Background and aims Expected normal tidal volume values have not been described in children treated on NIV yet. Describe tidal volumen/weight (V_T/W) values according to the ventilators and interface used while receiving non-invasive ventilation (NIV) in the Paediatric Intensive Care Unit (PICU).

Methods Prospective study done in 4 Spanish PICU of University Hospitals during 2013. Consecutive register using a dataweb (www. NIVresearch.net), of patients diagnosed with acute respiratory failure (ARF) treated with NIV. V_T/W was measured at the beginning of NIV, 1, 2, 6, 12 and 24 h. Failure was defined as need for intubation. Bonferroni’s adjust for mean comparison was used.

Results 607 episodes were included. 58% male. Median age was 11.3 months (P25 2.1–P75 60.5). Mean values V_T/W (ml/kg) between groups success/failure were: starting NIV was 9.4 ± 5 DS vs 8.3 ± 3.4 DS, at 1 h 8.5 ± 4DS vs 7 ± 3DS (p 0.02), at 2 h 9.2 ± 5 DS vs 7.8 ± 6DS, at 6 h 9.7 ± 5.4 DS vs 8.5 ± 3.8DS, at 12 h 9 ± 5DS vs 9.5 ± 14DS and at 24 h 9 ± 4,8DS vs 10.6 ± 8DS. Mean V_T/W shows significant differences between BiPAP Vision and conventional ventilator at start point (p 0.022), 1 h (p 0.01), 6 h (p 0.016), 12 h (p 0.01), 24 h (p 0.01). Differences of V_T/weight between several types of interfaces reached statistical significance.

Conclusions V_T/weight differs differs according to the ventilator and interface used. V_T/weight differences between success and failure group suggest that this variable could be a new predictive factor for NIV failure. Further studies with multivariable adjustment should explore this hypothesis.