Patient-ventilator interactions in new modes of patient-triggered ventilation.

Abstract

Recently, synchronized modes of conventional mechanical ventilation became available for neonatal ventilatory support, but there has been little information regarding details of patient-ventilator interactions during pressure support, volume support, or any other volume-targeted modes of synchronized ventilation in newborn infants. Our objective was to obtain comparative data on patient-ventilator interactions and stability of delivered tidal volume (V(T)) for the different modes of synchronized mechanical ventilation in stable ventilated newborn infants. We examined the effects of pressure support ventilation (PSV) and volume guarantee (VG) modes of a prototype Dräger Babylog ventilator on peak and mean airway pressures (PIP and Paw), inspiratory time (t(in)), and V(T) in 23 ventilated newborn infants. Twelve infants were studied while on assist/control (AC) and 11 on synchronized intermittent mandatory ventilation (SIMV). Mean birth weight was 1,650 +/- 1,180 g, gestational age 31 +/- 6 weeks, and age at time of study was 19 +/- 26 days. Data for 400-600 breaths from each infant were downloaded directly from the ventilator pressure and volume-monitoring module, and analyzed using ANOVA for repeated measures. Mean values and breath-to-breath variability were compared for 20-min periods of AC or SIMV followed by PSV, PSV+VG, and back to baseline AC or SIMV. PSV and PSV+VG led to shorter t(in) and thus to lower Paw, compared to AC. Mean PIP was similar across all AC modes but more variable during VG, reflecting the servocontrol of PIP. V(T) did not differ between AC modes, but was significantly less variable with VG added. PSV and PSV+VG led to lower and less variable PIP and Paw, compared to SIMV, because t(in) was shorter and every breath was supported in PSV and PSV+VG. V(T) was similar in SIMV, PSV, and PSV+VG, but less variable with PSV+VG. Arterial blood gas tensions were similar across all ventilation modes. We conclude that the ventilator prototype functioned as intended. Breath-to-breath tidal volume variability was significantly reduced in VG modes, although not completely eliminated.