Approximately two thirds of all infants admitted to neonatal intensive care nurseries require intermittent positive pressure ventilation (ANZNN 2000). The majority are ventilated because of lung immaturity and hyaline membrane disease. Although the respiratory difficulties resolve in most of these infants, a significant proportion (up to 20% of all infants of less than 32 weeks' gestation (ANZNN 2000) and 30% of those of birth weight less than 1 kg (Jobe 2001a)) develop bronchopulmonary dysplasia (BPD) with oxygen dependency at 36 weeks' corrected age. The resulting burden of illness includes increased duration of respiratory support and hospital stay, the need for home oxygen, more readmissions to hospital and greater mortality rate. As a result, much research has been directed at reducing rates of BPD.
Bronchopulmonary dysplasia is a clinically defined entity based on the requirement for supplemental oxygen at either 28 days postnatal age or at 36 weeks corrected age (Jobe 2001a; Jobe 2001b). It is characterized by the histopathological findings of impaired alveolarization, altered pulmonary microvasculature and pulmonary fibrosis. The development of BPD has been linked to lung immaturity, intrauterine growth retardation (Bardin 1997; Gortner 1999), infection (Hannaford 1999), oxidant stress (Warner 1998), ventilator-induced lung injury (VILI) (Coalson 1999; Clark 2000) and in-utero inflammation (Watterberg 1996). VILI has been targeted as a potentially preventable cause of bronchopulmonary dysplasia, and much research has been devoted to developing ventilation strategies to try to avoid the overdistension, atelectasis and shear stresses that are thought to lead to inflammation and consequently BPD. The fact that injury has been demonstrated as early as the first few resuscitative breaths highlights the potential importance of early use of protective ventilation strategies in the neonate.
In assisted ventilation, the ventilator performs two functions. Firstly, it maintains oxygenation by providing a mean airway pressure and a suitable inspired oxygen concentration. Secondly, it removes carbon dioxide by ventilating the lung with a "tidal volume" that it moves in and out of the lung with each breath. The clinician sets the tidal volume (VT) either directly, in volume-targeted ventilation, or indirectly, in pressure-limited ventilation, by setting the airway pressure difference (the peak inspiratory pressure [PIP] minus positive end-expiratory pressure or [PEEP]).
There has been concern about the effects of high-pressure strategies on the respiratory system ("barotrauma"); however, recent attention has focused on lung collapse and overdistension (or atelectasis and "volutrauma") as the major instigators of inflammation in the preterm lung. This is supported by animal studies of two kinds of "high-pressure" ventilation models, one where a high peak inspiratory pressure was set to deliver large tidal volumes, and a second where the chest was enclosed in a cast, so despite the same high PIPs the tidal volume remained low (Hernandez 1989). Postmortem examination revealed a significant decrease in lung inflammation in the low tidal volume group despite the high pressure. Further support comes from a randomised controlled trial comparing high tidal volume ventilation (12 ml/kg) with low volume (6 ml/kg) strategies in adults with acute lung injury, which was stopped prematurely when interim analysis revealed a significant reduction in both mortality and duration of ventilation in the latter group (ARDS Network 2000).
The neonate is theoretically at much greater risk of volutrauma, as the chest wall is more compliant than that of the older child or adult, providing less protection against lung overdistension (Clark 2000). The variable lung compliance seen in the infant with hyaline membrane disease provides a particular challenge, as it can change rapidly after surfactant administration or with spontaneous resolution of the RDS (Goldsmith 1996). In addition, an infant's spontaneous inspiratory efforts, when combined with ventilator inflations, may result in excessive tidal volumes that can also exacerbate lung injury.
Traditionally, neonatologists treating term and preterm infants have employed continuous flow, time-cycled, pressure-limited ventilation (TCPL). In this mode, the assistance provided by the ventilator is controlled in two ways. The duration of the breath is preset by the clinician as the inspiratory time (hence "time-cycled" (Goldsmith 1996)). The magnitude of each inflation is determined by the change in airway pressure, i.e. the difference between PIP and the baseline or positive end-expiratory pressure (PEEP). The tidal volume for any breath depends on both this pressure difference, which drives gas movement, and the lung compliance (or distensibility). Although tidal volume is indirectly determined by the clinician when the PIP and PEEP are set, VT may not be consistent where compliance changes. For example, administration of artificial surfactant may lead to a rapid improvement in compliance and delivery of excessive tidal volumes if the PIP is not reduced. Similarly, when the infant breathes spontaneously after a period of apnea the combination of the infant's efforts and the ventilator breath may lead to overinflation and volutrauma.
These problems led to the development of volume-targeted ventilation strategies, which try to deliver a consistent tidal volume. These modes have been in use in pediatric and adult practice for many years; however, technological limitations in older ventilators precluded their use in the preterm neonate because they were unable to deliver accurately the small tidal volumes required when ventilating small preterm infants. This was because the compressible gas volume in the ventilator circuit was much larger than the lung volume and the uncuffed endotracheal tubes often lost an unknown amount of the inflation. With modern microprocessor controlled neonatal ventilators with very sensitive and accurate flow sensors that can be placed at the Wye piece between the ventilator circuit and the endotracheal tube it is now possible to measure and control the tidal volume. Although little information exists regarding the optimal VT for preterm infants, typical VT lies between 4 and 6 ml/kg. Volume-targeted modes of ventilation have been shown to be feasible in the NICU (Cheema 2001). They reduce variability in tidal volume delivery and provide ventilation at lower mean airway pressures than traditional time-cycled pressure-limited modes in short-term crossover studies while maintaining adequate ventilation (Herrera 2002; Abubakar 2001).
Volume-targeted ventilators allow the clinician to set the tidal volume directly (set VT). The ventilator's internal computer makes adjustments to peak inspiratory pressure or inflation times from inflation to inflation to try to deliver that volume. Different systems measure either expired VT, which allows for leaks around the endotracheal tube, or inspired VT which does not take leaks into account. There are different forms of volume-targeted ventilation.
Volume Controlled (or Volume Support) ventilation
Here, the clinician selects a desired tidal volume. The duration of inflation then depends on the time it takes for that volume to be delivered (i.e. volume-cycled, not time-cycled), and inflation ends as soon as that volume has been delivered. If the volume is not delivered in the preset inspiratory time the volume may not be delivered. These are not continuous flow ventilators and so a spontaneously breathing baby may not be able to inspire easily during ventilator expiration. The rate of flow, PIP and inspiratory time may all vary from breath to breath. The only constant is the tidal volume delivered into the ventilator circuit. This mode of ventilation is available on the VIP Bird, and a similar "volume support" mode is available on the Siemens Servo 300 ventilator. The tidal volume is measured at the ventilator rather than at the endotracheal tube, so the inspiratory tidal volume which is set must take into account the dead space of the humidifier and ventilator tubing, the circuit compliance and gas compression in the circuit. This mode is designed to deliver an inspiratory volume and so does not compensate for any leak at the endotracheal tube. It is possible to calculate the effects of the above parameters for any given patient (Goldsmith 1996) but it has been recommended that the set tidal volume at the ventilator be a minimum 25-50% greater than the desired VT for delivery to the infant (Sinha 2000). A modified version of volume controlled ventilation, known as pressure-regulated volume control ventilation (PRVC), combines the features of volume control with an upper limit of PIP, and is also available on the Siemens Servo 300 ventilator (Goldsmith 1996).
Volume Guarantee Ventilation
The second form of volume-targeted ventilation is known as volume guarantee and is available on the Draeger Babylog 8000+ ventilator. It is a form of time-cycled, pressure-limited ventilation (i.e. preset inspiratory time determines duration of inflation and maximum pressure set by clinician limits maximum PIP), where a preset expiratory tidal volume is selected. It uses a flow sensor placed at the endotracheal tube to measure the inspired and expired tidal volume (VTe) of each breath (it analyses expiratory rather than inspiratory flow to account for possible leaks). The operator sets a target expiratory tidal volume. The ventilator then adjusts the peak inflating pressure to try and maintain the set expired tidal volume for the next inflation. Where the measured tidal volume is less than the value selected by the clinician, it makes small stepwise increases in PIP from breath to breath (<3 cm H2O) until the set VTe is reached. If the measured volume exceeds that set, it reduces the inflating pressure in similar decrements. The ventilator adjusts PIP to accommodate changes in endotracheal leak (up to about 60%), lung compliance, airway resistance and the infant's spontaneous inspirations. The ventilator ensures that for each inflation the tidal volume delivered to the baby is close to that set by the operator. It thus provides a volume-targeted strategy, within the limits of time-cycled, pressure-limited ventilation. Because the flow measurement occurs at the endotracheal tube, the circuit compliance does not affect the delivered volume.
Volume-controlled and volume guarantee ventilation are distinct entities, despite being linked by a strategy of presetting VT. The differences in flow patterns, inspiratory times and degree of limitation of PIP may interact with the spontaneously breathing infant in very different ways.
Other types of neonatal volume ventilation
Some other ventilators provide a time-cycled, pressure-limited "volume-limited" strategy (Bearcub 750 PSV, SLE 5000), where the pressure support for any inflation is aborted if the measured inspired tidal volume (VTi) exceeds a preset upper limit. Both inspired and expired tidal volumes can be displayed, but inflating pressures are not automatically adjusted if VTi falls to less than the pre-selected value (Sinha 2000). As this mode does not provide assurance of tidal volume delivery, it will be excluded from this review. Where additional methods of volume-targeted ventilation are developed or discovered in the literature they were assessed for suitability for inclusion in this review.
Characteristics of examples of volume-controlled, volume-guarantee and volume-limited ventilators are shown in Additional Table 01.
Neuromuscular paralysis of mechanically ventilated infants may help to deliver more consistent tidal volumes, as irregular infant inspiratory efforts are eliminated. There is no evidence from randomised trials that the use of pancuronium in preterm infants with hyaline membrane disease reduces mortality or the incidence of bronchopulmonary dysplasia (Cools 2002). There was, however, a benefit of neuromuscular paralysis where asynchronous breathing was demonstrated, reducing rates of both intraventricular haemorrhage and air leak. Although paralysis eliminates the variability generated by spontaneous breathing, the issue of variability of lung compliance persists, and volume-targeted modes have a theoretical role in the paralysed ventilated infant. In addition, the clinical decision to paralyse a ventilated infant may reflect the failure of a particular ventilation strategy to effectively stabilise the patient.
Although most infants requiring intermittent positive pressure ventilation admitted to intensive care units are preterm babies with surfactant deficiency, substantial numbers of more mature infants require ventilation. They have different underlying pulmonary or neurological pathological processes and may respond differently to different modes of ventilation. Although bronchopulmonary dysplasia is less common in these infants, ventilator-induced lung injury is a potential problem in all age groups; protective strategies of mechanical ventilation may therefore have a role in all neonates being ventilated.
This review investigated all infants of less than 28 days corrected age who were ventilated with intermittent positive pressure ventilation. The primary objectives were to determine:
Whether the use of volume-targeted (VT) ventilation strategies compared with time-cycled pressure-limited ventilation reduced rates of mortality and bronchopulmonary dysplasia.
In addition these strategies were compared to determine any differences in the incidences of complications of prematurity and adverse neurological outcomes.
Subgroup analyses
Three subgroup analyses were planned based on:
1. mode of volume-targeted ventilation
2. age at recruitment into study
3. maturity / birth weight of the infants
1. In view of the differences between the two modes of volume-targeted ventilation, subgroups were defined according to:
i) Volume-controlled ventilation
ii) Volume guarantee ventilation
2. Subgroups were defined according to postnatal age at time of recruitment into studies:
i) Early recruitment, i.e. commencement of ventilation strategy at birth or within the first four hours of life
ii) Late recruitment, i.e. beyond four hours of age. This subgroup included
trials in which volume ventilation was tested as a rescue strategy
3. In view of the increased risk of BPD in the smallest, most immature infants, subgroups were defined according to:
i) Birth weight, with a cut-off of 1000g
ii) Gestational age, with a cut-off of 30 weeks' gestation
Randomised and quasi-randomised studies were eligible for inclusion in this review.
All intubated infants of less than 28 days corrected age who were being mechanically ventilated with intermittent positive pressure ventilation at the time of study entry. Infants of all gestational ages and both paralysed and non-paralysed infants were eligible.
Volume-targeted versus time-cycled, pressure-limited modes of mechanical ventilation. The review only included studies looking at volume-targeted ventilators that assured tidal volume delivery.
The two primary outcomes were death, and death or requirement for supplemental oxygen at defined time points (as below).
The secondary objectives of this review were to compare each volume-targeted method of ventilation with time-cycled pressure-limited ventilation with respect to:
a) failure of mode of ventilation (clinical decision to change to different mode of ventilation)
b) addition of neuromuscular paralysis where previously not paralysed
c) ventilation data
f ) patent ductus arteriosus
g) incidence of air leak
k) Surviving infants with bronchopulmonary dysplasia
The search was performed using the standard strategy of the Neonatal Review Group of the Cochrane Collaboration. MEDLINE (1966-Oct 2004) was searched using the MeSH terms: infant, newborn and respiration, artificial and the text word: volume. These terms were also used in a search of the Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 3, 2004) and CINAHL. No language restrictions were applied. A review (1981-2004) of abstracts published by the Society for Pediatric Research and the European Society for Pediatric Research completed the literature search. This was combined with cross-referencing of previous reviews and the use of expert informants.
The standard methods of the Neonatal Review Group of the Cochrane Collaboration were used. Trial searches, assessments of methodology and extraction of data were performed independently by two review authors with comparison and resolution of any differences found at each stage. Review authors based quality assessment on 1) blinding of randomisation, 2) blinding of intervention, 3) blinding of outcome measurements and 4) completeness of follow-up.
For categorical data (e.g. number dying or developing bronchopulmonary dysplasia) the relative risk (RR), risk difference (RD) and number needed to treat (NNT) with 95% confidence intervals were calculated. Continuous data (e.g. number of ventilator days, or duration of oxygen dependency) were analysed using weighted mean difference (WMD). The fixed effect model was used. An evaluation of heterogeneity was conducted using the I squared statistic to determine the suitability of pooling results.
A sensitivity analysis limited to true randomised trials only was planned if quasi-randomised trials were identified during the literature search.
For the summary of the included studies see also table of included studies.
We found eight randomised trials which compared time cycled, pressure limited ventilation with a volume-targeted mode (Lista 2004; Piotrowski 1997; Sinha 1997; Keszler 2004; Abubakar 2001; Cheema 2001; Herrera 2002; Olsen 2002). Three of these were crossover trials that did not report the outcomes specified in our protocol and were excluded (Herrera 2002; Cheema 2001; Abubakar 2001) for various reasons. There were methodological concerns about the trial by Cheema 2001 (see Table of Excluded Studies). Abubakar 2001 did not randomise the first ventilator of the crossover study and the same order was used for each baby, so there may have been an element of fatigue in the findings. The trial by Herrera 2002 did not address any of the specified outcomes of this review, and so was not included in the meta-analysis. A trial by Olsen 2002 was excluded because it compared different modes of triggered ventilation as well as the use of either volume or time-cycled modes. The remaining four randomised controlled trials recruiting 178 infants met our inclusion criteria (Keszler 2004; Lista 2004; Piotrowski 1997; Sinha 1997).
Study populations:
Each trial examined preterm neonates (defined as less than or equal to 37 weeks' gestation), except Keszler 2004 (who looked at infants < 34 weeks). Piotrowski 1997 (weight < 2500 g) and Lista 2004 (gestation <32 weeks) examined cohorts that were potentially smaller and more premature than the study by Sinha 1997. Only Piotrowski 1997 presented their results stratified for birthweight. Three of the four studies specified a diagnosis of respiratory distress syndrome (clinically, biochemically and radiologically) as a prerequisite for inclusion. The other (Piotrowski 1997) included infants ventilated for any reason. Exclusion criteria were similar across the trials and included the following: lethal congenital anomalies, muscle relaxation, suspected sepsis, lack of arterial access, narcotics, endotracheal tube leaks (> 30%) and severe IVH, asphyxia, pneumothorax and meconium aspiration.
Three studies recruited infants within the first 24 hours of life and the fourth (Piotrowski 1997) by 72 hours. Lista 2004 randomised infants by a mean of three hours of age and Keszler 2004 by 2.3 hours. Sinha 1997 randomised infants by a median age of eight hours in the VC group and five hours in the TCPL arm, and Piotrowski 1997 by a mean age of 12.1 hours (pressure limited) and 15.6 hours (volume controlled).
Antenatal steroids and surfactant were available in all participating units and there were no significant differences in their use within individual studies.
Interventions:
Lista 2004 and Keszler 2004 used the volume guarantee mode on the Draeger Babylog 8000+ as the volume-targeting strategy. Lista 2004 used pressure support ventilation and Keszler 2004 assist control ventilation in both arms. Piotrowski 1997
investigated pressure regulated volume controlled (PRVC) ventilation as delivered
by the Siemens Servo 300 and compared it with intermittent mandatory ventilation
(IMV) using the Bear Medical Systems Cub Model 2001 and Sechrist IV 100B
ventilators. Sinha 1997 studied the volume control modality on the VIP Bird and used the assist control mode in both arms.
The ventilation settings were well defined in each study, and infants were weaned according to unit protocol. Volume-targeted ventilation operates with a maximum pressure limit, usually set slightly above that on pressure-limited ventilation, to allow accurate tidal volume delivery with variable lung compliance. This is particularly an issue with the volume guarantee (VG) mode on the Draeger Babylog. One of the studies using this ventilator specified the increase in PIP made for the transition to VG (Keszler 2004) but one did not (Lista 2004).
The use of triggering in one arm of the study but not the other is a potential source of bias. In the study by Piotrowski 1997 the control arm ventilation mode was intermittent mandatory ventilation, rather than a triggered mode that was used with the PRVC arm of that study. The converse was true for the study by Sinha 1997 where the experimental volume control (VC) modality was non-triggered but a triggered ventilation mode was used in the control group. This led to a switch from VC to a time-cycled pressure-limited ventilation strategy for weaning purposes (once the set rate was < 40), and so both groups were ventilated with a TCPL mode. In the other two randomised controlled trials (Lista 2004; Keszler 2004) triggered ventilation was used for both groups of subjects and the only difference was the addition of the volume guarantee setting in the experimental arms.
Major outcomes:
All four included studies reported death during hospitalisation.
BPD defined as requirement for supplemental oxygen at 28 days of age was
reported by Lista 2004 and Piotrowski 1997. Sinha 1997 and Lista 2004
reported BPD as defined by supplemental oxygen requirement at 36 weeks. The
combined outcome of death or BPD was not reported by any study. No outcome
data on growth or long-term neurodevelopment were reported.
Methodological quality was described using the standard method for conducting a systematic review as described in the Cochrane Collaboration Handbook. Additional details of each study appear in the table of included studies.
Method of Study Allocation.
The four included studies were randomised controlled trials. Sinha 1997 used block randomisation and sealed envelopes for group assignment. Sealed envelopes and random number generation were used in the studies by Keszler 2004 and Piotrowski 1997. Lista 2004 describes random number generator use to allocate grouping but not the method of concealment.
Masking of Caregivers.
Because of the nature of neonatal ventilation and the need for adjustments to ventilatory support to maintain adequate oxygenation and normocarbia, none of the studies included in this review attempted to mask the caregivers to the group assignment.
Completeness of Study Outcome.
Piotrowski 1997 excluded three infants (of 60) after randomisation; two who didn't fulfil enrolment criteria and one for whom the allocated ventilator was unavailable. Outcome assessment was otherwise described as complete. In the study by Lista 2004 there was an uneven distribution between the volume targeted and non-volume targeted groups which, if allocation was even, may reflect post-randomisation deletions not documented in the reported results.
Masking of Outcome assessors.
In three of the studies, the allocated treatment method of each patient was known to those assessing the trial outcomes. In the other, Sinha 1997, severity of lung disease was assessed by a radiographer blinded to the treatment assignment.
Four studies were identified which fulfilled our entry criteria. They recruited a total of 178 neonates. There was no disagreement regarding inclusion/exclusion of studies, quality assessment or data extraction. Available data were pooled and analysed as listed below. Where possible, subgroup analysis according to the type of volume targeting (volume guarantee vs volume control) was performed. A planned subgroup analysis based on age at enrolment was not possible, as the results of individual studies were not presented stratified for this variable.
There was no evidence of substantial heterogeneity in any of the pooled analyses i.e. I squared values were all < 40%.
There are no major concerns about the methodology used in the four trials included in this review, although there is imbalance in numbers between the control and volume targeted groups in Lista 2004 which may suggest post-randomisation loss to follow-up. It would be difficult, if not impossible, to blind caregivers to the allocated treatment and this could have affected various outcomes (e.g. clinician awareness of allocation might affect weaning and hence total duration of ventilation).
There were no significant differences in mortality between volume targeted and pressure limited groups. Both the primary outcome of combined death and BPD and the secondary outcome of long term neurodevelopmental outcome were not addressed. However, this review found a some clinically important benefits of volume targeting. These included significant reductions in duration of intermittent positive pressure ventilation and rates of pneumothorax and severe intraventricular haemorrhage. In spite of the variety of ventilators and ventilation strategies used in the four included studies we found no evidence of substantial heterogeneity in any of the pooled analyses. However, the possibility exists that different types of volume targeted strategies may have different safety and efficacy profiles.
There were no significant differences between volume targeted and TCPL groups in rates of surfactant administration or antenatal steroid use. The fact that some studies varied in the use of triggering between volume targeted and non-volume ventilation targeted strategies could have had an impact on outcomes such as duration of ventilation and airleak (Greenough 2004).
We were unable to perform a planned subgroup analysis of studies based on time of initiation of the allocated ventilator strategy. It is, therefore, unclear whether the benefits seen in infants allocated to volume targeting are affected by the time of starting this strategy. Overall the small numbers of infants randomised and the resultant low power and wide confidence intervals may have affected the findings. Subgroup analysis of the tiniest (< 1 kg) babies was limited by a paucity of data. Further research is required, as all neonates are susceptible to adverse pulmonary and neurological outcomes, especially the smallest and most vulnerable.
There is a sound theoretical basis for the use of volume-targeted ventilation strategies in the neonate, and this review did not identify any adverse outcomes associated with their use compared with conventional time-cycled, pressure-limited ventilation. However, this review failed to show any benefits in the clinically significant long-term outcomes of death and neurodevelopmental impairment. Volume-targeted ventilation resulted in significant reductions in rates of intraventricular haemorrhage and pneumothorax and in the number of days of IPPV required. Small numbers of studies and infants enrolled in trials to date mean that caution should be exercised in the widespread application of volume targeting in neonatal intensive care.
Further randomised controlled trials, powered to assess effects on important outcomes such as death, chronic lung disease or neurodevelopmental disability, are required to determine whether volume-targeted ventilation should routinely be used in preference to conventional pressure-limited modes in neonates.
Study | Methods | Participants | Interventions | Outcomes | Notes | Allocation concealment |
Keszler 2004 | Randomisation was blinded (using sealed envelopes) but neither intervention nor outcome assessment were blinded. Follow-up was complete. | Single centre, 18
preterm (<34 weeks') infants. Eligible if: RDS (ventilated by 6 hours
of life and expected to require ventilation for > 24 hours). Excluded: congenital cardiac, respiratory or CNS anomalies, paralysis or sedation or ETT leak > 30%. | Both groups: Babylog 8000 ventilator with set backup rate of 40/minute; target pCO2 of 35-45 torr. Experimental group: AC-VG, tidal volume set at 5 ml/kg, changed by 0.5 ml/kg to maintain normocarbia (n=9). Control group: AC with PIP set to achieve 4-6 ml/kg tidal volume, using PIP changes of 1-2 cmH2O to maintain target CO2 levels (n=9). | Death during study, PTX, PIE, tidal volumes. | A | |
Lista 2004 | Randomisation
was by random number sequencing but blinding was not specified. Method of
concealment not specified. All patients were followed up. | Two centres, 53 preterm (25-32 weeks'). Eligible if: at least 1 course of antenatal steroids, ventilated for RDS in first 24 hours, treated with surfactant within 3 hours. Excluded: lethal anomalies, requiring muscle relaxants at entry, IVH greater than grade 2, actual or suspected sepsis. | Both groups:
Babylog 8000+ set at rate 40/min, PEEP 3.5-4, FiO2 to maintain saturations
90-96%, IT 0.4-0.5 sec). Target blood gas parameters - pH> 7.25, pO2
50-75 mmHG, pCO2 40-65 mmHg). Experimental group: PSV + VG (n=30) with target
volume of 5 ml/kg throughout study. Control: PSV with PIP set to achieve tidal volumes of 5 ml/kg and PIP weaned to achieve blood gas goals (n=23). | Death in hospital, PDA, BPD (in oxygen at 28 days), BPD (in oxygen at 36 weeks), IVH, PVL, ROP, PIE, PVL, need for postnatal steroids. The study was also designed to compare inflammatory markers in the two groups. | Although it is not stated that there were post- randomisation deletions the imbalance in numbers between the two groups raises this concern. | B |
Piotrowski 1997 | Sealed envelopes were used to blind. Neither intervention nor outcome measurements were blinded. All infants received follow-up. | Single centre, 57 infants of birthweight < 2500g. Eligible if: needed ventilation for lung disease, age < 72 hours and servo ventilator available. Excluded if: "terminal state", PTX or other air leak, sepsis or meconium aspiration. | Both groups: PEEP 3-5 cm water, IT 0.5 sec, target oxygen saturations 88-95, pCO2 <55. Infants extubated once rate<12/min, FiO2 <0.25, and after a 30-60 min trial of ETT CPAP. Experimental group: Siemens Servo ventilator set at 5-6 ml/kg tidal volume plus 4-5 ml of compressible volume (n=27). Control: Bear Cub or Sechrist ventilator with PIP set to achieve chest movement (n=30). | Death, BPD (in oxygen at 28 days), any air leak, PTX, PIE, any IVH, IVH grade 3 or 4, PDA, sepsis, use of muscle relaxants, duration of ventilation. | A | |
Sinha 1997 | Blinding of randomisation used sealed envelopes. Intervention was not blinded, nor was any outcome assessment other than chest x-ray findings. Follow-up was complete. | Single centre, 50 infants. Eligible if: birthweight > 1200g, RDS requiring ventilation and surfactant. Excluded if: sepsis/pneumonia, congenital malformation or no arterial access. | Both groups: VIP Bird ventilator in assist control
mode with IT at 0.3-0.5 sec. Target pH 7.27-7.40, pCO2 4.5 to 6 kPa, pO2
8-11kpa. Experimental group: volume controlled ventilation with tidal volume set at 5-8 ml/kg (n=25). Control group: time-cycled, pressure-limited ventilation with PIP adjusted to achieve 5-8 ml/kg (n=25). | Death, failed allocated treatment, IVH or PVL (not reported separately), BPD (in oxygen at 36 weeks, PTX, PDA or improved AaDO2. | A |
Study | Reason for exclusion |
Abubakar 2001 | The order of ventilatory modes was not randomised in this crossover trial which means that an effect of fatigue cannot be excluded, and the results did not include the outcome measures of this review. |
Cheema 2001 | This short term crossover study did not address any of the outcome measurements of this review. Also, the crossover was made from TCPL to volume guarantee mode without changing the Maximum PIP, which may have interfered with the ventilator's capacity to deliver the set tidal volumes and hence affected the outcomes. |
Herrera 2002 | A short term crossover trial that did not address the outcome measures of this review. |
Olsen 2002 | This study was excluded for two reasons. It was a short term crossover trial that did not discuss the outcome measurements of this review. It also compared PSV (an inspiratory and expiratory triggering mode where each inflation is supported) to SIMV (where there is only inspiratory triggering and not all infant breaths are supported) as well as addressing the volume question. The outcomes can thus not be attributed to the volume targeted strategy alone. |
Keszler M, Abubakar K. Volume guarantee: stability of tidal volume and incidence of hypocarbia. Pediatric Pulmonology 2004;38:240-5.
Lista 2004 {published data only}
Lista G, Colnaghi M, Castoldi F, Condo V, Reali R, Compagnoni G, Mosca F. Impact of targeted-volume ventilation on lung inflammatory response in preterm infants with respiratory distress syndrome (RDS). Pediatric Pulmonology 2004;37:510-4.
Piotrowski 1997 {published data only}
Piotrowski A, Sobala W, Kawczynski P. Patient-initiated, pressure-regulated, volume-controlled ventilation compared with intermittent mandatory ventilation in neonates: a prospective, randomised study. Intensive Care Medicine 1997;23:975-81.
Sinha 1997 {published data only}
Sinha S, Donn S, Gavey J, McCarty M. Randomised trial of volume controlled versus time cycled, pressure limited ventilation in preterm infants with respiratory distress syndrome. Archives of Disease in Childhood Fetal and Neonatal Edition 1997;77:F202-5.
Abubakar KM, Keszler M. Patient-ventilator interactions in new modes of patient-triggered ventilation. Pediatric Pulmonology 2001;32:71-5.
Cheema 2001 {published data only}
Cheema IU, Ahluwahlia JS. Feasibility of tidal volume-guided ventilation in newborn infants: a randomised crossover trial using the volume guarantee modality. Pediatrics 2001;107:1323-8.
Herrera 2002 {published data only}
Herrera CM, Gerhardt T, Claure N, Everett R, Musante G, Thomas C, Bancalari E. Effects of volume-guaranteed synchronized intermittent mandatory ventilation in preterm infants recovering from respiratory failure. Pediatrics 2002;110:529-33.
Olsen 2002 {published data only}
Olsen SL, Thibeault DW, Truog WE. Crossover trial comparing pressure support with synchronized intermittent ventilation. Journal of Perinatology 2002;22:461-6.
* indicates the primary reference for the study
NSW Pregnancy and Newborn Services Network. Report of the Australian and New Zealand Neonatal Network 1999. Donoghue D, Cust A, editor(s). Sydney: University of Sydney, 2000.
Acute Respiratory Distress Syndrome Network. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. New England Journal of Medicine 2000;342:1301-8.
Bardin C, Zelkowitz P, Papageorgiou A. Outcome of small-for-gestational age and appropriate-for-gestational age infants born before 27 weeks of gestation. Pediatrics 1997;100:E4.
Clark RH, Slutsky AS, Gerstmann DR. Lung protective strategies of ventilation in the neonate: what are they? Pediatrics 2000;105:112-4.
Coalson JJ, Winter VT, Siler-Khodr T. Neonatal chronic lung disease in extremely immature baboons. American Journal of Respiratory and Critical Care Medicine 1999;160:1333-6.
Cools F, Offringa M. Neuromuscular paralysis for newborn infants receiving mechanical ventilation (Cochrane Review). In: The Cochrane Database of Systematic Reviews, Issue 2, 2005.
Goldsmith JP, Karotkin EH. Assisted Ventilation of the Neonate. Third edition. Pennsylvania: Saunders Company, 1996.
Gortner L, Wauer RR, Stock GJ, Reiter HL, Reiss I, Jorch G et al. Neonatal outcome in small for gestational age infants: do they really do better? Journal of Perinatal Medicine 1999;27:484-9.
Greenough A, Milner AD, Dimitriou G. Synchronised mechanical ventilation for respiratory support in newborn infants. In: Cochrane Database of Systematic Reviews, Issue 3, 2004.
Hannaford K, Todd DA, Jeffery H, John E, Byth K, Gilbert G. Role of ureaplasma urealyticum in lung disease of prematurity. Archives of Disease in Childhood Fetal & Neonatal Edition 1999;81:F162-7.
Hernandez LA, Peevy KJ, Moise AA, Parker JC. Chest wall restriction limits high airway pressure-induced lung injury in young rabbits. Journal of Applied Physiology 1989;66:2364-8.
Jobe AH, Bancalari E. Bronchopulmonary dysplasia. American Journal of Respiratory and Critical Care Medicine 2001;163:1723-9.
Jobe AH, Ikegami M. Prevention of bronchopulmonary dysplasia. Current Opinion in Pediatrics 2001;13:124-9.
Lemons JA, Bauer CR, Oh W, Korones SB, Papile LA, Stoll BJ. Very low birth weight outcomes of the National Institute of Child Health and Human Development Neonatal Research Network, January 1995 through December 1996. Pediatrics 2001;107:E1.
Ozdemir A, Brown M, Morgan WJ. Markers and mediators of inflammation in neonatal lung disease. Pediatric Pulmonology 1997;23:292-306.
Sinha SK, Donn SM. Manual of Neonatal Respiratory Care. First edition. New York: Futura, 2000.
Warner BB, Stuart LA, Papes RA, Wispe JR. Functional and pathological effects of prolonged hyperoxia in neonatal mice. American Journal of Physiology 1998;275:L110-7.
Watterberg KL, Demers LM, Scott SM, Murphy S. Chorioamnionitis and early lung inflammation in infants in whom bronchopulmonary dysplasia develops. Pediatrics 1996;97:210-5.
Comparison or outcome | Studies | Participants | Statistical method | Effect size |
---|---|---|---|---|
01 Volume-targeted vs pressure limited ventilation | ||||
01 Death in hospital | 3 | 178 | RR (fixed), 95% CI | 0.62 [0.30, 1.29] |
02 Failure of mode of ventilation | 1 | 50 | RR (fixed), 95% CI | 1.00 [0.22, 4.49] |
03 Need for muscle relaxant | 1 | 57 | RR (fixed), 95% CI | 0.32 [0.07, 1.40] |
04 Duration of intermittent positive pressure ventilation (days) | 2 | 103 | WMD (fixed), 95% CI | -2.93 [-4.28, -1.57] |
05 Patent ductus arteriosus | 3 | 160 | RR (fixed), 95% CI | 0.83 [0.61, 1.14] |
06 Air leak (any) | 1 | 75 | RR (fixed), 95% CI | 0.48 [0.14, 1.66] |
07 Pneumothorax | 3 | 178 | RR (fixed), 95% CI | 0.23 [0.07, 0.76] |
08 Pulmonary interstitial emphysema | 2 | 128 | RR (fixed), 95% CI | 0.87 [0.19, 4.06] |
09 Any cranial ultrasound abnormality (IVH or PVL) | 1 | 50 | RR (fixed), 95% CI | 0.09 [0.01, 1.56] |
10 Any IVH | 1 | 57 | RR (fixed), 95% CI | 0.44 [0.20, 0.98] |
11 Severe IVH (grade 3 or 4) | 2 | 110 | RR (fixed), 95% CI | 0.32 [0.11, 0.90] |
12 Periventricular leucomalacia | 1 | 53 | RR (fixed), 95% CI | 0.38 [0.04, 3.97] |
13 BPD (supplemental oxygen at 28 days) | 2 | 110 | RR (fixed), 95% CI | 0.87 [0.39, 1.96] |
14 BPD (supplemental oxygen at 36 weeks) | 2 | 103 | RR (fixed), 95% CI | 0.34 [0.11, 1.05] |
02 Volume-targeted vs pressure limited ventilation - infants less than 1000g | ||||
01 Death in hospital | 1 | 20 | RR (fixed), 95% CI | 0.25 [0.03, 1.86] |
02 Patent ductus arteriosus | 1 | 20 | RR (fixed), 95% CI | 1.00 [0.17, 5.77] |
03 Air leak (any) | 1 | 20 | RR (fixed), 95% CI | 0.20 [0.01, 3.70] |
04 Pneumothorax | 1 | 20 | RR (fixed), 95% CI | 0.33 [0.02, 7.32] |
05 Pulmonary interstitial emphysema | 1 | 20 | RR (fixed), 95% CI | 0.33 [0.02, 7.32] |
06 Any IVH | 1 | 20 | RR (fixed), 95% CI | 0.43 [0.15, 1.20] |
07 Severe IVH (grade 3 or 4) | 1 | 20 | RR (fixed), 95% CI | 0.33 [0.09, 1.27] |
08 BPD (supplemental oxygen at 28 days) | 1 | 20 | RR (fixed), 95% CI | 0.60 [0.19, 1.86] |
Characteristics | Volume-Controlled | Volume-Guarantee | Volume-Limited |
Ventilator model | 1. VIP Bird 2. Siemens Servo 300 | Draeger Babylog 8000+ | Bear Cub 750 PSV |
Trigger type | 1. Pressure sensor 2. Pressure or flow sensor | Flow (hot wire anemometer) | Flow (hot wire anemometer) |
Sensor position | At ventilator | At Wye piece | At Wye piece |
PIP | Variable, set max | Variable, set max | Pressure-cycled |
Volumes measured that affect ventilation | VTi (displays VTi & VTe) | VTi, VTe | VTi (displays VTi, VTe) |
Set maximal VT | Volume-cycled | Inflation stopped if VTi > 130% set VT | Inflation stops at max set VT |
Adjusts for low VT | Yes | Yes | No |
Modes available | IMV, SIMV, AC, Termination sensitivity | SIMV, AC, PSV | SIMV, AC, PSV |
Prof Colin J Morley, MA DCH MD FRCP FRCPCH FRACP
Professor/Divisional Director
Division of Neonatal Services
Royal Women's Hospital
132 Grattan St
Carlton
Victoria AUSTRALIA
3053
Telephone 1: +61 3 93442000 extension: 2527
E-mail:
morleyc@cryptic.rch.unimelb.edu.au
The review is published as a Cochrane review in The
Cochrane Library, Issue 3, 2005 (see http://www.thecochranelibrary.com for
information). Cochrane reviews are regularly updated as new evidence emerges
and in response to comments and criticisms, and The Cochrane Library should
be consulted for the most recent version of the Review. |