Oxy-PICU RCT – Oxygenation Targets in Ventilated Children

January 16, 2024

Written by Laura Murphy

Spoon Feed
A conservative (SpO₂ 88-92%) oxygenation target resulted in a small but significant greater probability of better outcome compared to a liberal (SpO₂ >94%) target in children admitted to the PICU on invasive mechanical ventilation.

Is less more when it comes to O₂? Probably…
This was an impressive and robust study; it was the largest individually randomized controlled trial in pediatric critical care. It was a pragmatic, multicenter, open-label RCT including 2,040 children older than 38 weeks corrected gestational age and younger than 16 years in 15 UK PICUs*. Children were admitted emergently and enrolled within 6 hours of acceptance to a participating PICU with face-to-face contact with PICU or emergency medical transport staff.

The primary outcome was duration of organ support at 30 days (rank-based endpoint with a score assigned between 1 to 30 depending on days of organ support, with a score of 31 for death)**. This was compared between two groups using a two-sample rank-sum test to estimate probabilistic index (i.e. probability that the intervention is superior to the control). This type of outcome allowed for analysis of differences across the distribution of organ support rather than reliance on median values (which were similar in this study).

While the median length of stay on organ support was similar (5 days) in both groups, the probabilistic index of better outcome in the conservative group was 0.53 (95%CI 0.50-0.55, p=0.04; adjusted odds ratio 0.84, 95% CI 0.72-0.99). These findings were consistent across components of primary outcome. As figure 3B shows, the change between the groups is asymmetric; there are more children in the conservative group with shorter length-of-stay, and death and very long-stay are less frequent in conservative group. Prespecified adverse events were similar in both groups (3% in conservative group vs 4% in liberal group).

Patients were analyzed in their randomized subgroup (intention-to-treat analysis). Notably, the conservative group had a median SpO₂ of 94% (above target), though authors state that this was largely due to lack of need for supplemental oxygen in this group. Median FiO₂ was lower in the conservative group (0.27 vs 0.35). While on mechanical ventilation, the rate of non-adherence was low (3.8% in conservative, 1.5% liberal), with deviations for acute deterioration, staffing (lack of awareness of the trial) or other clinical priorities. That said, this deviation from protocol is more likely to underestimate (rather than overestimate) the difference between the two groups.

While secondary analyses were exploratory, some interesting findings include decreased time to liberation from invasive mechanical ventilation in the conservative group, adjusted hazard ratio 1.11 (95%CI 1.01 to 1.21), and lower median cost, £32,749 vs £34,725, adjusted mean difference £-2,142, (95%CI –£4320 to £34). Cost saving implications across an entire health system could be significant.

Bottom line: While the effect of the intervention is small, given the large number of critically ill patients treated with the intervention, the clinical effect across the population is large. This study is the first trial to support the existing expert opinion-based guidelines of SpO₂ target of 88-92% in children with severe respiratory failure.

How will this change my practice?
Outside of specific populations*, I will target SpO₂ of 88-92% for critically ill children requiring mechanical ventilation with supplemental O₂. This trial provides support to the existing guidelines and suggests that better adherence to this provides both clinical benefit and cost savings. 

*Excluded premature infants, children with congenital cardiac disease, pulmonary hypertension, sickle cell disease, acute neurologic injury.

**This included respiratory support (noninvasive or invasive positive pressure ventilation), cardiovascular support (vasoactive or fluid bolus), renal support, analgesia or sedation, exchange transfusion, neurologic and metabolic support.

Source
Conservative versus liberal oxygenation targets in critically ill children (Oxy-PICU): a UK multicentre, open, parallel-group, randomised clinical trial. Lancet. 2023 Dec 1:S0140-6736(23)01968-2. doi: 10.1016/S0140-6736(23)01968-2. Epub ahead of print.