Co-ordinated multidisciplinary intervention to reduce time to successful extubation for children on mechanical ventilation: The SANDWICH cluster stepped-wedge RCT

Background: Daily assessment of patient readiness for liberation from invasive mechanical ventilation can reduce the duration of ventilation. However, there is uncertainty about the effectiveness of this in a paediatric population. Objectives: To determine the effect of a ventilation liberation intervention in critically ill children who are anticipated to have a prolonged duration of mechanical ventilation (primary objective) and in all children (secondary objective). Design: A pragmatic, stepped-wedge, cluster randomised trial with economic and process evaluations. Setting: Paediatric intensive care units in the UK. Participants: Invasively mechanically ventilated children (aged < 16 years). Interventions: The intervention incorporated co-ordinated multidisciplinary care, patient-relevant sedation plans linked to sedation assessment, assessment of ventilation parameters with a higher than usual trigger for undertaking an extubation readiness test and a spontaneous breathing trial on low levels of respiratory support to test extubation readiness. The comparator was usual care. Hospital sites were randomised sequentially to transition from control to intervention and were non-blinded. Main outcome measures: The primary outcome measure was the duration of invasive mechanical ventilation until the first successful extubation. The secondary outcome measures were successful extubation, unplanned extubation and reintubation, post-extubation use of non-invasive ventilation, tracheostomy, post-extubation stridor, adverse events, length of intensive care and hospital stay, mortality and cost per respiratory complication avoided at 28 days. Results: The trial included 10,495 patient admissions from 18 paediatric intensive care units from 5 February 2018 to 14 October 2019. In children with anticipated prolonged ventilation (n = 8843 admissions: Control, n = 4155; intervention, n = 4688), the intervention resulted in a significantly shorter time to successful extubation [cluster and time-adjusted median difference –6.1 hours (interquartile range –8.2 to –5.3 hours); adjusted hazard ratio 1.11, 95% confidence interval 1.02 to 1.20; p = 0.02] and a higher incidence of successful extubation (adjusted relative risk 1.01, 95% confidence interval 1.00 to 1.02; p = 0.03) and unplanned extubation (adjusted relative risk 1.62, 95% confidence interval 1.05 to 2.51; p = 0.03), but not reintubation (adjusted relative risk 1.10, 95% confidence interval 0.89 to 1.36; p = 0.38). In the intervention period, the use of post-extubation non-invasive ventilation was significantly higher (adjusted relative risk 1.22, 95% confidence interval 1.01 to 1.49; p = 0.04), with no evidence of a difference in intensive care length of stay or other harms, but hospital length of stay was longer (adjusted hazard ratio 0.89, 95% confidence interval 0.81 to 0.97; p = 0.01). Findings for all children were broadly similar. The control period was associated with lower, but not statistically significantly lower, total costs (cost difference, mean £929.05, 95% confidence interval –£516.54 to £2374.64) and significantly fewer respiratory complications avoided (mean difference –0.10, 95% confidence interval –0.16 to –0.03). Limitations: The unblinded intervention assignment may have resulted in performance or detection bias. It was not possible to determine which components were primarily responsible for the observed effect. Treatment effect in a more homogeneous group remains to be determined. Conclusions: The intervention resulted in a statistically significant small reduction in time to first successful extubation; thus, the clinical importance of the effect size is uncertain.