difference in mortality before hospital discharge in participants with ALI and ARDS. This review does not present data about subgroups (ALI or ARDS) that could benefit from high levels of PEEP because clinical heterogeneity is present among the studies included for such intervention. Further studies should help to assess benefit with respect to the different ways of applying PEEP and to the advantages and disadvantages associated with different ALI and ARDS patient populations.

Quality of the evidence

The quality of evidence is good. Four studies (Meade 2008; Mercat 2008; Talmor 2008; Villar 2006) met all criteria considered in the evaluation of risk of bias.

Potential biases in the review process

Risk of bias was noted in three of the included studies (Amato 1998; Brower 2004; Huh 2009). Amato 1998 provided insufficient information about the sequence generation process; Brower 2004 described reporting bias; and Huh 2009 provided insufficient information about concealment of the allocation sequence and about exclusions.

Agreements and disagreements with other studies or reviews

To date, several reviews have examined the use of high levels of PEEP in participants with ALI and ARDS. Evidence from these reviews indicates the trend toward a decrease in mortality with the use of high levels of PEEP, especially in participants with ARDS.

Gordo-Vidal 2007 evaluated four studies published from 1998 to 2006, three of which used a protective ventilatory strategy involving a low tidal volume and a high PEEP. These studies included contained statistical heterogeneity; moreover, the review failed to report a decrease in mortality (RR 0.73, 95% CI 0.49 to 1.10; P = 0.129). Furthermore, the sampling reported in Gordo-Vidal 2007 could not take into account the studies of greatest statistical weight in this present meta-analysis (i.e. Meade 2008, Mercat 2008) because those investigations had not yet been published.

Three reviews (Oba 2009; Phoenix 2009; Yang 2011) included the same studies as ours.

The aim of the study of Oba 2009 was to test the hypothesis that in participants with ALI and ARDS, the use of high levels of PEEP resulted in lower mortality, especially in more severely ill participants. For the analysis, investigators created a hypothetical group in which participants were treated with low VT, instead of conventional tidal volumes, and low PEEP to match tidal volumes between the two groups studied in each trial. In this meta-analysis, a small but significant decrease in mortality with high PEEP may have been shown to exist (RR 0.89; 95% CI 0.80 to 0.99; P = 0.03). As in our review, clinical heterogeneity was present, and investigators found that the effects of high PEEP were greater in participants with higher ICU severity scores.

Phoenix 2009 identified six eligible studies, of which one had evaluated mortality at day 28 (Ranieri 1999). This review furthermore included five studies (Amato 1998; Brower 2004; Meade 2008; Mercat 2008; Villar 2006) that evaluated mortality before hospital discharge. Authors in this review used a random-effects model, included no data about statistical heterogeneity, and showed a reduction in mortality (RR 0.87; 95% CI 0.78 to 0.96; P = 0.007). Those who performed the meta-analysis that was restricted to studies that included PEEP levels as the main variable investigated (e.g. Brower 2004; Meade 2008; Mercat 2008) obtained the same results as we did. Although that study argued that the evidence obtained supported the use of high levels of PEEP in unselected groups of participants with ALI or ARDS in general, or in participants with more severe disease in particular, we are not in agreement with that conclusion because we believe that a determination of which subgroup of participants (ALI or ARDS) would benefit from high levels of PEEP is necessary.

The study of Yang 2011 includes the same six studies as Phoenix 2009 but assesses mortality at day 28 and the rate of barotrauma. In the analysis of mortality at day 28, investigators reported a significant difference between the two groups (odds ratio (OR) 0.81, 95% Cl 0.68 to 0.95; P = 0.01) and furthermore no significant difference in the rate of barotrauma (OR 0.91, 95% Cl 0.55 to 1.51; P = 0.72). We assessed mortality within 28 days of randomization and found no differences between the two groups, although our analysis differs from the review of Yang 2011 in the included studies.

Three reviews (Briel 2010; Dasenbrook 2011; Putensen 2009) as in our main analysis included randomized controlled trials of participants with a diagnosis of ALI or ARDS (as defined by the American-European Consensus Conference) that compared higher versus lower levels of PEEP at the same tidal volume in both groups (control and experimental).

Putensen 2009 identified several studies in participants with ALI and ARDS that attempted to determine whether low tidal volume, high PEEP, or a combination of the two improved outcome. In a meta-analysis that included studies of high versus low PEEP with the same tidal volume, investigators found, as did we in our review, no significant decrease in mortality (OR 0.86, 95% CI 0.72 to 1.02; P = 0.08).

Briel 2010 performed a systematic review and meta-analysis of individual-patient data from three randomized studies that compared higher versus lower PEEP levels in 2299 participants with acute lung injury. Results showed, overall, no statistically significant difference in hospital mortality between groups (RR 0.94, 95% CI 0.86 to 1.04; P = 0.25). However in the subgroup of participants with ARDS, higher levels of PEEP were associated with improved survival (RR 0.90, 95% CI 0.81 to 1.00; P = 0.049). This review included an explicit study protocol and analysis plan, along with access to trial protocols, case report forms and complete, unedited data sets and standardized outcome definitions across trials (except for rescue therapies).

The review of Dasenbrook 2011 included four studies with ventilation strategies that included higher PEEP during volume and pressure limitation in participants with ALI and ARDS. This study reported that higher levels of PEEP were not associated with a significantly different mortality at day 28 (RR 0.90, 95% CI 0.79 to 1.02) or barotrauma.

None of these three reviews (Briel 2010; Dasenbrook 2011; Putensen 2009) demonstrated a decrease in mortality with the use of high PEEP levels. We obtained the same results in the main analysis, which included studies with similar characteristics.

Authors' conclusions

Implications for practice

In summary, the present systematic review provides evidence that the use of high levels of PEEP does not reduce mortality before hospital discharge, although the forest plot indicated a trend towards a mortality benefit in the higher-PEEP group. Moreover, in practice, we cannot recommend the routine use of high levels of PEEP in patients with ALI and ARDS because no agreement has been reached among the various studies reported here as to which group of patients should receive that specific intervention and how those high PEEP levels should be applied.

Implications for research

Further research is needed to evaluate the use of high levels of PEEP. This research should:

  •   Include a randomized trial of only participants with ARDS that involves the use of high compared with low levels of PEEP with the same tidal volume in both groups.

  • · Establish the most appropriate strategy for the application of high levels of PEEP.

Identification of these conditions will help to clarify the appropriate use of high levels of PEEP in participants with ALI and ARDS. 

Acknowledgements

We would like to thank Jane Cracknell (Managing Editor, Cochrane Anaesthesia Review Group (CARG)) for her invaluable help in carrying out the whole review; Karen Hovhannisyan (CARG Trials Search Co-ordinator) for his help in designing the search strategies; and Salvador Benito and Juan Gagliardi for their help and editorial advice during the development of this review.

We would like to thank Stephan Kettner (content editor), Nathan Pace (statistical editor), Arash Afshari (peer reviewer in the protocol), Stephen Frohlich (peer reviewer in the review), Rodrigo Cavallazzi (peer reviewer in the protocol and in the review) and Anne Lyddiatt (consumer representative) for their help and editorial advice during the preparation of this systematic review.

We are grateful to Xiaoli Ge and Guangying Zhuo, who translated two Chinese papers, and to Dr Donald F. Haggerty, a retired career investigator and native English speaker, who edited the final version of the manuscript.

Data and analyses

Download statistical data

Comparison 1. High versus low levels of PEEP
Outcome or subgroup titleNo. of studiesNo. of participantsStatistical methodEffect size
1 Mortality before hospital discharge32299Risk Ratio (M-H, Fixed, 95% CI)0.90 [0.81, 1.01]
2 Oxygen efficiency (PaO2/FIO2). Day 152337Mean Difference (IV, Random, 95% CI)41.31 [24.11, 58.52]
3 Oxygen efficiency (PaO2/FIO2). Day 352059Mean Difference (IV, Random, 95% CI)42.51 [25.00, 60.02]
4 Oxygen efficiency (PaO2/FIO2). Day 741379Mean Difference (IV, Random, 95% CI)24.77 [-0.92, 50.45]
5 Oxygen efficiency (PaO2/FIO2); Day 1. Patients with ARDS2152Mean Difference (IV, Fixed, 95% CI)17.79 [1.37, 34.21]
6 Oxygen efficiency (PaO2/FIO2); Day 3. Patients with ARDS2151Mean Difference (IV, Fixed, 95% CI)35.30 [14.71, 55.90]
7 Barotrauma62504Risk Ratio (M-H, Random, 95% CI)0.97 [0.66, 1.42]
8 Ventilator-free days  Other dataNo numeric data
9 Ventilator-free days (only means)2644Mean Difference (IV, Random, 95% CI)1.89 [-3.58, 7.36]
10 Length of stay in the ICU  Other dataNo numeric data
11 Mortality before hospital discharge (studies with or without other interventions)52447Risk Ratio (M-H, Fixed, 95% CI)0.88 [0.79, 0.98]
12 Mortality within 28 days of randomization51921Risk Ratio (M-H, Random, 95% CI)0.83 [0.67, 1.01]
13 Mortality before hospital discharge: older participants42394Risk Ratio (M-H, Fixed, 95% CI)0.89 [0.80, 0.99]
14 Mortality before hospital discharge. PEEP according to the mechanical characteristics of the lung3915Risk Ratio (M-H, Random, 95% CI)0.76 [0.57, 1.01]
15 Mortality before hospital discharge. PEEP according to FIO2 and PaO221532Risk Ratio (M-H, Fixed, 95% CI)0.90 [0.79, 1.04]
16 Mortality before hospital discharge. High PEEP and low tidal volume versus low PEEP and high tidal volume2148Risk Ratio (M-H, Fixed, 95% CI)0.62 [0.44, 0.87]
17 Mortality in the Intensive Care Unit (ICU). Patients with ARDS3205Risk Ratio (M-H, Random, 95% CI)0.67 [0.48, 0.95]
18 Mortality before hospital discharge. Sensitivity analysis. Studies of good quality31845Risk Ratio (M-H, Random, 95% CI)0.87 [0.76, 1.01]
19 Mortality before hospital discharge.Sensitivity analysis. Exclusion of the study with large effect size41464Risk Ratio (M-H, Random, 95% CI)0.83 [0.69, 0.99]
Analysis 1.1.

Comparison 1 High versus low levels of PEEP, Outcome 1 Mortality before hospital discharge.

Analysis 1.2.

Comparison 1 High versus low levels of PEEP, Outcome 2 Oxygen efficiency (PaO2/FIO2). Day 1.

Analysis 1.3.

Comparison 1 High versus low levels of PEEP, Outcome 3 Oxygen efficiency (PaO2/FIO2). Day 3.

Analysis 1.4.

Comparison 1 High versus low levels of PEEP, Outcome 4 Oxygen efficiency (PaO2/FIO2). Day 7.

Analysis 1.5.

Comparison 1 High versus low levels of PEEP, Outcome 5 Oxygen efficiency (PaO2/FIO2); Day 1. Patients with ARDS.

Analysis 1.6.

Comparison 1 High versus low levels of PEEP, Outcome 6 Oxygen efficiency (PaO2/FIO2); Day 3. Patients with ARDS.

Analysis 1.7.

Comparison 1 High versus low levels of PEEP, Outcome 7 Barotrauma.

Analysis 1.8.

Comparison 1 High versus low levels of PEEP, Outcome 8 Ventilator-free days.

Ventilator-free days
StudyHigh PEEPLow PEEPP Value
Brower 2004 Means: 13.8 Means: 14.50,50
Brower 2004 SD: 10.6 SD: 10.4 
Brower 2004 No. of patients: 276 No. of patients: 273 
Mercat 2008 Median: 7 Median: 30,04
Mercat 2008 Interquartile range: 0.0-19 Interquartile range: 0.0-17 
Mercat 2008 No. of patients: 385 No. of patients: 382 
Talmor 2008 Median: 11.5 Median: 70,50
Talmor 2008 Interquartile range: 0.0-20.3 Interquartile range: 0.0-17 
Talmor 2008 No. of patients: 30 No. of patients: 31 
Villar 2006 Means: 10.9 Means: 60,008
Villar 2006 SD: 9.4 SD: 7.9 
Villar 2006 No. of patients: 50 No. of patients: 45 
Analysis 1.9.

Comparison 1 High versus low levels of PEEP, Outcome 9 Ventilator-free days (only means).

Analysis 1.10.

Comparison 1 High versus low levels of PEEP, Outcome 10 Length of stay in the ICU.

Length of stay in the ICU
StudyHigh PEEPLow PEEPP Value
Huh 2009 Means: 25.1 days. Means: 21.4 days0,643
Huh 2009 SD: 5.6 SD: 5.3 
Huh 2009 No. of patients: 30 No. of patients: 27 
Meade 2008 Median: 13 days. Median: 13 days.0,98
Meade 2008 Interquartile range: 8-23 Interquartile range: 9-23 
Meade 2008 No. of patients: 475 No. of patients: 508 
Talmor 2008 Median: 15,5 days. Median: 13 days.0,16
Talmor 2008 Interquartile range: 10,8-28,5 Interquartile range: 7-22 
Talmor 2008 No. of patients: 30 No. of patients: 508 
Analysis 1.11.

Comparison 1 High versus low levels of PEEP, Outcome 11 Mortality before hospital discharge (studies with or without other interventions).

Analysis 1.12.

Comparison 1 High versus low levels of PEEP, Outcome 12 Mortality within 28 days of randomization.

Analysis 1.13.

Comparison 1 High versus low levels of PEEP, Outcome 13 Mortality before hospital discharge: older participants.

Analysis 1.14.

Comparison 1 High versus low levels of PEEP, Outcome 14 Mortality before hospital discharge. PEEP according to the mechanical characteristics of the lung.

Analysis 1.15.

Comparison 1 High versus low levels of PEEP, Outcome 15 Mortality before hospital discharge. PEEP according to FIO2 and PaO2.

Analysis 1.16.

Comparison 1 High versus low levels of PEEP, Outcome 16 Mortality before hospital discharge. High PEEP and low tidal volume versus low PEEP and high tidal volume.

Analysis 1.17.

Comparison 1 High versus low levels of PEEP, Outcome 17 Mortality in the Intensive Care Unit (ICU). Patients with ARDS.

Analysis 1.18.

Comparison 1 High versus low levels of PEEP, Outcome 18 Mortality before hospital discharge. Sensitivity analysis. Studies of good quality.

Analysis 1.19.

Comparison 1 High versus low levels of PEEP, Outcome 19 Mortality before hospital discharge.Sensitivity analysis. Exclusion of the study with large effect size.

Appendices

Appendix 1. Terminologies  

AECC

American-European Consensus Conference definitions.

ALI criteria: acute onset, PaO2/FIO2 ≤ 300 (regardless of PEEP level), bilateral pulmonary infiltrates and lack of evidence of left heart failure.

ARDS has the same clinical characteristics as ALI, except that the PaO2/FIO2 in ARDS is ≤ 200 (Bernard 1994).

ALIAcute lung injury.
APACHE IIAcute Physiology and Chronic Health Evaluation II. Classification system of severity of diseases.
ARDSAcute respiratory distress syndrome.
Days-free without organ failureNumber of days a participant was without organ failure from day 1 to day 28.
Days not spent in ICUNumber of days a participant was not in the ICU from day 1 to day 28.
Driving PressurePlateau pressure-PEEP.
FIO2Fraction of inspired oxygen.
FRCFunctional residual capacity.
ICUIntensive care unit.
LISSLung injury severity score (Murray 1988). Range 0 (normal) to 4 (most severe). LISS > 2.5 = ARDS.
LOSLength of stay.
MVMechanical ventilation.
PaCO2Partial pressure of arterial carbon dioxide.
PaO2Partial pressure of arterial oxygen.
PaO2/FiO2Relation between partial pressure of arterial/fractional inspired oxygen.
PBWPredicted body weights.
PEEPPositive end-expiratory pressure.
Persistent ARDSParticipants with ARDS ventilated with standard setting for 24 hours who persist with PaO2/FIO2 ≤ 200.
PflexUpward shift in the slope of the pressure-volume curve.
PwedgePulmonary wedge pressure.
Recruitment manoeuvreManoeuvre for opening of collapsed alveoli to improve gas exchange and lung volume end-expiration and decreased VILI.
RCTRandomized controlled trial.
SpO2Arterial oxygen saturation (measured via pulse oximetry).
Static compliance (Cst)Determined by dividing tidal volume by the difference between pressure at the end of inflation hold and PEEP.
StrainRatio between the amount of gas volume delivered during tidal breath and the amount of aerated lung receiving it.
VFDNumber of days between successful weaning from mechanical ventilation and day 28 after study enrolment (Schoenfeld 2002).
VILIVentilator-induced lung injury.
VTVolume tidal.

Appendix 2. Search strategy for CENTRAL, T he Cochrane Library

#1 MeSH descriptor Positive-Pressure Respiration explode all trees
#2 MeSH descriptor Positive-Pressure Respiration, Intrinsic explode all trees
#3 MeSH descriptor Continuous Positive Airway Pressure explode all trees
#4 MeSH descriptor Intermittent Positive-Pressure Breathing explode all trees
#5 MeSH descriptor Intermittent Positive-Pressure Ventilation explode all trees
#6 PEEP
#7 (positive airway or end?expiratory) near pressure
#8 APRV or CPAP or CPAP or PEEP* or auto?PEEP or positive?pressure
#9 (#1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8)
#10 MeSH descriptor Acute Lung Injury explode all trees
#11 MeSH descriptor Respiratory Paralysis explode all trees
#12 MeSH descriptor Respiratory Insufficiency explode all trees
#13 MeSH descriptor Respiratory Distress Syndrome, Adult explode all trees
#14 MeSH descriptor Respiratory Distress Syndrome, Newborn explode all trees
#15 Acute near (hypox* or respiratory)
#16 respirat* near (distress or failure)
#17 lung injury or AHRF or ARDS or ALI
#18 (#10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16 OR #17)
#19 (#8 AND #18)

Appendix 3. Search strategy for MEDLINE (Ovid SP)

1. exp Positive-Pressure-Respiration/ or exp Positive-Pressure-Respiration-Intrinsic/ or exp Continuous-Positive-Airway-Pressure/ or exp Intermittent-Positive-Pressure-Breathing/ or exp Intermittent-Positive-Pressure-Ventilation/
2. (occult or auto or non?therapeutic) adj6 PEEP) or (positive pressure adj3 (respirat* or ventil*) or (positive airway or end?expiratory) adj3 pressure) or (APRV or CPAP or nCPAP or PEEP* or auto?PEEP or positive?pressure).mp.
3. 1 or 2
4. exp Acute Lung Injury/ or exp Respiratory-Paralysis/ or exp Respiratory-Insufficiency/ or exp Respiratory-Distress-Syndrome-Newborn/ or exp Respiratory-Distress-Syndrome-Adult/
5. ((Acute adj3 (hypox* or respiratory) or (respirat* adj3 (distress or failure) or lung injury or (AHRF or ARDS or ALI).mp.
6. 4 or 5
7. 3 and 6
8. (randomized controlled trial or controlled clinical trial).pt. or randomized.ab. or placebo.ab. or clinical trials as topic.sh. or randomly.ab. or trial.ti.) not (animals not (humans and animals)).sh.
9. 7 and 8

Appendix 4. Search strategy for EMBASE (Ovid SP)

1. exp positive end expiratory pressure/ or exp intermittent positive pressure ventilation/
2. (occult or auto or non?therapeutic) adj6 PEEP) or (positive pressure adj3 (respirat* or ventil*) or (positive airway or end?expiratory) adj3 pressure) or (APRV or CPAP or nCPAP or PEEP* or auto?PEEP or positive?pressure).mp.
3. 1 or 2
4. exp acute lung injury/ or exp diaphragm paralysis/ or exp respiratory failure/ or respiratory distress/ or exp respiratory distress syndrome/ or exp adult respiratory distress syndrome/ or exp neonatal respiratory distress syndrome/
5. (Acute adj3 (hypox* or respiratory) or (respirat* adj3 (distress or failure) or lung injury or (AHRF or ARDS or ALI).mp.
6. 4 or 5
7. 3 and 6
8. (singl* or doubl* or tripl*) adj3 blind) or crossover).ti,ab. or multicenter.ab. or placebo.sh. or controlled study.ab. or random*.ti,ab. or trial*.ti,ab.) not (animal not (human and animal).sh.
9. 7 and 8

Appendix 5. Search strategy for LILACS (BIREME interface)

("POSITIVE-PRESSURE" or "AIRWAYPRESSURE" or "PEEP" or "positive pressure" or "presión positiva" or "pressão positiva" or "via aérea positiva" or "ruta aérea positiva" or "APRV" or "CPAP" or "nCPAP") and ("RESPIRATORYINSUFFIENCY" or "RESPIRATORYDISTRESS" or "Lung Injury" or "Acute hypox$" or "respirat$ distress" or "AHRF" or "ARDS" or "ALI" or "Herida pulmonar" or "Lesão de pulmão" or "Aflição respiratória" or "Pena respiratoria" or "Insuficiencia respiratoria" or "Insufficiency respiratório")

Contributions of authors

Conceiving the review: Roberto Santa Cruz (RSC).

Designing the review: RSC, Juan Rojas (JR), Agustin Ciapponi (AC).

Co-ordinating the review: RSC.

Undertaking manual searches: RSC.

Screening search results: RSC, Rolando Nervi (RN), Roberto Heredia (RH).

Organizing retrieval of articles: RSC.

Screening retrieved articles against inclusion criteria: RSC, RN, RH.

Appraising quality of articles: RSC.

Abstracting data from articles: RSC.

Writing to authors of articles for additional information: RSC.

Providing additional data about articles: RSC.

Obtaining and screening data on unpublished studies: RSC.

Providing data management for the review: RSC, JR, AC.

Entering data into Review Manager (RevMan 5.1): RSC, JR, AC.

Entering RevMan statistical data: RSC, JR, AC.

Performing other statistical analysis not using RevMan: RSC.

Performing double entry of data: RSC.

Interpreting data: RSC, JR, AC.

Making statistical inferences: RSC, JR, AC.

Writing the review: RSC.

Providing guidance on the review: JR, AC.

Securing funding for the review: none known.

Serving as guarantor for the review (one author): RSC.

Taking responsibility for reading and checking the review before submission: RSC.

Declarations of interest

None known.

Sources of support

Internal sources

  • No sources of support, Argentina.

External sources

  • No sources of support, Argentina.

Differences between protocol and review

The data collection and analysis sections of the review contain the following changes from corresponding sections of the protocol: In the selection of studies, Appendices 6 and 7 (present in the protocol) were excluded from the review because the data included in these appendices were developed within the text of the review.

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Amato 1998

MethodsRandomized controlled trial. Time period of study: December 1990 through July 1995.
Participants53 participants, aged > 14 and < 70 (two centres). Included: ARDS, LISS > 2.5, Pwedge < 16 mm Hg. Conditions excluded: previous lung or neuromuscular disease, MV > 1 week, uncontrolled terminal disease, previous barotrauma, previous lung biopsy or resection, uncontrollable and progressive acidosis, signs of intracranial hypertension, and documented coronary insufficiency.
Interventions

Control (24): MV: VT: 12 mL/kg, PEEP to optimize FIO2 < 0.6 with adequate systemic oxygen delivery.

Intervention (29): MV: VT: ≤ 6 mL/kg, recruiting manoeuvres, driving pressure < 20 cm H2O, PEEP: 2 cm H2O above Pflex or 16 cm H2O if no Pflex.

Outcomes

Primary: mortality at day 28.

Secondary: mortality before hospital discharge, barotrauma, weaning rate adjusted for APACHE II score.

Other outcomes: mortality in the intensive care unit (ICU), death after weaning of MV, nosocomial pneumonia, use of paralysing agents of > 24 hours, neuropathy after extubation, dialysis required, packed red cells infused.

NotesDiscontinued during the fifth interim analysis because of a significant survival difference between the groups.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Unclear riskInsufficient information about the sequence generation process.
Allocation concealment (selection bias)Low riskOpaque sealed envelopes with a 1:1 assignment scheme.
Blinding of participants and personnel (performance bias)
All outcomes
Low riskIncomplete blinding (blinding of participants but not of personnel), but outcomes not influenced.
Blinding of outcome assessment (detection bias)
All outcomes
Low riskNo blinding, but outcomes not influenced.
Incomplete outcome data (attrition bias)
All outcomes
Low riskAnalysis on the basis of the intention-to-treat principle. Minor protocol violations in both groups: 4 out of 29 participants from intervention group and 1 out of 24 participants from control group.
Selective reporting (reporting bias)Low riskPublished reports included all expected outcomes.
Other biasLow riskReview author believed the study to be free of other sources of bias.

Brower 2004

MethodsRandomized controlled trial. Time period of study: October 1999 through February 2002.
Participants549 participants, aged > 13 (23 centres). Included: ALI and ARDS (PaO2/FIO2 < 300). Excluded: ≥ 36 hours had elapsed since the eligibility criteria were met, participation in other trials involving ALI within the preceding 30 days, pregnancy, increased intracranial pressure, severe neuromuscular disease, sickle cell disease, severe chronic respiratory disease, body weight greater than 1 kg per centimetre of height, burns over 40% of body-surface area, severe chronic liver disease, vasculitis with diffuse alveolar haemorrhage, a coexisting condition associated with an estimated 6-month mortality rate greater than 50%, previous bone marrow or lung transplant, refusal to be included by attending physician.
Interventions

Control (273) and intervention (276): MV: VT: 6 mL/kg PBW, respiratory rate (breaths/min) 6 to 35 to achieve arterial pH > 7.30, plateau pressure ≤ 30 cm H2O, recruiting manoeuvres: in the first 80 participants (intervention group). Target ranges for oxygenation with PEEP/FIOcombination: PaO2 between 55 and 80 mm Hg or SpO2 between 88% and 95%.

PEEP.

Control: PEEP/FIO2 combination: mean PEEP: 8.6 cm H2O.

Intervention: PEEP/FIO2 combination (programming with higher levels of PEEP): mean PEEP: 13.5 cm H2O.

Outcomes

Primary: mortality before hospital discharge.

Secondary: VFD, days not spent in ICU, days free without organ failure.

Other outcomes: barotrauma, breathing without assistance by day 28.

Notes

In the first 171 participants (85 in control group, 86 in intervention group), the higher-PEEP protocol was different from the other 378 participants, but the adjusted mortality rates in both phases was small and not significant.

Discontinued during the second interim analysis on the basis of the specified futility-stopping rule.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskParticipants randomly allocated because authors used random permuted blocks (restricted randomization).
Allocation concealment (selection bias)Low riskA centralized interactive voice system used.
Blinding of participants and personnel (performance bias)
All outcomes
Low riskIncomplete blinding (blinding of participants but not personnel), but the primary outcome was not influenced.
Blinding of outcome assessment (detection bias)
All outcomes
Low riskNo blinding, but the outcomes not influenced.
Incomplete outcome data (attrition bias)
All outcomes
Low riskNo missing outcome data.
Selective reporting (reporting bias)High riskPrimary outcome (mortality before hospital discharge) was different from that given in the protocol (mortality at 28 days). Certain secondary outcomes in the protocol not assessed in the study.
Other biasLow riskSignificant differences between the two groups in baseline characteristics for the mean age and the mean PaO2/FIO2, but after adjustment for differences in baseline variables, these differences did not change the main results.

Huh 2009

MethodsRandomized controlled trial. Time period of study: July 2004 through September 2006.
Participants57 participants (one centre). Included: ARDS (PaO2/FIO2 ≤ 200). Conditions excluded: elevated intracranial pressure, severe acute cardiac disease, high risk of mortality within 3 months for reasons other than ARDS (cancer patients in the terminal stage of their disease), persistent haemodynamic instability.
Interventions

Control (27) and intervention (30): MV: VT: 6 mL/kg PBW, with allowances of up to 8 mL/kg PBW if the SpO2 were below 88%, or less than 7.2 of arterial pH by severe hypercapnoea.FIO2, PEEP, and respiratory rate were set to achieve an SpO2 between 88% and 92%.

PEEP.

Control: PEEP/FIO2 combination with the goal of obtaining a lower PEEP level compatible with an oxygenation target (SpO2 between 88% and 92%). Mean PEEP: 9 cm H2O.

Intervention: recruiting manoeuvres (extended-sigh method), next PEEP added set to 25 cm H2O and then lowered until a decrease of greater than 2% of saturation from the previous SpO2 and drop of static compliance.

Outcomes

Primary: improvement in oxygenation (PaO2/FIO2).

Secondary: respiratory mechanics (PEEP and dynamic compliance), ICU stay, duration of sedatives and paralysing agents, mortality at day 28, mortality at day 60, duration of MV.

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskAllocation intervention was done by random number table.
Allocation concealment (selection bias)Unclear riskInsufficient information to permit judgement of "low risk" or "high risk".
Blinding of participants and personnel (performance bias)
All outcomes
Low riskIncomplete blinding (blinding of participants but not personnel), but outcomes not influenced.
Blinding of outcome assessment (detection bias)
All outcomes
Low riskNo blinding, but outcomes not influenced.
Incomplete outcome data (attrition bias)
All outcomes
Unclear riskInsufficient reporting of exclusions. Three out of 30 participants missing from control group and 1 out of 31 participants missing from intervention group.
Selective reporting (reporting bias)Low riskPublished reports included all expected outcomes.
Other biasLow riskReview author believed the study to be free of other sources of bias.

Meade 2008

MethodsRandomized controlled trial. Time period of study: August 2000 through March 2006.
Participants983 participants (30 hospitals). Included: ALI and ARDS (PaO2/FIO2 ≤ 250) during invasive MV. Excluded if: left atrial hypertension, anticipated MV < 48 hours, inability to wean from experimental strategies, severe chronic respiratory disease, neuromuscular disease, intracranial hypertension, morbid obesity, pregnancy, lack of commitment to life support, conditions with expected 6-month mortality risk > 50%, participation in a confounding trial.
Interventions

Control (508) and intervention (475): MV: ventilator mode: volume-assist control (intervention), pressure control (control). VT: ≤ 6 mL/kg PBW, plateau pressure ≤ 40 cm H2O (intervention), ≤ 30 cm H2O (control), respiratory rate (breaths/min) ≤ 35, pH ≥ 7.30. Recruiting manoeuvres: intervention group.

PEEP.

Control: PEEP/FIO2 combination: mean PEEP: 9 cm H2O.

Intervention: PEEP/FIO2 combination (programming with higher levels of PEEP): mean PEEP: 12.6 cm H2O.

After the first 161 participants, PEEP levels modified in intervention group for the remaining 822 participants, but mean PEEP in all participants in intervention group the same.

Target ranges for oxygenation with PEEP/FIO2 combination: PaO2 between 55 and 80 mm Hg; SpObetween 88% and 93%.

Use of rescue therapies1 (both groups) in participants with refractory hypoxaemia (PaO2 < 60 mm Hg for at least 1 hour while receiving an FIO2 of 1.0), refractory acidosis (pH ≤ 7.10 for at least 1 hour), or refractory barotrauma (persistent pneumothorax with two chest tubes on the involved side or increasing subcutaneous or mediastinal emphysema with two chest tubes).

Outcomes

Primary: all-cause hospital mortality (mortality before hospital discharge).

Secondary: mortality during MV, mortality in ICU, mortality at day 28, barotrauma, refractory hypoxaemia, refractory acidosis, refractory barotrauma, use of rescue therapies in response to refractory hypoxaemia, refractory acidosis or refractory barotrauma, days of MV, days of ICU, days of hospitalizations.

Notes

After the first 161 participants, PEEP levels were modified in the intervention group for the remaining 822 participants, but these changes did not alter the outcomes.

Rescue therapies: prone ventilation, inhaled NO, high-frequency oscillation, jet ventilation, extracorporeal membrane oxygenation.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskAuthors used random permuted blocks. Participants randomly allocated.
Allocation concealment (selection bias)Low riskAssignment performed by central computerized telephone system. A programming error occurred late that created an unexpected difference in the number of participants allocated to each group, but this problem did not alter the results.
Blinding of participants and personnel (performance bias)
All outcomes
Low riskIncomplete blinding (blinding of participants but not of personnel), but the primary outcome was not influenced.
Blinding of outcome assessment (detection bias)
All outcomes
Low riskBlinding of outcome assessment because one analyst was blinded to allocation.
Incomplete outcome data (attrition bias)
All outcomes
Low riskAnalysis on the basis of the intention-to-treat principle. Seven participants were withdrawn from the study at various times (ranging from study days 1 to 11).
Selective reporting (reporting bias)Low riskPublished reports included all expected outcomes.
Other biasLow riskA significant difference was noted between the two groups in baseline characteristics for the mean age and rate of sepsis, but these differences were minimal after the data were pooled.

Mercat 2008

MethodsRandomized controlled trial. Time period of study: September 2002 through December 2005.
Participants767 participants, aged > 18 (37 centres). Included: ARDS. Excluded: known pregnancy, participation in another trial within 30 days, increased intracranial pressure, sickle cell disease, severe chronic respiratory disease requiring oxygen therapy or home MV, actual body weight exceeding 1 kg/cm of height, severe burns, severe chronic liver disease, bone marrow transplant or chemotherapy-induced neutropenia, pneumothorax, expected duration of MV ≤ 48 hours, decision to withhold life-sustaining treatment.
Interventions

Control (382) and intervention (385): MV: ventilator mode (both groups): volume-assist control, VT: 6 mL/kg PBW, plateau-pressure limit ≤ 30 cm H2O, respiratory rate (breaths/min) ≤ 35 adjusted for a pH between 7.30 and 7.45, recruiting manoeuvres: allowed but not recommended.

Target ranges for oxygenation: PaO2 between 55 and 80 mm Hg; SpObetween 88% and 95%.

PEEP:

Control: Total PEEP between 5 and 9 cm H2O.

Intervention: PEEP level to achieve plateau pressures between 28 and 30 cm H2O.

Use of rescue therapies1 (both groups) when the oxygenation goal was not met despite FIO2 ≥ 0.8.

Outcomes

Primary: mortality at day 28.

Secondary: mortality at day 60, mortality before hospital discharge censored on day 60, VFD, days-free without organ failure and barotrauma between day 1 and day 28.

Notes

1. Rescue therapies: prone ventilation, inhaled NO, almitrine bismesylate.

Discontinued during the 18th interim analysis because of absence of 10% absolute reduction in mortality between groups.

Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskAuthors performed random allocation in permuted blocks stratified by centre.
Allocation concealment (selection bias)Low riskAssignment was performed by centralized-interactive telephone system.
Blinding of participants and personnel (performance bias)
All outcomes
Low riskIncomplete blinding (blinding of participants but not personnel), but outcomes not influenced
Blinding of outcome assessment (detection bias)
All outcomes
Low riskThe main analyses were conducted in a blind fashion.
Incomplete outcome data (attrition bias)
All outcomes
Low riskOne participant (control) was excluded because family withdrew consent after randomization. One participant (intervention) was lost to follow-up after discharge on day 29 and was included in the analysis on the basis of the intention-to-treat principle.
Selective reporting (reporting bias)Low riskPublished reports included all expected outcomes.
Other biasLow riskReview author believed the study to be free of other sources of bias.

Talmor 2008

MethodsRandomized controlled trial.
Participants61 participants, (one centre). Included: ALI and ARDS according to the AECC. Excluded: recent injury or other pathologic condition of the oesophagus, major bronchopleural fistula and solid-organ transplantation.
Interventions

In both groups, control (31) and intervention (30), the goals of MV included: a VT of 6 mL/kg PBW; a recruiting manoeuvre to standardize the history of lung volume, a PaO2 between 55 and 120 mm Hg or an SpO2 between 88% and 98%, an arterial pH of 7.30 to 7.45, and a PaCO2 of 40 to 60 mm Hg.

PEEP:

Control: PEEP/FIO2 combination.

Intervention: Transpulmonary pressure (airway pressure minus pleural pressure) was determined, airway pressure was recorded during MV, and pleural pressure was estimated by an oesophageal balloon catheter. PEEP levels were set to achieve a transpulmonary pressure of 0 to 10 cm H2O at end-expiration.

All measurements were performed within 72 hours of patient inclusion.

Outcomes

Primary: improvement in arterial oxygenation (PaO2/FIO2).

Secondary: indices of lung mechanics and gas exchange (respiratory system compliance and the ratio of physiological dead space to tidal volume), VFD, length of stay in the ICU, days not spent in ICU, mortality at day 28, mortality at day 180, days of ventilation among survivors.

Notes 
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskPatients randomly allocated because authors used random allocation with the use of a block randomization scheme.
Allocation concealment (selection bias)Low riskOpaque sealed envelopes that were randomly ordered.
Blinding of participants and personnel (performance bias)
All outcomes
Low riskIncomplete blinding (blinding of participants but not of personnel), but the outcomes not influenced.
Blinding of outcome assessment (detection bias)
All outcomes
Low riskNo blinding, but outcomes not influenced by lack of blinding.
Incomplete outcome data (attrition bias)
All outcomes
Low riskIn the experimental group, one participant who could not be assessed was included in the analysis on the basis of the intention-to-treat principle.
Selective reporting (reporting bias)Low riskPublished reports included all expected outcomes.
Other biasLow riskReview author believed the study to be free of other sources of bias.

Villar 2006

MethodsRandomized controlled trial. Time period of study: March 1999 through March 2001.
Participants95 participants, aged > 15. Included: persistent ARDS. Excluded: patients with acute cardiac clinical conditions, pregnancy, neuromuscular diseases, high risk of mortality within 3 months for reasons other than ARDS (severe neurologic damage, age > 80 years, and cancer patients in the terminal stage of their disease), or more than two extrapulmonary organ failures.
Interventions

Control (45) and intervention (50): MV: ventilator mode (both groups): volume-assist control, respiratory rate to maintain PaCO2 between 35 and 50 cm H2O.

Control: VT: 9 to 11 mL/kg PBW, PEEP ≥ 5 cm H2O and FIO2 to optimize SpO2 > 90% and PaO2 between 70 and 100 mm Hg.

Intervention: VT: 5 to 8 mL/kg PBW, PEEP: 2 cm H2O above Pflex or 15 cm H2O if no Pflex; FIO2 to optimize SpO2 > 90% and PaO2 between 70 and 100 mm Hg.

Outcomes

Primary: mortality in the ICU.

Secondary: hospital mortality (mortality before hospital discharge), VFD, extrapulmonary organ failure, and barotrauma.

NotesDiscontinued prematurely because the absolute mortality difference between control and intervention groups satisfied the stopping rule.
Risk of bias
BiasAuthors' judgementSupport for judgement
Random sequence generation (selection bias)Low riskAuthors used blocked randomization (restricted randomization) stratified by centre.
Allocation concealment (selection bias)Low riskOpaque sealed envelopes that were randomly ordered.
Blinding of participants and personnel (performance bias)
All outcomes
Low riskIncomplete blinding (blinding of participants but not of personnel), but outcomes not influenced.
Blinding of outcome assessment (detection bias)
All outcomes
Low riskNo blinding, but outcomes not influenced.
Incomplete outcome data (attrition bias)
All outcomes
Low riskThree out of 53 participants missing from intervention group and five out of 50 participants missing from control group, because a centre failed to adhere to the randomization methodology. The final analysis was performed with the remaining 95 participants.
Selective reporting (reporting bias)Low riskPublished reports included all expected outcomes.
Other biasLow riskReview author believed the study to be free of other sources of bias.

Characteristics of excluded studies [ordered by study ID]

StudyReason for exclusion
Amato 1995The 28 participants in this study were included in the large study published in 1998.
Badet 2009This study assessed the role of three ventilation strategies with recruitment manoeuvres and optimal PEEP (PEEP set to 24 cm H2O and then lowered stepwise to step above which PaO2 decreased by ≥ 20%) in a random order in 12 participants with ALI or ARDS. Arterial blood gas values and static compliance were recorded at the end of each strategy. No control group, thus not an RCT.
Burns 200114 participants with ARDS were treated with each of ten tidal volume-PEEP combinations, applied in random order, to assess changes in oxygenation and respiratory compliance. No control group, thus not an RCT.
Carvalho 1997This study was a descriptive analysis of another published study (Amato 1995) and included two groups of participants with different levels of PEEP, in which physiological variables were studied.
Dellamonica 2011In 30 participants with ALI, two levels of PEEP were applied in a random sequence to provide a simple and accurate bedside method for assessing alveolar recruitment while monitoring strain on the lungs. We had no control group, so the trial was not an RCT.
Grasso 2005In 19 participants with ARDS divided into two groups and with different levels of PEEP, changes in oxygenation, static lung elastance and the shape of the volume-pressure curve were assessed. Excluded because the outcomes are physiological.
Grasso 2007 In 15 participants with ARDS, two ventilatory strategies with different levels of PEEP were applied in random order. Physiological parameters and plasma inflammatory mediators were measured. No control group, thus not an RCT.
Kallet 2007This was a review of the different ways of applying PEEP in participants with ARDS.
Oczenski 2004This was a randomized controlled study of participants with early extrapulmonary ARDS. In this study, two groups were assigned, but both had equal levels of PEEP and differed only in the inclusion of a recruitment manoeuvre.
Ranieri 1999This was a randomized controlled trial of participants with ARDS. Two groups were assigned with different VT and levels of PEEP. Excluded because the primary outcome is physiological (pulmonary and systemic concentrations of inflammatory mediators) and mortality is reported as a post hoc analysis.
Richard 2003In this prospective cross-over study, 15 participants with ALI were treated with three ventilatory strategies combining different levels of PEEP and VT to investigate changes in alveolar recruitment and oxygenation. No control group, thus not an RCT.
Toth 2007This was a study in which 18 participants with ARDS underwent recruitment manoeuvres and decremental PEEP to reach optimal PEEP (PEEP set to 26 cm H2O and then lowered stepwise to step above which PaO2 decreased by > 10%) for assessment of respiratory and haemodynamic changes. No control group, thus not an RCT.

Characteristics of ongoing studies [ordered by study ID]

Kacmarek 2007

Trial name or titleARDSnet Protocol vs Open Lung Approach in ARDS
MethodsRandomized controlled trial
ParticipantsIncluded: participants with ARDS in the first 48 hours after diagnosis, PaO2/FIO2 must be < 200 with standard ventilator setting, no lung recruitment manoeuvres or adjunct therapy, total time on mechanical ventilation < 96 hours at time of randomization. Excluded: age < 18 years or > 80 years, weight < 35 kg PBW, body mass index > 60, intubated secondary to acute exacerbation of a chronic pulmonary disease, acute brain injury (ICP > 18 mm Hg), immunosuppression 2° to chemotherapy or radiation therapy, severe cardiac disease, pregnancy, sickle cell disease, neuromuscular disease, more than two organ failures, barotrauma, persistent haemodynamic instability, penetrating chest trauma, enrolment in another interventional study.
Interventions

Control: PEEP/FIO2 combination with target ranges for oxygenation: PaO2 between 55 and 80 mm Hg; SpO2 between 88% and 95%.

Intervention: uses a technique with recruitment manoeuvres and optimal PEEP (the optimal PEEP level is determined by a decremental PEEP trial involving a series of pressure measurements taken after the recruitment manoeuvre).

Outcomes
  • Mortality at day 60.

  • Mortality in the intensive care unit (ICU),

  • Mortality before hospital discharge.

  • Mortality at day 28,

  • Mortality at day 180.

  • Mortality at day 365.

  • Ventilator-free days.

  • Length of ICU stay.

  • Development of extra-pulmonary organ failures.

  • Duration of hospitalization.

  • Incidence of barotrauma.

  • Systemic inflammatory mediator levels.

  • Lung function 6 months after discharge.

  • Lung function 12 months after discharge.

  • Need for rescue therapy.

  • Ventilation-associated pneumonia rate.

Starting dateFebruary 1, 2007.
Contact informationRobert Kacmarek, Massachussets General Hospital, Boston, Massachussets, United States.
Notes 

Pintado 2003

Trial name or titlePilot study of positive end-expiratory pressure in acute respiratory distress syndrome (PEEP-HUPA).
MethodsRandomized controlled pilot trial.
ParticipantsIncluded: participants with ARDS after 24 hours under mechanical ventilation. Excluded: age < 18 years, pregnancy, neuromuscular diseases, intracranial hypertension (head trauma), left ventricular dysfunction, mechanical ventilation for longer than 72 hours, previous barotrauma, participants with terminal stage of an illness and high risk of mortality within 90 days, participants who refused to consent to the study.
Interventions

In both groups: MV: VT of 6 mL/kg PBW; plateau pressure ≤ 35 cm H2O, respiratory rate (breaths/min) 30 to 35 to maintain a pH of 7.30 to 7.45 and FIO2 to optimize SpO2 between 88% and 95% or PaO2 between 55 and 80 mm Hg.

PEEP:

Compliance-guided PEEP group: The PEEP level is programmed on a daily basis, according to the method described by Suter in 1978. The static compliance (Cst) is calculated at different levels of PEEP at a constant VT. The maximum value of Cst in individual patients is considered to be the best PEEP.

FIO2-driven-PEEP group: PEEP/FIO2 combination.

Outcomes

Primary: evolution of arterial oxygenation during the 28 days after study randomization.

Secondary:mortality at day 28, ventilator-free days.

Starting dateJanuary 2003.
Contact informationMaría del Consuelo Pintado, Critical Care Unit, Universitary Hospital Príncipe de Asturias, Spain.
NotesThis study has been completed.

Ancillary