First-attempt neonatal intubation went from 45% to 61%

Design

single-centre quality improvement project with plan-do-study-act cycles and statistical process control, plus a sustainability phase

Population

642 tracheal intubations in a tertiary neonatal intensive care unit, April 2023 to May 2025

Primary outcome

first-attempt intubation success rate

Effect

45% at baseline to a sustained process mean of 61%; infants under 1000 g 38% to 62%; video laryngoscopy use 0% to 70-90%

A tertiary neonatal unit ran a structured quality improvement project against a baseline first-attempt intubation success rate of 45%, aiming for above 55%. Root cause analysis pointed at two drivers of failure: poor glottic visualisation, and the infant destabilising during the attempt. The interventions followed directly - phase in video laryngoscopy as the primary device, and standardise oxygenation during laryngoscopy with nasal high flow.

Across 642 intubations, control chart analysis showed the first-attempt success rate rise from 45% to a sustained process mean of 61%, held through a ten-month sustainability phase. Video laryngoscopy use went from 0% at baseline to 70% to 90%. Improvement was seen across all clinician grades and all locations. The most striking subgroup is the one that matters most: in infants under 1000 g, first-attempt success rose from 38% to 62%.

This is a single-centre before-and-after design, so secular improvement cannot be excluded - but statistical process control with a sustained shift, and a process measure moving in step with the outcome, is about as much causal reassurance as a QI study can offer. The transferable part is not the equipment alone. It is the pairing of a device that lets the whole team see the glottis with an oxygenation strategy that buys time, and the fact that both were standardised rather than left to individual preference.

• Pair video laryngoscopy with apnoeic oxygenation by nasal high flow - the project changed both, and either alone is a weaker intervention.
• Standardise the approach rather than allowing device choice per operator; the process measure was uptake, and uptake was what moved.
• Track first-attempt success on a control chart, not as a monthly percentage - the shift is what tells you it worked.
• Audit the under-1000 g group separately; it starts lowest and has the most to gain.
• A shared screen also makes supervision possible, which is where a training unit gets a second benefit.

The statistics, in plain English

This is a quality improvement design, not a randomised comparison, so the 16 percentage point rise is an association with a package of changes over time rather than a measured treatment effect. What makes it more than a trend is the control chart method: a sustained shift in the process mean, held through a separate sustainability phase, is harder to explain by chance or by gradual drift than a single before-and-after comparison would be. The process measure - video laryngoscopy use rising from 0% to 70% to 90% - moving alongside the outcome adds to that. No confidence intervals are reported, which is usual for statistical process control but does mean the precision of the 61% figure is not stated.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.