A Stepped Wedge Cluster Randomised Trial Of Video Versus Direct Laryngoscopy For Intubation Of Newborn Infants

INTRODUCTION Many newborn infants have difficulty breathing after birth. Some of these babies have a tube inserted into their "windpipe" (trachea) - an endotracheal tube (ETT) - through which they are given breathing support (ventilation). When clinicians attempt to intubate (insert an ETT), they use an instrument called a laryngoscope to view the airway in order to identify the entrance to the trachea (larynx). Standard laryngoscopes have a "blade" (which, despite its name, is not sharp) with a light at the tip. Doctors insert the blade into the baby's mouth to view the larynx. Traditionally, clinicians used a standard laryngoscope to look directly into the baby's mouth to view the larynx (direct laryngoscopy, DL). When clinicians attempt to intubate newborns with DL, less than half of first attempts are successful. Also adverse effects - such as falls in the blood oxygen levels (fall in oxygen saturation (SpO2), or "desaturation"), slowing down of the heart rate (bradycardia), oral trauma - are relatively common.

In recent years, video laryngoscopes (VL) have been developed. In addition to a light, VL have a video camera at the tip of the blade. This camera acquires a view of the larynx and displays it on a screen that the clinician views when attempting intubation (indirect laryngoscopy). In a randomised study performed at the National Maternity Hospital, Dublin, Ireland, more infants were successfully intubated at the first attempt when clinicians used VL compared to DL [79/107 (74%) versus 48/107 (45%), P<0.001]. While this study was large enough to show that VL resulted infants being successfully intubated at the first attempt in one hospital, it couldn't give information about how it might work in a range of hospitals, and it wasn't large enough to see what effect VL had on adverse events. There is a large difference in cost between a standard laryngoscope (approx. €300) and a video laryngoscope (approx. €21,000). This is a matter of concern for all hospitals, particularly in settings where resources are more limited.

The investigators aim to assess whether VL compared to DL results in more infants being intubated at the first attempt without physiological instability.

STUDY DESIGN A recent single centre study reported that that more newborn infants were successfully intubated at the first attempt when VL was used to indirectly view the airway compared to DL. This study was not large enough to determine the effect of VL on adverse effects that are seen commonly (e.g. desaturation) or more rarely (e.g. bradycardia, receipt of chest compressions or adrenaline, oral trauma) during intubation attempts.

For the current study, the investigators chose a stepped-wedge cluster randomised controlled design, where the participating centre, rather than the individual infant, will be the unit of randomisation. This design has been found appropriate to test the effects of an intervention that encompasses a behavioural aspect and to implement interventions while studying them at the same time. In this study, all centres will begin in the "control group"; where clinicians will routinely attempt intubation with DL, as is their usual practice. At specified intervals, centres will be randomly assigned to cross over to the "intervention group", where clinicians will routinely attempt intubation with VL. All participating centers will have included patients in both arms by the end of the study.

SAMPLE SIZE ESTIMATION To determine the intra-cluster correlation (that means the correlation between two observations from the same centre), the investigators used the dataset of the MONITOR trial that included infants from 7 delivery rooms worldwide. In this trial, the intra-cluster correlation for intubation in the delivery room was reported as 0.1.

This complete stepped-wedge cluster-randomized design includes 21 time periods (including the baseline) and 20 centres that will be including patients, with each randomised to a unique sequence. Each time period lasts a fortnight. Each time period, 1 centre will switch their treatment from DL to VL. With all centres including 2 patients each time period, 42 patients will be included per centre which will provide a total sample size of 840 patients.

Assuming a control proportion of 0.4, this sample will achieve 90% power (0.9091) to detect a treatment proportion of 0.55, assuming a conservative ICC of 0.05. The power is not very sensitive to ICC values up to 0.1 (power of >90% to detect difference 40% versus 56%). The test statistic used is the two-sided Wald Z-Test.

TREATMENT OF SUBJECTS

DIRECT LARYNGOSCOPY (DL, control period) At the start of the study, clinicians at participating centres will attempt intubation using a standard laryngoscope to perform DL as is their normal practice.

VIDEO LARYNGOSCOPY (VL, intervention period) For each centre, a lot will be drawn which indicates the month in which endotracheal intubation will be routinely attempted with VL rather than DL. In the month before the switch, centres will be provided with a C-MAC VL by the manufacturers, Karl Storz-Endoskop (Tuttlingen, Germany). The system will be provided on loan for the duration of the study and will consist of an 8" high-definition monitor with connecting cable and reusable straight Miller type blades size 0 and size 1. The equipment will be demonstrated by representatives from Karl Storz, and clinicians who intubate babies at participating hospitals will be encouraged to practice with the equipment on mannequins. We will have an virtual meeting with each centre in the week before they are due to switch to review the protocol, data collection and to answer any queries that they may have.

All other procedures in the delivery room and NICU will be performed according to international and local guidelines. All other aspects of the approach to intubation at the participating centre are at the discretion of the local clinicians and should remain the same for the duration of the study; e.g.:

• The drugs used before intubation attempts (e.g. opiate, atropine, curare-like drug)
• The route by which intubation is usually attempted (i.e. oral or nasal)
• Whether they use a stylet is routinely used
• Whether supplemental oxygen is given during attempts