Endotracheal intubation versus face mask for newborns resuscitated with positive pressure ventilation at birth
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
In newly born infants resuscitated with intermittent positive pressure ventilation, does PPV via an endotracheal tube as compared to a face mask reduce mortality and morbidity?
Planned subgroup analyses:
-
gestational age: term (37 weeks gestation and above), preterm (less than 37 weeks) infants
-
professional category or experience of resuscitator
Background
Internationally agreed consensus statements advise how to resuscitate newborns and state that effective ventilation is the key to success (ILCOR 1999; ILCOR 2000). It is advised that positive pressure ventilation (PPV) be initiated using manual ventilation devices via a face mask, and stated that most infants can be adequately ventilated using this technique. It is noted that endotracheal intubation may be indicated at several points during neonatal resuscitation: when tracheal suctioning for meconium is required, if bag‐mask ventilation is ineffective or prolonged, when chest compressions are performed, when tracheal administration of medications is desired, or in special resuscitation circumstances such as congenital diaphragmatic hernia or extremely low birth weight (ILCOR 1999; ILCOR 2000). It is also noted that the timing of endotracheal intubation may depend on the skill and experience of the resuscitator.
While endotracheal intubation is an important skill in neonatal medicine, clinicians appear to have difficulty with the procedure. Difficulty with intubation during resuscitation was a frequently cited concern in a study of the care of infants born at 27 and 28 weeks of gestation in the UK (CESDI(UK)). Studies of paediatric residents (Carbine 2000; Falck 2003; Lane 2004), neonatal fellows and neonatologists (Lane 2004) in tertiary academic centres in the US suggest that half or of more of intubation attempts are unsuccessful. US studies of intubation performed in tertiary academic centres by neonatologists, fellows, residents and respiratory therapists in which the detection of exhaled carbon dioxide was used to confirm correct tube placement suggest that oesophageal intubation is not infrequent (Roberts 1995; Aziz 1999; Repetto 2001; Lane 2004). While trends towards greater proficiency with greater experience were observed in some of these studies (Falck 2003; Lane 2004), they did not reach statistical significance.
Studies which have recorded the first three inflations given to term or near‐term infants during neonatal resuscitation concluded that volumes sufficient for tidal exchange were rarely produced via a mask and that success of resuscitative efforts depended on stimulation of the infants' own respiratory efforts (Milner 1984; Field 1986; Hoskyns 1987). The tidal volumes delivered via a face mask were substantially less than those the authors had delivered via an endotracheal tube in comparable infants (Milner 1984).
Objectives
In newly born infants resuscitated with intermittent positive pressure ventilation, does PPV via an endotracheal tube as compared to a face mask reduce mortality and morbidity?
Planned subgroup analyses:
-
gestational age: term (37 weeks gestation and above), preterm (less than 37 weeks) infants
-
professional category or experience of resuscitator
Methods
Criteria for considering studies for this review
Types of studies
All randomised and quasi‐randomised controlled trials will be included
Types of participants
Term and preterm infants resuscitated using positive pressure ventilation at birth.
Types of interventions
Intubation for PPV versus PPV via a face mask. International consensus statements advise that when PPV is initiated via a mask, endotracheal intubation may nevertheless be indicated later in the course depending on response. Studies which include infants randomised to receive PPV via a face mask who are subsequently intubated for resuscitation will be eligible for inclusion in this review and will be analysed in the group to which they were originally randomised (by intention to treat).
Types of outcome measures
The primary outcomes for this review will be:
-
Death in the delivery room
-
Death during hospitalisation or to latest follow‐up
The secondary outcomes for this review will be:
-
Apgar scores at one and five minutes
-
Heart rate at five minutes
-
Need for intubation for respiratory support beyond the delivery room
-
Duration in hours of respiratory support ie. nasal continuous airway pressure and ventilation via an endotracheal tube considered separately and in total
-
Duration in days of supplemental oxygen requirement
-
Chronic lung disease: the need for supplemental oxygen at 28 days of life; the need for supplemental oxygen at 36 weeks gestational age for infants born before 32 weeks gestation
-
Air leaks (pneumothorax, pneumomediastinum, pneumopericardium, pulmonary interstitial emphysema) reported either individually or as a composite outcome
-
Seizures including clinical and electroencephalographic
-
Hypoxic ischaemic encephalopathy (grade I ‐ III Sarnat 1976)
-
Cranial ultrasound abnormalities: any intraventricular haemorrhage (IVH), grade 3 or 4 (IVH) according to Papile classification (Papile 1978) and cystic periventricular leukomalacia
-
Long term neurodevelopmental outcome (rates of cerebral palsy on physician assessment, developmental delay, ie. IQ 2 standard deviations less than the mean on validated assessment tools eg Bayley's Mental Developmental Index).
Search methods for identification of studies
The standard search strategy of the Cochrane Neonatal Group will be used. PubMed (1966 to 2004) and CINAHL (1982 to 2004) will be searched using the MeSH headings "Infant, Newborn", "Resuscitation", "Respiration, Artificial" AND "Intubation, Intratracheal". The Cochrane Central Register of Controlled Trials (CENTRAL, The Cochrane Library, Issue 2, 2004) will be searched using textwords Newborn AND Resuscitation. No language restrictions will be applied. The abstracts of the Society for Pediatric Research, and the European Society for Pediatric Research will be searched from 1995‐2004.
Data collection and analysis
The standard methods of the Neonatal Review Group of the Cochrane Collaboration will be employed. Trial searches, assessments of methodology and extraction of data will be performed independently by each reviewer with comparison and resolution of any differences found at each stage. Methodology will be assessed regarding blinding of randomisation, intervention and outcome measurements as well as completeness of follow‐up. Where any queries arise or where additional data are required, the authors will be contacted. Categorical data (e.g. number dying or developing bronchopulmonary dysplasia) will be extracted for each intervention group, and relative risk (RR), relative risk reduction, risk difference (RD) and number needed to treat (NNT) calculated. Mean and standard deviation will be obtained for continuous data (e.g. number of days of respiratory support, or duration of oxygen dependency) and analysis performed using the weighted mean difference (WMD). For each measure of effect the 95% confidence intervals will be given. The fixed effect model will be applied for meta‐analysis. Heterogeneity will be evaluated using the I2 statistic to determine the suitability of pooling results. We will perform subgroup analyses according to gestational age [term (37 weeks gestation and above) and preterm (less than 37 weeks)] and according to the professional category or experience of the resuscitator.
