The potentials of PSV-VG to improve infant-ventilator synchrony and to decrease pressure needed to ventilate premature lungs are promising, even though the changes were small. However, its benefits during acute illness and on the final outcome remain to be proven.

INTRODUCTION: Disturbance of cerebral blood flow (CBF) has been associated with neonatal brain injury. Because CBF is greatly influenced by mechanical ventilation, it is important to consider the effect of ventilation mode on CBF. We have shown that pressure-support ventilation combined with volume-guarantee mode (PSV-VG) would lower ventilation pressures and improve infant-ventilator synchrony when compared with synchronized intermittent mandatory ventilation (SIMV). OBJECTIVE: Our aim was to compare the effect of PSV-VG and SIMV on CBF. METHODS: To study an on/off effect, 33 preterm infants (mean gestational age: 26.6 ± 2.2 weeks) were switched from SIMV to PSV-VG and back again to SIMV. By using Doppler techniques, anterior cerebral artery pulsatility index (PIaca), superior vena cava flow (SVCF), and aortic minute distance (MDao) were measured in each phase. Infants were grouped according to patency of ductus arteriosus. RESULTS: When the duct was closed, the PIaca decreased significantly (P = .002), which indicates higher CBF, SVCF increased by 20.1% ± 7.4% (P = .01), and MDao increased by 14.6% ± 4.1% (P = .007) during PSV-VG. In the group with patent ductus arteriosus, although MDao decreased by 10.5% ± 4.4% (P = .04) during PSV-VG, the PIaca and SVCF remained constant. We also observed noticeable changes in the pattern of SVCF Doppler waveform as infant-ventilator interaction changed with switching the ventilation mode. Such pattern changes are described here for the first time. CONCLUSIONS: The interaction between the ventilation mode, the shunt across the duct, and probably CBF autoregulation determines the effect of mechanical ventilation on CBF. Studying the SVCF Doppler waveform pattern may be a useful tool for assessing ventilator-patient interaction.