Objective The authors previously showed that 48% of infants <800 g were ventilated with tidal volume (VT) < dead space (DS) using volume guarantee (VG) ventilation. Here, The authors sought to confirm those findings under the rigorous conditions of a bench study.

Design and methods The authors measured the time to wash out CO2 from a 45-ml test lung using end-tidal CO₂ monitor (ETCO₂). The test lung was filled with 100% CO₂, then ventilated using VG at VT ranging from DS+2 ml to DS–1.5 ml. With ventilation, ETCO₂ declined exponentially as CO₂ was washed out, the rate being proportional to VT – effective instrumental DS. The time from initiation of ventilation to threshold of accurate detection was determined in triplicate.

Results Halving the theoretical ‘alveolar ventilation’ (DS+2 ml to DS+1 ml) only increased the elimination time by 26%, not the 100%, as predicted by conventional physiology. CO₂ washout was less efficient, but still occurred even at VT=DS and VT=DS–1.5 ml. Halving the theoretical ‘alveolar ventilation’ by decreasing respiratory rate from 80 to 40 breaths/min only increased elimination time by 35%, not 100%, as predicted by conventional physiology. Twenty minutes of continuous positive airway pressure prior to ventilation did not alter the elimination time, verifying that CO₂ did not diffuse or leak out of the test lung. Size of the endotracheal tube (ETT; 2.5, 3.0 and 3.5 mm) flow rate (4, 6 and 10 l/min) and inspiratory time (0.25 vs 0.35 s) did not affect the results.

Conclusions Contrary to conventional physiology, effective CO₂ elimination appears to be possible with VT<DS even at conventional rates. With small ETT a spike of fresh gas likely penetrates through the DS, rather than pushing it ahead.