Home > Health & Research > Health Education Campaigns & Programs > Cochrane Neonatal Review > Positive end expiratory pressure for preterm infants requiring conventional mechanical ventilation for respiratory distress syndrome or bronchopulmonary dysplasia

Positive end expiratory pressure for preterm infants requiring conventional mechanical ventilation for respiratory distress syndrome or bronchopulmonary dysplasia

Skip sharing on social media links
Share this:

Authors

Nicolas Bamat1, David Millar2, Sanghee Suh3, Haresh Kirpalani4

Background - Methods - Results - Characteristics of Included Studies - References - Data Tables & Graphs


1Pediatric Residency Program, Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, USA [top]
2Regional Neonatal Intensive Care Unit, Royal Jubilee Maternity Service, Belfast, UK [top]
3Department of Pediatrics, Children's Hospital of Philadelphia, Philadelphia, Pennsylvania, USA [top]
4Department of Pediatrics, University of Pennsylvania School of Medicine and Dept of Clinical Epidemiology and Biostatistics, McMaster University, Philadelphia, Pennsylvania, USA [top]

Citation example: Bamat N, Millar D, Suh S, Kirpalani H. Positive end expiratory pressure for preterm infants requiring conventional mechanical ventilation for respiratory distress syndrome or bronchopulmonary dysplasia. Cochrane Database of Systematic Reviews 2012, Issue 1. Art. No.: CD004500. DOI: 10.1002/14651858.CD004500.pub2.

Contact person

Nicolas Bamat

Pediatric Residency Program
Children's Hospital of Philadelphia
34th Street and Civic Center Boulevard
Philadelphia Pennsylvania 19104
USA

E-mail: nbamat@gmail.com

Dates

Assessed as Up-to-date: 08 September 2011
Date of Search: 10 November 2010
Next Stage Expected: 08 November 2012
Protocol First Published: Issue 4, 2003
Review First Published: Issue 1, 2012
Last Citation Issue: Issue 1, 2012

Abstract

Background

Conventional mechanical ventilation (CMV) of neonates has been used as a treatment of respiratory failure for over 30 years. While CMV facilitates gas exchange, it may simultaneously damage the lung. Positive end expiratory pressure (PEEP) has received less attention than other ventilation parameters when considering this balance of benefit and possible harm. While an appropriate level of PEEP may exert substantial benefits in ventilation, both inappropriately low or high levels may lead to harm. An appropriate level of PEEP for neonates may also be best achieved by an individualized approach.

Objectives

  1. To compare the effects of different levels of PEEP in preterm newborn infants requiring CMV for respiratory distress syndrome (RDS).
  2. To compare the effects of different levels of PEEP in preterm infants requiring CMV for bronchopulmonary dysplasia (BPD).
  3. To compare the effects of different methods for individualizing PEEP to an optimal level in preterm newborn infants requiring CMV for RDS.

Search methods

The search was performed in accordance with the standard search strategy for the Cochrane Neonatal Review Group. The Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library), Ovid MEDLINE, EMBASE, study references and experts were utilized for study identification.

Selection criteria

All randomized and quasi-randomized controlled trials studying preterm infants (less than 37 weeks gestational age) requiring CMV with endotracheal intubation and undergoing randomization to either different PEEP levels (RDS or BPD) or two or more alternative methods for individualizing PEEP levels (RDS only) were included. Cross-over trials were included but we limited the findings to those in the first cross-over period.

Data collection and analysis

Data collection and analysis were performed in accordance with the recommendations of the Cochrane Neonatal Review Group.

Results

An initial evaluation identified 10 eligible articles. Ultimately, a single study met our inclusion criteria. The study addressed the effects of different levels of PEEP in preterm newborn infants requiring CMV for RDS. Only short term physiologic measures were reported. All results were limited to a small sample size without statistically significant results. No trials addressing the effect of PEEP in infants with BPD or strategies to individualize the management of PEEP were identified.

Authors' conclusions

There is insufficient evidence to guide selection of appropriate PEEP levels for RDS or CMV. There is a need for well designed clinical trials evaluating the optimal application of this important and frequently applied intervention.

Plain language summary

PEEP for preterm infants receiving conventional mechanical ventilation for respiratory distress syndrome or bronchopulmonary dysplasia

Adequate gas exchange is readily accomplished in full term infants with appropriately developed lungs. In contrast, premature infants with respiratory distress syndrome (RDS) or bronchopulmonary dysplasia (BPD) often require medical support to achieve gas exchange. Conventional mechanical ventilation (CMV) is a common therapy used to accomplish this.

CMV allows for oxygenated air to be driven into an infant's lungs from a ventilator through a tube that is placed inside the infant's trachea. The gas exchange is facilitated by a set of pressures applied during a respiratory cycle supported or maintained by the ventilator. One of these pressures is known as positive end expiratory pressure (PEEP) and can be thought of as a continuous pressure that is applied throughout the respiratory cycle. It plays an important role in keeping the lungs open and preventing collapse so that all areas of the lungs can participate in gas exchange.

While it is generally accepted that some level of PEEP is important and necessary to accomplish adequate gas exchange, it is not clear what level of PEEP results in the greatest benefit. While too little PEEP likely fails to provide adequate gas exchange, too much PEEP may lead to over distension of the lungs resulting in harm.

This review was performed to assess what the best level of PEEP is for preterm infants requiring CMV for either RDS or BPD. It also searched for evidence of any strategies that have been effective in determining the best level of PEEP for infants on an individual case by case basis.

The results of this review highlight that there has been very little good quality research in the form of randomized controlled trials assessing the effects of different PEEP levels. Only a single study, performed in infants receiving CMV for RDS, met our criteria for inclusion. This study was small and the results were not sufficient to lead to any recommendations on the best PEEP level for infants with RDS. No study met the inclusion criteria for the review for either BPD or for strategies attempting to identify an individualized PEEP level. This review draws attention to the need for more randomized controlled trials addressing these unanswered questions.

[top]

Background

Conventional mechanical ventilation (CMV) of neonates has been used as a treatment of respiratory failure for 30 years, and the technology has developed dramatically in this time. However, the basis of CMV remains unchanged: humidified, heated, oxygenated gas is driven into the infant's lungs for a set period of time (inspiratory time or Ti) at a given pressure (peak inspiratory pressure or PIP), followed by an expiratory phase of set duration (expiratory time or Te) where positive end expiratory pressure (PEEP) is maintained. This cycle is repeated a number of times per minute, giving a respiratory rate or frequency.

The understanding that while mechanical ventilation facilitates gaseous exchange it may simultaneously damage the lung (volu-baro-trauma) (Michna 1999; Tremblay 2002) has become widely accepted. Most attention has focused on the potentially damaging effects of PIP. The need to consider other parameters in examining this balance was highlighted by the concept of a 'mean airways pressure' by Boros in 1979 (Boros 1979). This value encapsulated various parameters of ventilation such as PIP and PEEP into a single pressure value. While convenient, it hindered the study of the potentially important contributions of each component part. Although acknowledged, the contribution of the level of PEEP on this balance has not been well studied in the neonatal literature (Monkman 2003).

The risks and benefits of PEEP should be considered when setting this pressure parameter. The beneficial effects of PEEP may be related to an increase in functional residual capacity, alveolar recruitment, reduced work of breathing, or an improvement in the distribution of ventilation to perfusion (Gregory 1971; Richardson 1978; Richardson 1986; Thome 1998). Both Corbridge 1990 and Froese 1993 demonstrated in animal models that zero end expiratory pressure (ZEEP), or near-ZEEP, is deleterious. However, concerns have been raised about the potential risks of inappropriately high PEEP, specifically air leak syndromes, gas trapping, and deleterious effects on cardiac function (Reynolds 1971; Simbruner 1986; Hausdorf 1987). Impaired venous return is generally considered the main cause of depressed cardiac function with the application of PEEP.

Adjusting the level of PEEP is important in neonates requiring CMV. If the level of PEEP is too low, alveolar collapse leads to impaired gaseous exchange and a probable increase in lung damage. Higher PEEP levels appear to reduce lung damage in a rabbit model of neonatal respiratory distress syndrome (Sandhar 1988). Alternatively, PEEP levels set too high may decrease tidal and minute ventilation, impair expiration, decrease cardiac output, and cause lung damage though over-distension leading to pulmonary interstitial emphysema, pneumothorax, and other types of air leaks syndromes.

In the ventilation of adults with adult respiratory distress syndrome, the concept of a 'best' PEEP that trades off these competing factors has been recognized (Amato 1998). Randomized clinical trials comparing different levels of PEEP have been performed (Brower 2004; Meade 2008; Mercat 2008) and a systematic review comparing different PEEP levels on patient important outcomes has recently been reported (Briel 2010). In contrast, this question has not been well addressed in neonates requiring CMV for RDS or BPD.

In this review, we will consider PEEP levels of < 5 cm H2O to comprise 'low' PEEP, and greater than/or equal to 5cm H20 to comprise 'high' PEEP. The reason for the cut-off at 5 cm H2O, although arbitrary, reflects a common value applied in many neonatal units. Additionally, da Silva 1994 (da Silva 1994) demonstrated potentially useful increases in functional residual capacity (FRC) from pressure of 2 to 5 cm H2O; and at 5 cm H2O the mean values for FRC were similar to the normal values of healthy term infants.

It is plausible that the appropriate level of PEEP for neonates depends on factors influencing the underlying pathophysiology on a case by case basis. For example, changes in pulmonary physiology secondary to differences in gestational age or prior treatments, such as exogenous surfactant therapy, may influence the impact of different PEEP levels.

It is possible then that rather than a single value of PEEP for all infants, an individualized level is more appropriate. In fact, in some of the earliest applications of continuous positive airway pressure (CPAP) therapy, Bonta (Bonta 1977) suggested that it was possible to determine an 'optimum' individual pressure for CPAP in an infant with respiratory distress syndrome (RDS). In that study 'optimum' was defined as that level of PEEP at which the maximum reduction in required oxygen was achieved. In an early attempt at individualization, Mathe and colleagues proposed examining characteristics of individual pressure-volume curves to identify the lower inflexion point of the inspiratory limb as the point of 'appropriate PEEP' (Mathe 1987). Controlled trials in newborn animal models suggest that setting PEEP levels above this level may provide an individualized technique to minimize lung damage (Monkman 2001). We wish to determine if there are effective methods of individualizing PEEP therapy during conventional mechanical ventilation and to determine which of these methods are associated with improved outcomes.

In summary, in this review we examine the following questions. For infants requiring CMV for RDS or BPD, what should be recommended as the clinical standard; high or low PEEP? Furthermore, are there methods of determining an appropriate individualized PEEP strategy for infants requiring CMV for RDS?

Objectives

There are three broad objectives to this review.

  1. To compare the effects of different levels of PEEP in preterm newborn infants requiring CMV for respiratory distress syndrome (RDS).
  2. To compare the effects of different levels of PEEP in preterm infants requiring CMV for bronchopulmonary dysplasia (BPD).
  3. To compare the effects of different methods for individualizing PEEP to an optimal level in preterm newborn infants requiring CMV for RDS.

For each objective, the effects of intervention were evaluated through several pre-specified measured outcomes. These can be found detailed by the comparisons and are comprised of the objectives above and, when applicable, subgroups within these objectives (see Methods) in Table 1 below.

[top]

Methods

Criteria for considering studies for this review

Types of studies

All randomized and quasi-randomized controlled trials were included. Randomized cross-over designs were eligible. We excluded trials of high frequency oscillatory ventilation (HFOV) or high frequency jet ventilation (HFJV).

Types of participants

Preterm infants (less than 37 weeks gestational age) requiring CMV with endotracheal intubation for RDS or BPD.

For objective 1: subgroup analysis was planned for gestational age (GA) (< 28 weeks GA, 28 to 31 weeks GA, greater than/or equal to 32 weeks age) and for pre versus post-surfactant therapy. As such, these characteristics were considered when evaluating types of participants.

Types of interventions

Randomization to different PEEP levels in infants requiring CMV for RDS or BPD.

Interventions for comparison included randomization to either:

  1. different PEEP levels in infants requiring CMV for RDS or BPD, which included both comparisons of ZEEP versus PEEP and comparisons of two different levels of PEEP;
    or
  2. two or more alternative methods for individualizing PEEP in infants requiring CMV for RDS.

Types of outcome measures

The considered outcomes varied slightly among the three objectives. Please refer to Table 1, 'Table of comparisons', for a detailed chart of all considered outcomes by objective, intervention, and subgroup. Outcomes from cross-over trials were limited to those measured at the end of the first cross-over period.

For comparisons within objective (1): different levels of PEEP in RDS

Primary outcomes
  1. Mortality at postnatal age of 28 days or 40 weeks corrected gestational age
  2. Neurodevelopmental outcome at two years of age, using Bayley Mental Developmental Index (MDI) or Psychomotor Developmental Index (PDI)
Secondary outcomes
  1. Incidence of BPD (defined as requiring oxygen supplementation at 28 days postnatal age or 36 weeks postmenstrual age)
  2. Incidence of air leak syndromes; specifically pulmonary interstitial emphysema, pneumothorax, pneumomediastinum, and pneumopericardium
  3. Duration of ventilatory support in days
  4. Fractional inspired oxygen needed to maintain arterial gas or oxygen saturation levels. Where data allowed, these were described as alveolar-arterial oxygen tension differences (AaDO2) in mmHg
  5. Arterial carbon dioxide (PaCO2) or transcutaneous carbon dioxide levels in mmHg
  6. Direct measures of cardiac output (or surrogate markers of cardiac output, if these became apparent during analysis of the data)
  7. Blood pressure measurements in mmHg and the requirement for additional boluses of fluid or inotropic support in ml
  8. Incidence of grade III-IV intraventricular hemorrhage or periventricular leukomalacia, or both, as defined by neuroimaging with cranial ultrasound (Papile 1978)
  9. Functional residual capacity in ml

For comparisons within objective (2): different levels of PEEP in BPD

Primary outcomes
  1. Mortality at two years
  2. Poor neurodevelopmental outcome using Bayley MDI or PDI at two years
Secondary outcomes
  1. Duration of ventilatory support in days
  2. Incidence of air leak syndromes; specifically pulmonary interstitial emphysema, pneumothorax, pneumomediastinum, and pneumopericardium
  3. Fractional inspired oxygen needed to maintain arterial gas or oxygen saturation levels. Where data allowed, these were described as AaDO2 in mmHg
  4. Direct measures of cardiac output (or surrogate markers of cardiac output, if these became apparent during analysis of the data)
  5. Blood pressure measurements in mmHg and the requirement for additional boluses of fluid or inotropic support in ml
  6. Incidence of grade III-IV intraventricular hemorrhage or periventricular leukomalacia, or both, as defined by neuroimaging with cranial ultrasound (Papile 1978)
  7. Functional residual capacity in ml
  8. Arterial carbon dioxide (PaCO2) or transcutaneous carbon dioxide levels in mmHg

For comparisons within objective (3): different methods to optimize PEEP in RDS

Primary outcomes
  1. Mortality by postnatal age of 28 days or 40 weeks postmenstrual age
  2. Neurodevelopmental outcome at two years of age, using Bayley MDI or PDI
Secondary outcomes
  1. Incidence of BPD defined as requiring oxygen supplementation at 28 days postnatal age or 36 weeks postmenstrual age
  2. Incidence of air leak syndromes; specifically pulmonary interstitial emphysema, pneumothorax, pneumomediastinum, and pneumopericardium
  3. Duration of ventilatory support in days
  4. Fractional inspired oxygen needed to maintain specified arterial gas or oxygen saturation levels. Where data allowed, these were described as AaDO2 in mmHg
  5. Arterial carbon dioxide (PaCO2) or transcutaneous carbon dioxide levels in mmHg
  6. Incidence of grade III-IV intraventricular hemorrhage or periventricular leukomalacia, or both, as defined by neuroimaging with cranial ultrasound (Papile 1978)
  7. Functional residual capacity in ml
  8. Direct measures of cardiac output (or surrogate markers of cardiac output, if these became apparent during analysis of the data)
  9. Blood pressure measurements in mmHg and the requirement for additional boluses of fluid or inotropic support in ml

Search methods for identification of studies

Electronic searches

We used the standard search strategy of the Cochrane Neonatal Review Group.

We searched the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library), Ovid MEDLINE and EMBASE for potential studies through November 2010. We did not apply any language or date of publication limits. When applicable, we used both controlled vocabulary and free text terms to maximize search sensitivity. A detailed demonstration of the search strategies for each database is provided in Appendix 1. Each search strategy was executed independently by two review authors, with any discrepancies followed by discussion and resolution mediated by a third author when necessary.

Searching other resources

Reference sections of studies considered for inclusion were searched independently by two authors. Additionally, we consulted experts in the field for knowledge of potential studies for inclusion.

Data collection and analysis

Selection of studies

We exported all electronic database results to RefWorks (ProQuest LLC; Ann Arbor, MI). We eliminated duplicates. Titles and abstracts were considered for each result. Studies clearly not meeting the criteria for inclusion were eliminated without further consideration. Studies in which meeting the inclusion criteria seemed likely or could not be fully assessed were considered in detail with the use of a standardized and previously prepared data extraction form. This form guided inclusion or exclusion determinations as well as extraction of data for candidate studies. All of these steps were performed independently by two review authors. A third, senior author resolved differences and uncertainties when present.

We attempted to contact the authors directly in cases where additional data or information were needed. We verified appropriate contact information and e-mail addresses through correspondence with experts in the field as well as known collaborators. A minimum of three separate attempts at correspondence were made in cases for which initial attempts were unsuccessful.

Data extraction and management

Extraction of relevant data eligible for analysis was guided by a standardized and previously prepared data extraction form that was applied to each candidate study. Original data were taken from published reports of trials and, when necessary, reorganized to facilitate the calculation of mean and standard deviation values for the outcomes of interest. These steps were independently performed, in duplicate, by two authors; with identical results then being confirmed. Data entry, construction of comparison tables and graphs were performed using RevMan software.

Assessment of risk of bias in included studies

Assessment of risk of bias in included studies was guided by a standardized and previously prepared data extraction form applied to each candidate study. Selection bias was evaluated by considering the adequacy of the randomization technique. Performance bias was evaluated by considering the blinding of the intervention. Detection bias was evaluated by considering the blinding of outcome assessment. Attrition bias and selective reporting were evaluated by considering the completeness of follow up and reported results. Each criterion was assessed as 'YES' or 'NO' with space provided in the data extraction form for notes in instances in which the determination was unclear or required further description. Studies with significant bias were excluded from the review.

These assessments were performed independently by two separate authors. A third, senior author resolved differences and uncertainties when present.

Measures of treatment effect

Categorical data were to be extracted and relative risk, relative risk (RR) reduction, risk difference (RD), and number needed to treat (NNT) calculated. No eligible categorical data has been identified for any of the pre-specified outcomes for any of the objectives in this review to date.

For continuous data, the mean and standard deviation were obtained. The analysis was performed using the weighted mean difference (WMD). For each measure of effect, 95% confidence intervals were calculated.

Data derived from cross-over trials were limited to those measured at the end of the first cross-over period. Categorical and continuous data from such trials were employed as detailed above, respectively.

Assessment of heterogeneity

Assessment of heterogeneity was to be performed by considering differences in study participants, interventions, study design and study quality. A decision to pool or not pool outcome results from differing studies, based on the degree of heterogeneity, was to be performed in duplicate by the study authors, with a third senior author resolving differences.

For pooled results, a Chi2 test was to be applied to the results to further evaluate the presence of heterogeneity, with a P value of 0.10 used to determine statistical significance. This was to be followed by determination of an I2statistic to assess the degree of heterogeneity. Heterogeneity will be further explored by an assessment of the subgroup comparisons previously stated for this review, when applicable. To date, all outcomes result from a single study. As such, an assessment of heterogeneity has not been necessary.

Assessment of reporting biases

Assessment of reporting bias was to be performed using funnel plots of effect estimates using RevMan software, when applicable. At this time, insufficient studies and associated effect estimates have been identified to warrant these strategies. As such, an assessment of reporting bias has not been performed to date.

Data synthesis

A fixed-effect model for meta-analysis was to be undertaken using RevMan 5 software. At this time, a single study has been identified for inclusion. As such, a meta-analytic synthesis has not been performed to date.

[top]

Results

Description of studies

Results of the search

The MEDLINE search returned 707 articles, the EMBASE search returned 168 articles, and the CENTRAL search returned 13 articles for further evaluation. Using reference software, the search results were merged and duplicates were deleted. An initial assessment of the remaining abstracts was performed, eliminating articles clearly not meeting eligibility.

Ultimately 10 studies were identified for potential inclusion. Six of these articles resulted from the electronic search strategy, three were provided by experts and one article was identified in searching the reference section of an article considered for inclusion.

Included studies

See: Characteristics of Included Studies

A single study (Aufricht 1995) met the inclusion criteria following detailed review.

This three period cross-over study randomized 13 infants from a single center in Vienna, Austria who were on CMV for RDS to three different PEEP levels. Mean gestational age and range were provided for the study participants as a whole but not for individual patients or by group in the initial randomization. Weight and age in days were reported for each individual participant. The patients were initially randomized to CMV with a PEEP of either 2, 4 or 6 cm H2O. Following the initial period, they were switched to the two remaining levels as determined by a random sequence that corresponded to each infant. The infant remained at each level for 30 minutes prior to being crossed-over to the next period. While on CMV, peak pressures were simultaneously altered to keep the inflation pressures constant. Other respiratory settings were not changed. Patients did not all receive the same sedation and ventilation techniques. Outcomes were measured at the end of each 30 minute period. These outcomes included two of our pre-specified secondary outcomes of interest, PaCO2 and AaDO2. The results of the measured outcomes were detailed for each infant after each cross-over period, allowing isolation of data after the first cross-over period.

Infants were randomized to a PEEP of 2 cm H2O (n = 4), 4 cm H2O (n = 6), or 6 cm H2O (n = 3). As defined for the review, low PEEP is < 5 cm H2O while high PEEP is greater than/or equal to 5 cm H2O. As such, the first two groups were collapsed, resulting in low PEEP (n = 10) versus high PEEP (n = 3). Lastly, a single infant from the group initially randomized to 2 cm H2O received deliberate hyperventilation to a pre-specified PaCO2 range as part of the ventilatory management. This infant was excluded from the analysis, resulting in measurements comparing patients undergoing low PEEP (n = 9) versus high PEEP (n = 3).

The study authors did not address the unequal distribution to the three treatment groups in the initial randomization. Because the original study employed a cross-over design ensuring an equal number of study periods for each treatment group at the conclusion of the study, it is possible that the study authors were less concerned about the uneven initial chance randomization.

Details can be found in 'Characteristics of Included Studies'.

Excluded studies

See: Characteristics of excluded studies for detailed information regarding excluded studies

Eight of the studies evaluated for potential inclusion were excluded following detailed assessment.

Four of the studies (Belenky 1976; Finer 2004; Hausdorf 1987; Stewart 1981) did not meet the methodological criteria upon further review. While the latter two were not randomized controlled trials, the former two did not randomize the patients requiring CMV to different PEEP levels. Three studies (Dimitriou 1999; Herman 1973; Simbruner 1989) required further information to determine classification. Correspondence with Dimitrou established that while the study methods were eligible, the necessary original data were no longer available. For the remaining two studies we could not secure the data to further evaluate the methodology, needed for consideration of their inclusion. A final study (Durbin 1976) employed a methodology that led to a lack of true randomization to our interventions of interest. As such, it was excluded following application of our pre-stated criteria for assessment of risk of bias.

Detailed information can be found for each study in 'Characteristics of excluded studies'.

Studies awaiting classification

See: Characteristics of studies awaiting classification for detailed information regarding excluded studies

One of the studies (Dinger 1999) that was evaluated could potentially meet the inclusion criteria. However, additional information and or data would be needed to confirm its inclusion. Correspondence with the study authors has been unsuccessful to date. Detailed information can be found in 'Characteristics of studies awaiting classification'.

Risk of bias in included studies

A single study (Aufricht 1995) met the inclusion criteria. In this study, patients were allocated to an initial PEEP level by a random sequence table. Further details on the methods for generating or obtaining this table were not provided. Blinding of personnel to the allocated intervention during the study or at the time of outcome assessment was not addressed but was unlikely given the study design. Data for all 13 patients were reported for each measured outcome and the study was free of suggestion of selective reporting. For one participant, PaCO2 was reported as a measured outcome although the ventilation technique for that particular infant involved deliberate hyperventilation to achieve a pre-specified PaCO2. The presentation of the study results allowed isolation of this infant, which was excluded from the analyses of outcomes as this technique could act as a potential confounder.

Effects of interventions

All comparisons of interest for this review and the considered outcomes for each comparison are provided in tabular form and can be found within Table 1 (Table of Comparisons; Comparisons A through O). There was an absence of eligible data to evaluate all outcomes for comparisons A-F, H-J and L-O and for outcomes 1 to 5 and 8 to 11 for comparisons G and K. The only reported outcomes are for comparison G, outcome 6 and outcome 7 and comparison K, outcome 6 and outcome 7.

A single study (Aufricht 1995), containing two outcomes of interest, was identified for inclusion. Mean gestational age and range was reported for the study subjects as a whole, but was not provided for each comparison group or for each individual subject. As such, subgroup comparisons based on gestational age could not be performed. Surfactant was not used in any patients in this study. As such, a subgroup comparison between pre and post surfactant was not performed.

Comparison G. Low PEEP versus high PEEP in preterm newborn infants requiring CMV for RDS

Outcome 6. AaDO2 (analysis 1.1)

The single included trial (Aufricht 1995) assessed this comparison and outcome.

The review analysis for this outcome showed no evidence of effect, with a lower but not significantly different AaDO2 in infants randomized to low PEEP when compared to high PEEP (mean difference (MD) -61.89 mmHg, 95% CI -189.05 to 65.27), as seen in Figure 3.

Outcome 7. PaCO2 or transcutaneous CO2 levels (analysis 1.2)

The single included trial (Aufricht 1995) assessed this comparison and outcome.

The review analysis for this outcome showed no evidence of effect, with a lower but not significantly different PaCO2 in infants randomized to low PEEP when compared to high PEEP (MD -1.82 mmHg, 95% CI -16.56 to12.92), as seen in Figure 1.

Comparison K. Low PEEP versus high PEEP in preterm newborn infants requiring CMV for RDS, prior to surfactant therapy

Outcome 6. AaDO2 (analysis 2.1)

All infants in Aufricht 1995 were managed without the use of surfactant therapy. As such, all participants from that study fell within the pre-surfactant subgroup. All details and results for this comparison were identical to those provided immediately above for Comparison G, Outcome 6. See Figure 4.

Outcome 7. PaCO2 or transcutaneous CO2 levels (analysis 2.2)

All infants in Aufricht 1995 were managed without the use of surfactant therapy. As such, all participants from that study fell within the pre-surfactant subgroup. All details and results for this comparison were identical to those provided immediately above for Comparison G, Outcome 7. See Figure 2.

Discussion

Summary of main results

The principal result of this study was the finding of lack of evidence with respect to the review's aims, which were listed previously under Objectives.

Multiple outcomes for comparisons were identified a priori to evaluate these objectives and are listed in Table 1. Ultimately, not a single eligible study addressing Objectives 2 or 3 was identified. For Objective 1, a single study meeting the eligibility criteria was identified (Aufricht 1995). While this study provided data for two of our pre-specified secondary outcomes of interest (PaCO2 and AaDO2), it did not provide any data for patient important outcomes such as our primary outcomes of mortality and neurodevelopment. For the two studied outcomes, PaCO2 and AaDO2, the trial did not find a significant difference between low PEEP and high PEEP.

Overall completeness and applicability of evidence

The findings of this review are insufficient to substantially address any of the objectives of this study. The results are limited to the effects of high versus low PEEP for infants undergoing CMV for RDS on PaCO2 and AaDO2 levels, and the study does not find a significant difference between different PEEP levels for these outcomes. The infants were not treated with surfactant despite evidence of RDS, deviating from current standard practice.

The findings of this review are limited to data from 12 patients in a single study from a center in Vienna, Austria (Aufricht 1995). This study is clearly underpowered to detect any important clinical differences at a statistically significant level. While the effect sizes of the interventions on the reported outcomes suggest a possible clinically important difference between low and high PEEP, the paucity of existing data is such that the effect size could vary considerably with the inclusion of additional data. The effect size on the existing outcomes provides an unreliable estimate of the true effect of the intervention and is a limitation of this current review.

As such, the results of this review can not provide evidence to guide current clinical practice for any of the three objectives. This underscores the need for well designed randomized trials evaluating the impact of different PEEP levels on patient outcomes in infants undergoing CMV for RDS or BPD.

Quality of the evidence

The findings of this review are limited to data from a single study (Aufricht 1995). The results are limited to the effects of high versus low PEEP for infants undergoing CMV for RDS on PaCO2 and AaDO2, and do not find a significant difference between the groups for these outcomes.

While blinding of the intervention and outcomes are not addressed in the study, the allocation of participants to different PEEP levels via random tables and the complete and transparent nature of the study results are strengths that support the internal validity of these results. A key shortcoming of this study is the inconsistency in the sedation and ventilation techniques applied between study participants. While some patients were sedated with chloral hydrate, paralyzed with pancuronium, and ventilated with controlled ventilation (N = 5), others received no sedation while spontaneously breathing on intermittent mandatory ventilation (IMV) (n = 7). While the groups were combined when reporting the measured outcome, it is possible that these different techniques had a confounding effect on the results. These methodological shortcomings detract from the internal validity of this study. An additional limitation was the short study period (30 minutes). It is possible that there was inadequate time to achieve stabilization at the randomized PEEP levels, and that the effects of these interventions would differ if measured following a longer intervention time.

One infant from the study was excluded for the purpose of our analyses. This infant was sedated with chloral hydrate and hyperventilated to a PaCO2 less than 35 mmHg with controlled ventilation. This ventilation technique disallowed the use of PaCO2 and AaDO2 as outcome measures for this infant.

Potential biases in the review process

This review is limited by several potential sources of bias.

The electronic search strategy, found in Appendix 1, was created a priori and with a deliberate effort to maximize the sensitivity of the strategy for identifying relevant trials. However, it must be noted that of the 10 studies ultimately considered for inclusion, only six resulted from the original electronic search. Of the four remaining articles, three were known and provided by one of our senior authors. The remaining article was identified through a search of the reference section of one of the articles being considered for inclusion. This suggests that our search strategy, despite providing nearly 900 articles for assessment, potentially lacked the sensitivity to identify all relevant articles. An assessment revealed that the missing of these articles may have arisen from inappropriate keyword and filter classifications on the part of the electronic databases used to identify studies in this review.

For the purpose of this review, we created a dichotomy by defining low PEEP as < 5 cm H2O and high PEEP as greater than/or equal to 5 cm H2O. Although there are some small animal data (da Silva 1994) to suggest 5 cm H2O as an appropriate cut-off value, this separation is somewhat arbitrary. Our single included study (Aufricht 1995) randomized patients to PEEP levels of 2, 4 and 6 cm H2O. As specified in our a priori methodology, the outcomes of the infants in the 2 and 4 cm H2O groups were combined into the low PEEP group and compared to the high PEEP group, composed of those infants randomized to 6 cm H2O. As the difference in the intervention between infants assigned to 2 cm H2O versus 4 cm H2O and between infants assigned to 4 cm H2O versus 6 cm H2O is quantitatively identical, the limitations of our arbitrary dichotomy is highlighted when applied to this trial.

A major barrier to the inclusion of studies in this review was the limitations of data from cross-over studies. Generally, the effects of interventions in any preceding cross-over period may influence the measured outcomes in subsequent cross-over periods. Studies involving mechanical ventilation are complicated by the possible recruitment and de-recruitment of preceding ventilation parameters. This was a concerning limitation in the context of our review objectives. As such, we determined to limit outcome data to those resulting from the first cross-over period. Several of the identified studies (Aufricht 1995; Dimitriou 1999; Dinger 1999; Herman 1973) employed a cross-over design in assessing the impact of different levels of PEEP in infants on CMV for RDS. These studies had short time intervals for each cross-over period, raising the concern that interventions following the initial randomization may not have had adequate time to achieve stabilization before the outcomes were measured. Of these studies, only Aufricht 1995 presented their outcome results in such a way that individual data from the first cross-over period could be isolated. Correspondence with Dr Dimitriou confirmed that data of this nature were not available for his study. It is our understanding that nearly 40 years after the completion of the study, the likelihood of retrieving the necessary information for Herman 1973 is sufficiently low to merit exclusion. To date, we have been unable to establish effective correspondence with the authors of Dinger 1999. While we have had to exclude such articles from the analysis at present, this criterion may have limited the inclusion of relevant data in this review.

As implied by the third objective of this review, the optimal level of PEEP may only be achieved by individualization of PEEP therapy. Of interest, Aufricht 1995 attempted to establish criteria for an individualized approach. In short, the study set out to show that the alveolar distension index (a measure of relative pulmonary over or under-distension, derived from characteristics of volume-pressure curves) may be helpful in predicting the effect of PEEP changes in ventilated infants on a case-by-case basis. However, the authors did not perform a randomized evaluation of this strategy. This idea of individualized PEEP finds support in studies dating back over 30 years. In some of the earliest applications of continuous positive airway pressure (CPAP) therapy, Bonta (Bonta 1977) argued that it was possible to determine an 'optimum' individual pressure for CPAP in infants with respiratory distress syndrome (RDS) by monitoring transmitted esophageal pressures and gradually increasing airway pressures until a maximum reduction in required oxygen was achieved. A similar approach was used several years later by Tanswell and colleagues (Tanswell 1980), who identified an 'appropriate' nasal continuous distending pressure as the one at which transpulmonary transmission of airway pressure to the esophagus suddenly increased, associating this level with decreases in oxygen requirements. The study methodology of Aufricht 1995 bears resemblance to the work conducted by Mathe and colleagues (Mathe 1987), in which inspiratory pressure-volume curves were used to determine an individualized PEEP level for newborn infants with RDS. These studies all suggest that considering baseline differences between infants may be critical to the determination of an appropriate PEEP level and that these differences should be considered when evaluating potential heterogeneity within and between studies.

Other studies and reviews

We are not aware of any other trials or reviews that specifically address the objectives of this review.

Authors' conclusions

Implications for practice

The principal result of this study was the finding of a lack of clinical research evidence of effect to address the study's objectives. At this time, we are not able to make any recommendations for clinical practice based on the results of this review.

Implications for research

The findings of this review should raise awareness of the lack of evidence guiding the selection of PEEP levels in preterm infants undergoing CMV for RDS or BPD.

Trials addressing the study objectives are warranted. This includes trials evaluating the effects of different levels of PEEP in infants undergoing CMV for both RDS and BPD. Trials comparing the effect of different strategies for determining an 'optimal', individualized PEEP level based on available patient data and characteristics are also warranted.

Trials evaluating the effects of these interventions should examine outcomes that are important to the patient, such as mortality and neurodevelopmental outcomes, and be aware of the limitations resulting from cross-over trials. In instances in which cross-over trials are performed, researchers should be mindful of the importance of presenting data in such a way that data can be isolated following the end of each cross-over period.

Acknowledgements

We wish to acknowledge Dr Gabriel Dimitriou and Dr George Simbruner for their helpful responses to our inquiries.

Contributions of authors

NB was responsible for data collection for the review, designing search strategies, designing data extraction form, undertaking searches, screening search results, organizing retrieval of papers, screening retrieved papers against inclusion criteria, appraising quality of papers, extracting data from papers, data management for the review, entering review data into RevMan, analysis of data, interpretation of data, editing and writing of the background of the review, primary writing of the review.

DM was responsible for conceiving the review, designing the review, writing the protocol, entering protocol data into RevMan, providing additional data about papers, primary writing of the background to the review, editing, providing general advice on the review.

SS was responsible for data collection for the review, undertaking searches, screening search results, organizing retrieval of papers, screening retrieved papers against inclusion criteria, appraising quality of papers, extracting data from papers, writing to authors of papers for additional information, editing and writing of the review.

HK was responsible for conceiving the review, designing the review, coordinating the review, resolving differences and uncertainties for screening of retrieved papers against inclusion criteria, writing to authors for additional information, providing additional data about papers, editing and writing the review, providing general advice on the review, securing funding for the review, performing previous work that was the foundation of the current review.

Declarations of interest

  • None noted.

Differences between protocol and review

The background of the review has been slightly modified from that of the protocol for improved clarity and organization.

The search strategy was modified to be in accordance with the current guidelines and standard search protocol of the neonatal review group as well as to increase the sensitivity of the search results.

Additional tables

  • Nonte noted.

Potential conflict of interest

  • None noted.

[top]

Characteristics of studies

Characteristics of Included Studies

Aufricht 1995

Methods

Three period cross-over study design. Randomization to intervention sequence via random tables, no blinding of intervention or outcome assessment, follow-up complete.

Participants

13 infants from a single center in Vienna, Austria with RDS diagnosed on the basis of clinical, radiological and blood gas findings. No inclusion or exclusion criteria stated. Infants with a mean gestational age of 30.8 (± 2.6) weeks were randomized at 1.5 (± 0.6) days of life. Participants did not receive surfactant. Sex of infants not specified. Gestational age not provided for individual subjects.

Interventions

Initially randomized to PEEP levels of 2 (N= 4), 4 (N= 6) or 6 (N=3) cm h3O from unspecified baseline. Maintained at PEEP level for 30 minutes prior to being crossed over to next level in random sequence. All infants intubated with uncuffed tubes and mechanically ventilated with Drager Babylog ventilator. Patients were sedated with chloral hydrate, paralyzed with pancuronium and ventilated with controlled ventilation (N=5); sedated with chloral hydrate and hyperventilated to PaCO2 less than 35mmHg with controlled ventilation (N=1); or received no sedation while spontaneously breathing on CMV (N=7).

Outcomes

Outcomes were measured following maintenance at specified PEEP level for 30 minutes. Pre-specified outcomes of interest included PaCO2 obtained via arterial blood gas and AaDO2. Other reported outcomes included pH, compliance and alveolar distention index.

Notes
  1. Data and results provided for each infant after each cross-over period, allowing isolation of data after first cross-over period and inclusion in study.
  2. "Peak pressures were simultaneously altered to keep inflation pressures (peak pressure, PEEP) constant. Other respiratory settings were not changed."
Risk of bias table
Bias Authors' judgement Support for judgement
Random sequence generation (selection bias) Low risk

Sequence determined by random tables.

Allocation concealment (selection bias) Unclear risk

Not addressed.

Blinding (performance bias and detection bias) Unclear risk

Not addressed, but unlikely given nature of intervention and methods.

Incomplete outcome data (attrition bias) Low risk

Follow-up complete.

Selective reporting (reporting bias) Low risk

All reported outcomes provided with complete results.

Other bias High risk

For one infant, PaCO2 provided as measured outcome result despite ventilation technique which deliberately hyperventilated infant to achieve pre-specified PaCO2 goal.

Characteristics of excluded studies

RDS = respiratory distress syndrome; PEEP = positive end expiratory pressure; IMV = intermittent mandatory ventilation; PaCO2 = arterial carbon dioxide pressure; AaDO2 = alveolar to arterial oxygen tension difference.

Belenky 1976

Reason for exclusion

Study did not meet methodological criteria for inclusion. Abstract suggested possibility for a comparison of PEEP (intervention) versus ZEEP (control) in infants receiving CMV for RDS. Detailed review of methods revealed that the control group intervention did not include CMV unless the infant experienced clinical deterioration.

Dimitriou 1999

Reason for exclusion

Study met inclusion criteria but employed a cross-over study design. Outcome data were not presented in such a way as to allow isolation of data obtained after the first cross-over period, as specified for inclusion. Author was contacted and replied that the original data necessary for isolation of the relevant data was no longer available.

Durbin 1976

Reason for exclusion

Abstract suggested possibility for a comparison of PEEP (intervention) versus ZEEP (control) in infants receiving CMV for RDS. Detailed review of methods revealed that infants randomized to the intervention group received either PEEP ("CPAP via an oroendotracheal tube") or continuous inflating pressure via a negative pressure tank, with only those infants deemed "seriously ill" receiving PEEP. This would have created a selection bias, with randomly selected controls compared against selected, "seriously ill, " patients in the intervention group of interest.

Finer 2004

Reason for exclusion

Study did not meet methodological criteria for inclusion. Initial uncertainty resulted from the use of the term "CPAP/PEEP" to describe non-invasive ventilation techniques in the context of delivery room resuscitation.

Hausdorf 1987

Reason for exclusion

Study did not meet methodological criteria for inclusion. Abstract suggested possibility for a comparison of ZEEP versus low PEEP versus high PEEP in infants receiving CMV for RDS. Detailed review of methods revealed that the three different levels were applied in sequence to each patient. Randomization of patients to different PEEP levels did not occur.

Herman 1973

Reason for exclusion

Study methodology unclear. Potentially a six period cross-over design with randomization via sealed envelopes but with no blinding for intervention or outcome assessment. Additional information and data would be required from study authors for consideration of inclusion. Extensive efforts at correspondence with the study authors have been made without success. Given remote date of publication and unavailability of study authors, the likelihood of obtaining this information is exceedingly small.

Simbruner 1989

Reason for exclusion

Report was available in abstract form only. The abstract suggested possibility for a comparison of low PEEP versus high PEEP in infants receiving CMV for RDS. Author was contacted for unpublished materials and replied that it was no longer available.

Stewart 1981

Reason for exclusion

Study did not meet methodological criteria for inclusion. Abstract suggested possibility for a comparison of different PEEP levels in infants receiving CMV for RDS. Detailed review of methods revealed that infants were not randomized to specific PEEP levels, instead undergoing specified alterations in PEEP levels from different individual baselines.

Footnotes

PEEP = positive end expiratory pressure; ZEEP = zero end expiratory pressure; CMV = conventional mechanical ventilation; RDS = respiratory distress syndrome; CPAP = continuous positive airway pressure

Characteristics of studies awaiting classification

Dinger 1999

Methods

Two period cross-over study design. Randomization to initial intervention "randomly" performed without further specification, no blinding of intervention or outcome assessment reported, follow-up complete.

Participants

20 infants from a single center in Dresden, Germany with RDS diagnosed on the basis of clinical and radiological findings. Inclusion criteria: RDS, need for endotracheal intubation, arterial-alveolar oxygen ratio of less than 0.2. No exclusion criteria stated. Infants with a median (range) gestational age of 28 (24-32) weeks were randomized 72 hours following surfactant replacement, which occurred within 2 hours of birth. Sex of infants not specified.

Interventions

Initially randomized to PEEP levels of 0.2 kPa (N= ?) or 0.4 kPa (N= ?) from baseline of 0.3 kPa. Maintained at PEEP level for 20 minutes prior to being crossed over to alternate level. All infants intubated and mechanically ventilated with time cycled, pressure limited ventilation using a Bear BP 2001 ventilator.

Outcomes

Outcomes were measured following maintenance at specified PEEP level for 20 minutes. Pre-specified outcomes of interest included PaCO2 obtained via arterial blood gas and FRC obtained via "a computerized multiple breath washin-washout technique using SF6". Other reported outcomes included pH, PaO2, compliance and specific compliance.

Notes
  1. Outcome data were not presented in such a way as to allow isolation of data obtained after the first cross-over period, as specified for inclusion. All attempts at correspondence with the study authors to obtain necessary data have been unsuccessful to date.
  2. Surfactant treatment with Curosurf, 100mg/kg.
Footnotes

RDS = respiratory distress syndrome; PEEP = positive end expiratory pressure; PaCO2 = arterial carbon dioxide pressure; SF6 = sulfur hexafluoride; PaO2 = arterial oxygen pressure; I:E = inspiratory:expiratory; AaDO2 = alveolar to arterial oxygen tension difference.

Summary of findings tables

  • None noted.

Additional tables

1 Table of comparisons

  1. PEEP versus ZEEP in preterm newborn infants requiring CMV for RDS
    Outcomes:
    1. Death at postnatal age of 28 days or 40 weeks corrected gestational age
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Incidence of BPD
    4. Incidence of air leak syndromes
    5. Duration of ventilatory support
    6. AaDO2.
    7. PaCO2 or transcutaneous CO2 levels
    8. Cardiac output
    9. Blood pressure measurements and requirement for fluid boluses/inotropic support
    10. Incidence of grade III-IV IVH
    11. Functional residual capacity
  2. PEEP versus ZEEP in preterm newborn infants of < 28 w GA requiring CMV for RDS
    Outcomes:
    1. Death at postnatal age of 28 days or 40 weeks corrected gestational age
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Incidence of BPD
    4. Incidence of air leak syndromes
    5. Duration of ventilatory support
    6. AaDO2
    7. PaCO2 or transcutaneous CO2 levels
    8. Cardiac output
    9. Blood pressure measurements and requirement for fluid boluses/inotropic support
    10. Incidence of grade III-IV IVH
    11. Functional residual capacity
  3. PEEP versus ZEEP in preterm newborn infants of 28-31w GA requiring CMV for RDS
    Outcomes:
    1. Death at postnatal age of 28 days or 40 weeks corrected gestational age
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Incidence of BPD
    4. Incidence of air leak syndromes
    5. Duration of ventilatory support
    6. AaDO2
    7. PaCO2 or transcutaneous CO2 levels
    8. Cardiac output
    9. Blood pressure measurements and requirement for fluid boluses/inotropic support
    10. Incidence of grade III-IV IVH
    11. Functional residual capacity
  4. PEEP versus ZEEP in preterm newborn infants of greater than/or equal to 32w GA requiring CMV for RDS
    Outcomes:
    1. Death at postnatal age of 28 days or 40 weeks corrected gestational age
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Incidence of BPD
    4. Incidence of air leak syndromes
    5. Duration of ventilatory support
    6. AaDO2.
    7. PaCO2 or transcutaneous CO2 levels
    8. Cardiac output
    9. Blood pressure measurements and requirement for fluid boluses/inotropic support
    10. Incidence of grade III-IV IVH
    11. Functional residual capacity
  5. PEEP versus ZEEP in preterm newborn infants requiring CMV for RDS, prior to surfactant therapy
    Outcomes:
    1. Death at postnatal age of 28 days or 40 weeks corrected gestational age
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Incidence of BPD
    4. Incidence of air leak syndromes
    5. Duration of ventilatory support
    6. AaDO2.
    7. PaCO2 or transcutaneous CO2 levels
    8. Cardiac output
    9. Blood pressure measurements and requirement for fluid boluses/inotropic support
    10. Incidence of grade III-IV IVH
    11. Functional residual capacity
  6. PEEP versus ZEEP in preterm newborn infants requiring CMV for RDS, post surfactant therapy
    Outcomes:
    1. Death at postnatal age of 28 days or 40 weeks corrected gestational age
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Incidence of BPD
    4. Incidence of air leak syndromes
    5. Duration of ventilatory support
    6. AaDO2.
    7. PaCO2 or transcutaneous CO2 levels
    8. Cardiac output
    9. Blood pressure measurements and requirement for fluid boluses/inotropic support
    10. Incidence of grade III-IV IVH
    11. Functional residual capacity
  7. Low PEEP versus high PEEP in preterm newborn infants requiring CMV for RDS
    Outcomes:
    1. Death at postnatal age of 28 days or 40 weeks corrected gestational age
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Incidence of BPD
    4. Incidence of air leak syndromes
    5. Duration of ventilatory support
    6. AaDO2
    7. PaCO2 or transcutaneous CO2 levels
    8. Cardiac output
    9. Blood pressure measurements and requirement for fluid boluses/inotropic support
    10. Incidence of grade III-IV IVH
    11. Functional residual capacity
  8. Low PEEP versus high PEEP in preterm newborn infants of < 28 w GA requiring CMV for RDS
    Outcomes:
    1. Death at postnatal age of 28 days or 40 weeks corrected gestational age
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Incidence of BPD
    4. Incidence of air leak syndromes
    5. Duration of ventilatory support
    6. AaDO2
    7. PaCO2 or transcutaneous CO2 levels
    8. Cardiac output
    9. Blood pressure measurements and requirement for fluid boluses/inotropic support
    10. Incidence of grade III-IV IVH
    11. Functional residual capacity
  9. Low PEEP versus high PEEP in preterm newborn infants of 28-31 w GA requiring CMV for RDS
    Outcomes:
    1. Death at postnatal age of 28 days or 40 weeks corrected gestational age
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Incidence of BPD
    4. Incidence of air leak syndromes
    5. Duration of ventilatory support
    6. AaDO2
    7. PaCO2 or transcutaneous CO2 levels
    8. Cardiac output
    9. Blood pressure measurements and requirement for fluid boluses/inotropic support
    10. Incidence of grade III-IV IVH
    11. Functional residual capacity
  10. Low PEEP versus high PEEP in preterm newborn infants of greater than/or equal to 32 w GA requiring CMV for RDS
    Outcomes:
    1. Death at postnatal age of 28 days or 40 weeks corrected gestational age
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Incidence of BPD
    4. Incidence of air leak syndromes
    5. Duration of ventilatory support
    6. AaDO2
    7. PaCO2 or transcutaneous CO2 levels
    8. Cardiac output
    9. Blood pressure measurements and requirement for fluid boluses/inotropic support
    10. Incidence of grade III-IV IVH
    11. Functional residual capacity
  11. Low PEEP versus high PEEP in preterm newborn infants requiring CMV for RDS, prior to surfactant therapy
    Outcomes:
    1. Death at postnatal age of 28 days or 40 weeks corrected gestational age
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Incidence of BPD
    4. Incidence of air leak syndromes
    5. Duration of ventilatory support
    6. AaDO2
    7. PaCO2 or transcutaneous CO2 levels
    8. Cardiac output
    9. Blood pressure measurements and requirement for fluid boluses/inotropic support
    10. Incidence of grade III-IV IVH
    11. Functional residual capacity
  12. Low PEEP versus high PEEP in preterm newborn infants requiring CMV for RDS, post-surfactant therapy
    Outcomes:
    1. Death at postnatal age of 28 days or 40 weeks corrected gestational age
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Incidence of BPD
    4. Incidence of air leak syndromes
    5. Duration of ventilatory support
    6. AaDO2.
    7. PaCO2 or transcutaneous CO2 levels
    8. Cardiac output
    9. Blood pressure measurements and requirement for fluid boluses/inotropic support
    10. Incidence of grade III-IV IVH
    11. Functional residual capacity
  13. ZEEP versus PEEP in preterm infants requiring CMV for BPD
    Outcomes:
    1. Death at two years
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Duration of ventilatory support
    4. Incidence of air leak syndromes
    5. AaDO2
    6. Cardiac output
    7. Blood pressure measurements and requirement for fluid boluses/inotropic support
    8. Incidence of grade III-IV IVH
    9. Functional residual capacity
    10. PaCO2 or transcutaneous CO2 levels
  14. Low PEEP versus high PEEP in preterm infants requiring CMV for BPD
    Outcomes:
    1. Death at two years
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Duration of ventilatory support
    4. Incidence of air leak syndromes
    5. AaDO2
    6. Cardiac output
    7. Blood pressure measurements and requirement for fluid boluses/inotropic support
    8. Incidence of grade III-IV IVH
    9. Functional residual capacity
    10. PaCO2 or transcutaneous CO2 levels
  15. Strategy X versus Strategy Y vs Strategy Z (...) for individualizing PEEP to an optimal level in preterm newborn infants requiring CMV for RDS
    Outcomes:
    1. Death at two years
    2. Neurodevelopmental outcome at two years, using Bayley MDI or PDI
    3. Incidence of BPD
    4. Incidence of air leak syndromes
    5. Duration of ventilatory support
    6. AaDO2
    7. PaCO2 or transcutaneous CO2 levels
    8. Incidence of grade III-IVH
    9. Functional residual capacity
    10. Cardiac output
    11. Blood pressure measurements and requirement for fluid boluses/inotropic support

For detailed description and definitions of outcome measures, refer to Types of outcome measures within Methods section.

PEEP = positive end expiratory pressure; ZEEP = zero end expiratory pressure; Low PEEP = PEEP < 5 cm h3O; High PEEP = PEEP greater than/or equal to 5 cm h3O; CMV = conventional mechanical ventilation; RDS = respiratory distress syndrome; MDI = Mental Development Index; PDI = Psychomotor Development Index; BPD = bronchopulmonary dysplasia; PaO2 = arterial partial oxygen pressure; SpO2 = oxygen saturation; PaCO2 = arterial partial carbon dioxide pressure; CO2 = carbon dioxide; FiO2 = fractional inspired oxygen; IVH = intraventricular hemorrhage, AaDO2 = alveolar arterial oxygen tension difference.

[top]

References to studies

Included studies

Aufricht 1995

Aufricht C, Frenzel K, Votava F, Simbruner G. Quasistatic volume-pressure curve to predict the effects of positive end-expiratory pressure on lung mechanics and gas exchange in neonates ventilated for respiratory distress syndrome. American Journal of Perinatology 1995;12(1):67-72.

Excluded studies

Belenky 1976

Belenky DA, Orr RJ, Woodrum DE, Hodson WA. Is continuous transpulmonary pressure better than conventional respiratory management of hyaline membrane disease? A controlled study. Pediatrics 1976;58(6):800-8.

Dimitriou 1999

Published data only (unpublished sought but not used)

Dimitriou G, Greenough A, Laubscher B. Appropriate positive end expiratory pressure level in surfactant-treated preterm infant. European Journal of Pediatrics 1999;158(11):888-91.

Durbin 1976

Durbin GM, Hunter NJ, McIntosh N, Reynolds EO, Wimberley PD. Controlled trial of continuous inflating pressure for hyaline membrane disease. Archives of Disease in Childhood 1976;51(3):163-9.

Finer 2004

Finer NN, Carlo WA, Duara S, Fanaroff AA, Donovan EF, Wright LL, Kandefer S, Poole WK and for the National Institute of Child Health and Human Development Neonatal Research Network. Delivery room continuous positive airway pressure/positive end expiratory pressure in extremely low birth weight infants: a feasibility trial. Pediatrics 2004;114(3):651-7.

Hausdorf 1987

Hausdorf G, Hellwege HH. Influence of positive end-expiratory pressure on cardiac performance in premature infants: a Doppler-echocardiographic study. Critical Care Medicine 1987;15(7):661-4.

Herman 1973

Published data only (unpublished sought but not used)

Herman S, Reynolds EO. Methods for improving oxygenation in infants mechanically ventilated for severe hyaline membrane disease. Archives of Disease in Childhood 1973;48(8):612-7.

Simbruner 1989

Published data only (unpublished sought but not used)

Simbruner G, Popow C, Aufricht C. The effect of PEEP in lung mechanics and blood gases in mechanically ventilated newborn infants. In: Pediatric Research. Vol. 26. 1989:508.

Stewart 1981

Published data only (unpublished sought but not used)

Stewart AR, Finer NN, Peters KL. Effects of alterations of inspiratory and expiratory pressures and inspiratory/expiratory ratios on mean airway pressure, blood gases and intracranial pressure. Pediatrics 1981;67(4):474-81.

Studies awaiting classification

Dinger 1999

Published data only (unpublished sought but not used)

Dinger J, Topfer A, Schaller P, Schwarze R. Effect of positive end expiratory pressure on functional residual capacity and compliance in surfactant treated preterm infants. Journal of Perinatal Medicine 2001;29(2):137-43.

Ongoing studies

  • None noted.

Other references

Additional references

Amato 1998

Amato MB, Barbas CS, Medeiros DM, Magaldi RB, Schettino GP, Lorenzi-Filho G et al. Effect of a protective-ventilation strategy on mortality in the acute respiratory distress syndrome. New England Journal of Medicine 1998;338(6):347-54.

Bonta 1977

Bonta BW, Uauy R, Warshaw JB, Motoyama EK. Determination of optimal continuous positive airway pressure for the treatment of IRDS by measurement of esophageal pressure. Journal of Pediatrics 1977;91(3):449-54.

Boros 1979

Boros SJ. Variations in inspiratory:expiratory ratio and airways pressures wave form during mechanical ventilation: the significance of mean airways pressure. Journal of Pediatrics 1979;94(1):114-7.

Briel 2010

Briel M, Meade M, Mercat A, Brower RG, Talmor D, Walter S et al. Higher vs lower positive end expiratory pressure in patients with acute lung injury and acute respiratory distress syndrome: systematic review and meta-analysis. JAMA 2010;303(9):865-73.

Brower 2004

Brower RG, Lanken PN, MacIntyre N et al National Heart, Lung, and Blood Institute ARDS Clinical Trials Network. Higher versus lower positive end expiratory pressures in patients with the acute respiratory distress syndrome: a randomized controlled trial. New England Journal of Medicine 2004;351(4):327-36.

Corbridge 1990

Corbridge TC, Wood LD, Crawford GP, Chudoba MJ, Yanos J, Sznajder JI. Adverse effects of large tidal volume and low PEEP in canine acid aspiration. American Review of Respiratory Disease 1990;142(2):311-5.

da Silva 1994

da Silva WJ, Abbasi S, Pereira G, Bhutani VK. Role of positive end-expiratory pressure changes on functional residual capacity in surfactant treated preterm infants. Pediatric Pulmonology 1994;18(2):89-92.

Froese 1993

Froese AB, McMulloch PR, Sugiura M, Vaclavik S, Possmayer F, Moller F. Optimizing alveolar expansion prolongs the effectiveness of exogenous surfactant therapy in the adult rabbit. American Review of Respiratory Disease 1993;148(3):569-77.

Gregory 1971

Gregory GA, Kitterman JA, Phibbs RH, Tooley WH, Hamilton WK. Treatment of the idiopathic respiratory distress syndrome with continuous positive airways pressure. New England Journal of Medicine 1971;284(24):1333-40.

Mathe 1987

Mathe JC, Clement A, Chevalier JY, Gaultier C, Costil J. Use of total inspiratory pressure-volume curves for determination of appropriate positive end expiratory pressure in newborns with hyaline membrane disease. Instensive Care Medicine 1987;13(5):332-6.

Meade 2008

Meade MO, Cook DJ, Guyatt GH et al; Lung Open Ventilation Study Investigators. Ventilation strategy using low tidal volumes, recruitment maneuvers, and high positive end expiratory pressure for acute lung injury and acute respiratory distress syndrome: a randomized controlled trial. JAMA 2008;299(6):637-45.

Mercat 2008

Mercat A, Richard JC, Vielle B et al; Expiratory Pressure (Express) Study Group. Positive end-expiratory pressure setting in adults with acute lung injury and acute respiratory distress syndrome: a randomized controlled trial. JAMA 2008;299(6):646-55.

Michna 1999

Michna J, Jobe AH, Ikegami M. Positive end-expiratory pressure preserves surfactant function in preterm lambs. American Journal of Respiratory and Critical Care Medicine 1999;160(2):634-9.

Monkman 2001

Monkman SL, Andersen CC, Nahmias C, Ghaffer H, Bourgeois JM, Roberts RS et al. Positive end expiratory pressure above lower inflection point minimizes influx of activated neutrophils into lung. Critical Care Medicine 2004;32(12):2471-5.

Monkman 2003

Monkman S, Kirpalani H. PEEP -- a "cheap" and effective lung protection. Paediatric Respiratory Reviews 2003;4(1):15-20.

Papile 1978

Papile L, Burstein J, Burstein R, Koffler H. Incidence and evolution of subependymal and intraventricular hemorrhage: a study of infants with birth weights less than 1, 500 gm. Journal of Pediatrics 1978;92(4):529-34.

Reynolds 1971

Reynolds EOR. Effect of alterations in mechanical ventilator settings on pulmonary gas exchange in hyaline membrane disease. Archives of Disease in Childhood 1971;46(246):152-9.

Richardson 1978

Richardson CP, Jung AL. Effects of continuous positive airway pressure on pulmonary function and blood gases of infants with respiratory distress syndrome. Pediatric Research 1978;12(7):771-4.

Richardson 1986

Richardson P, Bose CL, Carlstrom JR. The functional residual capacity of infants with respiratory distress syndrome. Acta Paediatrica Scandinavica 1986;75(2):267-71.

Sandhar 1988

Sandhar BK, Niblett DJ, Argiras EP, Dunhill MS, Sykes MK. Effects of positive end-expiratory pressure on hyaline membrane formation in a rabbit model of the neonatal respiratory distress syndrome. Intensive Care Medicine 1988;14(5):538-46.

Simbruner 1986

Simbruner G. Inadvertent positive end-expiratory pressure in mechanically ventilated newborn infants: detection and. Effect on lung mechanics and gas exchange. Journal of Pediatrics 1986;108(4):589-95.

Tanswell 1980

Tanswell AK, Clubb RA, Smith BT, Boston RW. Individualised continuous distending pressure applied within 6 hours of delivery in infants with respiratory distress syndrome. Archives of Disease in Childhood 1980;55(1):33-9.

Thome 1998

Thome U, Topfer A, Schaller P, Pohlandt F. The effect of positive end expiratory pressure, peak inspiratory pressure and inspiratory time on functional residual capacity in mechanically ventilated preterm infants. European Journal of Pediatrics 1998;157(10):831-7.

Tremblay 2002

Tremblay LN, Miatto D, Hamid Q, Govindarajan A, Slutsky AS. Injurious ventilation induces widespread pulmonary epithelial expression of tumour necrosis factor-alpha and interleukin-6 messenger RNA. Critical Care Medicine 2002;30(8):1693-700.

Other published versions of this review

  • None noted.

Classification pending references

  • None noted.

[top]

Data and analyses

1 G. Low PEEP versus high PEEP in preterm newborn infants requiring CMV for RDS

For graphical representations of the data/results in this table, please use link under "Outcome or Subgroup"

Outcome or Subgroup Studies Participants Statistical Method Effect Estimate
1.1 (6. Alveolar arterial oxygen tension difference (AaDO2) in mmHg) 1 12 Mean Difference (IV, Fixed, 95% CI) -61.89 [-189.05, 65.27]
1.2 (7. Arterial partial carbon dioxide pressure (PaCO2) in mmHg) 1 12 Mean Difference (IV, Fixed, 95% CI) -1.82 [-16.56, 12.92]

2 K. Low PEEP versus high PEEP in preterm newborn infants requiring CMV for RDS, prior to surfactant therapy

For graphical representations of the data/results in this table, please use link under "Outcome or Subgroup".

Outcome or Subgroup Studies Participants Statistical Method Effect Estimate
2.1 (6. Alveolar arterial oxygen tension difference (AaDO2) in mmHg) 1 12 Mean Difference (IV, Fixed, 95% CI) -61.89 [-189.05, 65.27]
2.2 (7. Arterial partial carbon dioxide pressure (PaCO2) in mmHg) 1 12 Mean Difference (IV, Fixed, 95% CI) -1.82 [-16.56, 12.92]

[top]

Figures

Figure 1 (Analysis 1.2)

Refer to figure 1 caption below.

Forest plot of comparison G. Low PEEP versus high PEEP in preterm newborn infants requiring CMV for RDS, outcome 7: Arterial partial carbon dioxide pressure (PaCO2) in mmHg following 30 minutes of randomization to PEEP level (Figure 1 summary).

Figure 2 (Analysis 2.2)

Refer to figure 2 caption below.

Forest plot of comparison K. Low PEEP versus high PEEP in preterm newborn infants requiring CMV for RDS, prior to surfactant therapy, outcome 7: Arterial partial carbon dioxide pressure (PaCO2) in mmHg following 30 minutes of randomization to PEEP level (Figure 2 summary).

Figure 3 (Analysis 1.1)

Refer to figure 3 caption below.

Forest plot of comparison G. Low PEEP versus high PEEP in preterm newborn infants requiring CMV for RDS, outcome 6: Alveolar arterial oxygen tension difference (AaDO2) in mmHg following 30 minutes of randomization to PEEP level (Figure 3 summary).

Figure 4 (Analysis 2.1)

Refer to figure 4 caption below.

Forest plot of comparison K. Low PEEP versus high PEEP in preterm newborn infants requiring CMV for RDS, prior to surfactant therapy, outcome 6. Alveolar to arterial oxygen tension difference (AaDO2) in mmHg following 30 minutes of randomization to PEEP level (Figure 4 summary).

Sources of support

Internal sources

  • No sources of support provided.

External sources

  • Eunice Kennedy Shriver National Institute of Child Health and Human Development National Institutes of Health, Department of Health and Human Services, USA
  • Editorial support of the Cochrane Neonatal Review Group has been funded with Federal funds from the Eunice Kennedy Shriver National Institute of Child Health and Human Development National Institutes of Health, Department of Health and Human Services, USA, under Contract No. HHSN275201100016C.

Feedback

  • None noted.

Appendices

1 Electronic databases search strategies

I. Ovid MEDLINE

Notes: Controlled vocabulary (MeSH) terms were searched as both keywords and subject headings, this is indicated by the "'Term X'.mp. or exp Term X/" format seen below. Text words were searched as keywords, as indicated by the "Term X".mp format seen below.

The search was limited by application of a filter for randomized clinical trials, as seen in item #24.

The search was applied to both the standard database as well as to "In-Process & Other Non-Indexed Citations." The search included studies in these databases through November 2010.

  1. Infant, Premature.mp. or exp Infant, Premature/
  2. Preterm Birth.mp
  3. Preterm Infant.mp
  4. Preterm Neonate.mp
  5. Premature Birth.mp or exp Premature Birth/
  6. Premature Infant.mp
  7. Premature Neonate.mp
  8. Infant, Newborn.mp or exp Infant, Newborn
  9. #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8
  10. Respiratory Distress Syndrome, Newborn.mp or exp Respiratory Distress Syndrome, Newborn/
  11. Hyaline Membrane Disease.mp or exp Hyaline Membrane Disease/
  12. Bronchopulmonary Dysplasia.mp or exp Bronchopulmonary Dysplasia/
  13. Chronic Lung Disease.mp
  14. Respiratory Distress Syndrome.mp
  15. RDS.mp
  16. BPD.mp
  17. #10 OR #11 OR #12 OR #13 OR #14 OR #15 OR #16
  18. Positive Pressure Respiration.mp or exp Positive Pressure Respiration/
  19. Positive Pressure Ventilation
  20. PEEP.mp
  21. Positive End Expiratory Pressure.mp
  22. #18 OR #19 OR #20 OR #21
  23. #9 AND #17 AND #22
  24. limit #23 to "therapy (sensitivity)"

II. CENTRAL

Notes: Studies found within Ovid MEDLINE and EMBASE were excluded from this search to avoid redundancy. The steps necessary to execute this are seen in steps #22 through #25.

  1. Preterm Birth
  2. Preterm Infant
  3. Preterm Neonate
  4. Premature Birth
  5. Premature Infant
  6. Premature Neonate
  7. Newborn Infant
  8. #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7
  9. Respiratory Distress Syndrome
  10. Hyaline Membrane Disease
  11. Bronchopulmonary Dysplasia
  12. Chronic Lung Disease
  13. RDS
  14. BPD
  15. #9 OR #10 OR #11 OR #12 OR #13 OR #14
  16. Positive Pressure Respiration
  17. Positive Pressure Ventilation
  18. Positive End Expiratory Pressure
  19. PEEP
  20. #16 OR #17 OR #18 OR #19
  21. #8 AND #15 AND #20
  22. "accession number" near pubmed
  23. "accession number" near2 embase
  24. #22 OR #23
  25. #21 NOT #24

III. Ovid EMBASE

Notes: Controlled vocabulary was examined and used by mapping terms to EMTREE terminology. Additonally, each term was searched as both "map to preferred terminology" and "search also for synonyms, explosion on preferred terminology". This was found to be the most sensitive search strategy.
It was noted that including multiple-word term with or without quotation resulted in a different number or search results. Furthermore, it was noted that neither including or excluding quotation marks consistently resulted in an more sensitive search result, with some terms having a greater return of results with quotations while others had a greater return of results without quotations. As such, each term was searched both with and without quotations. Ultimately these results were combined. This technique is reflected throughout the search strategy below. Terms searched without quotations were displayed with an "AND" between words.

The search was limited by application of a filter for randomized clinical trials, as seen in item #24.

  1. newborn/syn
  2. prematurity/syn
  3. 'preterm birth'/syn
  4. preterm AND birth/syn
  5. 'preterm infant'/syn
  6. preterm AND infant/syn
  7. 'preterm neonate'/syn
  8. preterm AND neonate/syn
  9. 'premature birth'/syn
  10. premature AND birth/syn
  11. 'premature infant'/syn
  12. premature AND infant/syn
  13. 'premature neonate'/syn
  14. premature AND neonate/syn
  15. #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 OR #11 OR #12 OR #13 OR #14
  16. 'neonatal respiratory distress syndrome'/syn
  17. neonatal AND respiratory AND distress AND syndrome/syn AND
  18. 'hyaline membrane disease'/syn
  19. hyaline AND membrane AND disease/syn
  20. 'lung dysplasia'/syn
  21. lung AND dysplasia/syn
  22. 'bronchopulmonary dysplasia'/syn
  23. bronchopulmonary AND dysplasia/syn
  24. 'chronic lung disease'/syn
  25. chronic AND lung AND disease/syn
  26. 'respiratory distress syndrome'/syn
  27. respiratory AND distress AND syndrome/syn
  28. RDS/syn
  29. BPD/syn
  30. #16 OR #17 OR #18 OR #19 OR #20 OR #21 OR #22 OR #23 OR #24 OR #25 OR #26 OR #27 OR #28 OR #29
  31. 'positive pressure ventilation'/syn
  32. positive AND pressure AND ventilation/syn
  33. 'positive pressure respiration'/syn
  34. positive AND pressure AND respiration/syn
  35. 'positive end expiratory pressure'/syn
  36. positive AND end AND expiratory AND pressure/syn
  37. PEEP/syn
  38. #31 OR #32 OR #33 OR #34 OR #35 OR #36 OR #37 OR #38
  39. #15 AND #30 AND #38
  40. #15 AND #30 AND #38 AND ([controlled clinical trial]/lim OR [randomized controlled trial]/lim

This review is published as a Cochrane review in The Cochrane Library, Issue 1, 2012 (see http://www.thecochranelibrary.com External Web Site Policy for information). Cochrane reviews are regularly updated as new evidence emerges and in response to feedback. The Cochrane Library should be consulted for the most recent recent version of the review.