Lung recruitment manoeuvres in mechanically ventilated children for reducing respiratory morbidity
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
To determine the evidence supporting the use of recruitment manoeuvres in mechanically ventilated children and identify the optimal method of lung recruitment. To ascertain the impact of recruitment manoeuvres on oxygenation, dynamic lung compliance, cytokine release, end expiratory lung volume, length of ventilation, length of stay and mortality.
Background
Critically ill children are commonly intubated and mechanically ventilated. Whilst this therapy is lifesaving, it is not without inherent problems (Dahlem 2003). The term used to describe the negative sequelae attributed to mechanical ventilation is ventilator associated lung injury (VALI), also known as ventilator induced lung injury (VILI). VALI is an established phenomenon which occurs secondary to positive pressure mechanical ventilation (Dyhr 2003; Frank 2002). VALI describes the reduced compliance, deteriorating shunt fraction, inflammatory response and oedema that results from high transpulmonary pressures at end inspiration and inadequate end expiratory lung volume at end expiration (Artigas 1998; Dyhr 2003; Schibler 2006; The ARDS Network 2000). High levels of inspired oxygen also contribute to VALI (Marraro 2005; Theil 2005; Sinclair 2004). VALI is considered to contribute to acute lung injury (ALI) and acute respiratory distress syndrome (ARDS), increased length of ventilation, length of stay and can lead to chronic pulmonary impairment (Villagra 2002).
Lung recruitment is postulated as a means to reduce the incidence of VALI in intubated and mechanically ventilated paediatrics (Duff 2007; Halbertsma 2007; Rimensberger 2000; Villagra 2002). It is proposed that recruitment manoeuvres be a routine adjunct to mechanical ventilation (Arnold 2002; Dyhr 2003; Frank 2002; Villagra 2002). Lung recruitment describes the process in which there is a deliberate elevation of airway pressures in the ventilated patient in order to maximise the number of alveoli participating in gas exchange (Arnold 2002; Dyhr 2003). The rationale behind using recruitment manoeuvres appears sound, particularly when lung protective ventilation strategies (LPVS) are used. LPVS entails low tidal volumes (6 ml/kg), adequate positive end expiratory pressure (PEEP) and clinician tolerance of relative hypoxia and hypercapnoea (Frye 2005; Hanson 2006; Meade 2008; Von Ungern‐Sternberg 2007). LPVS’s minimize high peak inspiratory pressures and inadequate functional residual capacity (FRC) by minimizing tidal volumes and maintaining PEEP at a level to maintain alveolar patency (Dyhr 2003; Von Ungern‐Sternberg 2007).
Chronic de‐recruitment of distal and dependant alveoli is an inevitable outcome when LPVS’s influence ventilation management (Hanson 2006; Hinz 2007; Wolf 2007). Additionally, a rapid, profound and inhomogeneous de‐recruitment of alveoli occurs with each disconnection from the circuit and this is exacerbated by the application of suction (Cunha‐Goncalves 2007; Heinze 2008; Lindgren 2007). Suctioning of the endotracheal tube to extricate secretions occurs regularly and routinely in intubated paediatrics. Recruitment manoeuvres restore end expiratory lung volume by overcoming threshold opening pressures and are most effective when applied after disconnection and suction (Dyhr 2003; Lindgren 2007).
Recruitment manoeuvres aim to optimize the number of alveoli participating in gas exchange and subsequently improve oxygenation (Dyhr 2003; Gattinoni 2008; Maggiore 2003; Wolf 2007). There are a range of methods of recruiting alveoli and a consensus is yet to be achieved as to which is the most effective at reducing respiratory morbidity (Dyhr 2003; Gattinoni 1993; Halter 2007; Hodgson 2009; Lim 2004a; Maggiore 2003). Manipulating PEEP and sustained inflation (SI) techniques have increased end expiratory lung volume (EELV) in both clinical and experimental studies (Barbas 2005; Duff 2007; Povoa 2004, Syring 2007; Tugrul 2003). It is unknown which method of recruitment is the most effective at restoring EELV (Borges 2006; Odenstedt 2005); or whether they improve oxygenation (Rimensberger 1999; Schreiter 2004). It is also unknown if recruitment manoeuvres are suitable for all mechanically ventilated paediatric patients (Borges 2006; Marcus 2002; Morrow 2007; Pelosi 2001; Sargent 2002); or improve outcomes (Duff 2007).
There is the potential that recruitment manoeuvres may result in adverse effects (Claesson 2003; Lim 2004b; Nunes 2004; Toth 2007). By increasing intrathoracic pressure there will be a reduction in cardiac output impacting on perfusion and an increase in intracranial pressure as a consequence of the returning pressure differential (Duff 2007; Graham 2006; Nielsen 2006).
Description of the condition
Ventilator associated lung injury (VALI), also referred to as ventilator induced lung injury (VILI), is an iatrogenic injury acquired as a consequence of positive pressure ventilation. VALI or VILI is characterized by reduced compliance, increased oxygen requirement and an inflammatory response (Mols 2006). Lung recruitment is proposed as a means by which the incidence of VALI may be minimized.
Description of the intervention
Lung recruitment is most commonly achieved by either manipulating end expiratory lung volume with PEEP or end inspiratory lung volume by inspiratory hold: sustained inflation. Lung recruitment is a process that increases airway pressure with the aim to re‐inflate collapsed alveoli and subsequently increasing the total number of alveoli participating in gas exchange.
How the intervention might work
It is suggested that lung recruitment may restore end expiratory lung volume promptly and result in greater alveolar stability. This in turn will minimize the shearing injury to the alveoli associated with cyclic opening and closing and reduce the extent of alveolar endothelial damage.
Why it is important to do this review
It is known that lung recruitment in adults post suctioning is effective (Almgren 2004;Dyhr 2003;Lapinsky 1999;Maggiore 2003).There is no consensus of opinion as to whether lung recruitment in children is appropriate or minimizes the incidence of VALI. There are various methods of recruiting lung, most commonly manipulating PEEP or via a sustained inflation. What is unknown is which of these methods is most appropriate to use in a paediatric population. It is also largely unknown what, if any, adverse effects may be experienced as a consequence of using recruitment manoeuvres.
Objectives
To determine the evidence supporting the use of recruitment manoeuvres in mechanically ventilated children and identify the optimal method of lung recruitment. To ascertain the impact of recruitment manoeuvres on oxygenation, dynamic lung compliance, cytokine release, end expiratory lung volume, length of ventilation, length of stay and mortality.
Methods
Criteria for considering studies for this review
Types of studies
We will include prospective, randomized clinical trials (RCTs) that compare ventilation management with recruitment manoeuvres to ventilation with no recruitment manoeuvres in paediatric patients. We will include all RCTs and quasi‐RCTs (RCTs in which allocation to treatment was obtained by alternation, use of alternate medical records, date of birth or other predictable methods) determining or evaluating the effect of recruitment manoeuvres administered to mechanically ventilated children. We will also include randomized cross‐over studies.
Types of participants
We will include paediatric participants aged from four weeks corrected to 18 years or participants that authors define as paediatrics. Given the developmental difference of paediatrics between these ages, we will conduct subgroup analysis wherever possible in age appropriate groupings. Participants will be intubated and undergoing mechanical ventilation.
In this review mechanical ventilation is defined as any invasive method of positive pressure ventilation via either an endotracheal tube or tracheostomy. We will define the population of paediatrics as per the authors in the individual studies.
We will exclude studies on pre‐term infants (defined as those aged less than 36 completed weeks' gestation) and neonates up to four weeks of age.
We will attempt to retrieve data from adolescents that may be included in adult studies.
Types of interventions
We will define recruitment manoeuvres as a deliberate effort to elevate pulmonary pressures in order to increase the percentage of alveoli participating in alveolar ventilation. We will communicate with authors of studies to determine the precise method of lung recruitment.
Types of outcome measures
Primary outcomes
1. Hospital mortality
2. Intensive care unit (ICU) mortality
3. Length of ICU stay
4. Length of hospital stay
5. Total ventilation hours
Secondary outcomes
We will include studies that measure the following outcomes as measures of the effect of recruitment manoeuvres.
1. Lung compliance
2. Oxygenation
3. Circulating cytokine levels
4. End expiratory lung volume
5. Cardiovascular stability (heart rate, blood pressure, arrhythmia incidence, cardiac output)
Search methods for identification of studies
We will obtain all relevant studies irrespective of language or publication status (published, unpublished, in press, and in progress) using the following methods.
Electronic searches
We will search the current issue of the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library); MEDLINE via OVID (January 1966 to present); EMBASE via OVID (January 1980 to present); CiNAHL via Ebsco Host (1982 to present); LILACS (1982 to present).
We will search the following electronic databases of higher degree theses for relevant unpublished trials: Index to Theses (1950 to date), Australian Digital Theses Program (1997 to date) and Proquest Digital Dissertations (1980 to date).
We will combine our MEDLINE search strategy with the Cochrane Highly Sensitive Search Strategy for identifying the Randomized Controlled Trials (RCTs) as it is suggested in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2008). We will adopt the search strategy for MEDLINE for searching in all other databases. See Appendices (CENTRAL, Appendix 1; MEDLINE, Appendix 2; EMBASE, Appendix 3; CINAHL, Appendix 4; LILACS, Appendix 5).
Searching other resources
We will handsearch citations.
We will not exclude studies on the basis of language.
We will contact authors known in the field to determine if unpublished work is available.
Data collection and analysis
Selection of studies
Six authors (JJC, AS, FB, KG, CG and CE) will undertake the review. We will use the search strategy described to obtain titles and abstracts of studies that may be relevant to the review. We (JJC and CG) will independently screen all titles and abstracts. We will discard studies that are not applicable, although initially we will retain studies and reviews that might include relevant data or information on trials. We (JJC, CG) will independently assess retrieved abstracts, and if necessary the full text of these studies, to determine which studies satisfy the inclusion criteria. We will describe our reasons for excluding studies table of excluded studies. We will resolve disagreement by discussion.
Data extraction and management
We will adapt the standardized Cochrane Anasthesia Review Group (CARG) data extraction form (Appendix 6) to capture relevant data specific to this review. We (JJC and CG) will use this form to extract data from relevant studies. We (JJC and CG) will independently perform data extraction and quality assessment of eligible trials. We will pilot the standardized form using a representative sample of trials to ensure consistency of reporting between the authors. We will revise the tools if we find inconsistencies. We will translate studies reported in non‐English language journals before assessment. Where more than one publication of one trial exists, we will only include the publication with the most complete data. Where relevant outcomes are only published in earlier versions, we will use this data. We will highlight any discrepancy between published versions. We will request any further information required from the original author by written correspondence and we will include any relevant information obtained in this manner in the review. We will resolve disagreements by consensus in consultation with CE.
Assessment of risk of bias in included studies
We will appraise the methodological quality of each trial and will include assessment of bias (selection, performance, detection and attrition). We will grade the method of treatment allocation and concealment of the allocation by using the GRADE approach as recommended by the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2008). We will assess the levels of quality of a body of evidence using the GRADE approach. We will assess other aspects of methodological quality using a standardized checklist with each individual component recorded as yes, no or unclear. The primary author will enter the data into the Review Manager Software (RevMan 5.0) with verification of data entry conducted independently. For each study we will construct a risk of bias graph and risk of bias summary figure from the risk of bias table.
Measures of treatment effect
We will summarize trials that meet the inclusion criteria in tables to enable comparison of trial characteristics and individual components of the quality assessment. We will tabulate the bibliographic details of trials excluded from the review with the reason for exclusion documented. We will discuss the level of agreement between review authors during the screening, data extraction and critical appraisal process in a narrative form. We will review the summary tables of included trials to identify substantial clinical heterogeneity amongst trials. If there are two or more randomized trials with comparable populations undergoing similar interventions, we will implement a meta‐analysis with a random‐effects model using RevMan 5.0 software. If there is clear evidence of heterogeneity among trials or their populations, we will undertake a narrative summary of the findings.
We will quantitatively analyse outcomes from comparable trials to estimate each trial's treatment effect with 95% confidence intervals (CI). We will compare the results graphically within forest plots with risk ratio (RR) as the point estimate for dichotomous outcomes and mean difference (MD) for continuous outcomes. We will calculate standardized mean difference (SMD) if different scales are used to measure continuous outcomes across trials. We will conduct a meta‐analysis using RevMan 5.0 of pooled data to provide a summary statistic of effect if the combined data has minimal statistical heterogeneity (Sutton 2008).
Unit of analysis issues
We will conduct a sensitivity analysis on data pooled within a meta‐analysis. We will analyse individual components of the standardized quality assessment separately to examine their impact on the review's findings. It is not feasible to blind health professionals providing the lung recruitment; therefore, we will not subject participant and caregiver blinding to sensitivity analysis. We will compare the results with or without trials by addressing adequate randomization, adequate concealed allocation, outcome assessor blinding, standard management and co‐interventions applied equally across groups, and loss to follow up of less than 20% with an intention‐to‐treat analysis. We will undertake a sensitivity analysis based on the choice of summary statistic and on the presence of outlying trials. In addition, we are aware that requests for missing data from trial authors may or may not be successful. We will consider assessment for publication bias through funnel plots if there are more than 10 included trials. A large number of trials are required to provide moderate power for detection of publication bias (Higgins 2002).
We will include crossover trials and cluster randomized trials in the review. We will consider the wash‐out period in each trial in determining whether any carryover effect is possible on subsequent measurements (Higgins 2008). We will also confirm that the order of treatments have been randomized (Higgins 2008). We will attempt to access paired and unpaired data (Higgins 2008). We will consider crossover studies only in reference to secondary outcomes.
Dealing with missing data
In the first instance we will contact the study authors to source missing data. If the study author either does not respond, or it is not possible to find them, then we will include the trial in question in the review but will analyse its inclusion and exclusion for overall effect on the results as part of the sensitivity analysis.
Assessment of heterogeneity
We will analyse heterogeneity using a Chi2 test on N‐1 degrees of freedom, with an alpha of 0.1 used for statistical significance and with the I² test (Higgins 2002). I² values of 25%, 50% and 75% correspond to low, medium and high levels of heterogeneity.
We will test for homogeneity between trials for each outcome using the Cochran's Q statistic with P less than or equal to 0.10. We will formally test for the impact of heterogeneity by using the I2 test (Higgins 2002). We will set an I2 threshold of greater than 50% to indicate that variation across trials due to heterogeneity is substantial. We will examine possible sources of substantial heterogeneity through a summary of trial characteristics and quality. We will use a fixed‐effect model if we find insignificant heterogeneity between trials. We will use a random‐effects model if significant heterogeneity exists among trials (Higgins 2008).
Clinical heterogeneity may exist due to the nature of the inclusion criteria. Positive pressure breaths may alter the effects of lung recruitment compared to spontaneous, pressure supported breathing. Therefore, we will undertake subgroup analysis to examine possible clinical variability when the I2 statistic is less than 50% but heterogeneity remains statistically significant. We will analyse outcome data from trial populations rather than individuals to explain possible sources of variability.
We will examine differences in populations based on:
age (corrected);
disease status ‐ in particular pulmonary pathology;
gestation.
Assessment of reporting biases
We will assess publication bias or a systematic difference between smaller and larger studies (small study effects) by preparing a funnel plot, assuming we source at least 10 studies.
Data synthesis
We will tabulate studies that meet the inclusion criteria to enable comparison of trial characteristics and individual components of the quality assessment. We will also tabulate the bibliographic details of trials excluded from the review with the reason for exclusion documented.
We will review the summary tables of included trials to identify clinical heterogeneity amongst trials. If there are two or more randomized trials with comparable populations undergoing similar interventions, we will implement a meta‐analysis with a random‐effects model using RevMan 5.0 software. If there is clear evidence of heterogeneity among trials, we will undertake a narrative summary of the findings (Sutton 2008).
We will quantitatively analyse outcomes from comparable trials to estimate each trial's treatment effect with 95% confidence intervals (CI). We will compare the results graphically within forest plots with relative risk (RR) as the point estimate for dichotomous outcomes and mean difference (MD) for continuous outcomes. We will calculate standardized mean difference (SMD) if different scales are used to measure continuous outcomes across trials. We will conduct a meta‐analysis of pooled data to provide a summary statistic of effect if the combined data has minimal statistical heterogeneity(Higgins 2008).
Subgroup analysis and investigation of heterogeneity
We will use subgroup analysis to explore possible sources of heterogeneity (e.g. participants, interventions). Heterogeneity among participants could be related to age, gestational age, lung pathophysiology and pre‐existing lung disease. Heterogeneity in treatments could be related to mode and length of ventilation. We will determine heterogeneity in recruitment techniques, timing and frequency of recruitment via communication with authors where necessary. We will also explore the impact of differing modes of ventilation and recruitment methods with a subgroup analysis. We will tabulate and assess adverse effects with descriptive techniques, as they are likely to be different for the various subgroups. Where possible, we will calculate the risk ratio with 95% CI for each adverse effect, either compared to no treatment or to a different method of lung recruitment.
Sensitivity analysis
If there are an adequate number of studies, we will perform a sensitivity analysis to explore the causes of heterogeneity and the robustness of the results. We will include the following factors in the sensitivity analysis, separating studies according to: quality of allocation concealment (adequate or unclear or inadequate); blinding (adequate or unclear or inadequate or not performed); analysis using both random‐effects or fixed‐effect models; intention‐to‐treat analysis and available case analysis (only for dichotomous data) (Higgins 2008).
Summary of Findings
We will use the principles of the GRADE system (Guyatt 2008) to assess the quality of the body of evidence associated with specific outcomes ‐ hospital mortality, ICU mortality, length of ICU stay, total ventilation hours, oxygenation, lung compliance and end expiratory lung volume in our review and construct a Summary of Findings (SoF) table using the GRADE software. The GRADE approach appraises the quality of a body of evidence based on the extent to which one can be confident that an estimate of effect or association reflects the item being assessed. The quality of a body of evidence for considers within the study risk of bias (methodologic quality), the directness of the evidence, heterogeneity of the data, precision of effect estimates and risk of publication bias.
