Patient-triggered ventilation: a comparison of tidal volume and chestwall and abdominal motion as trigger signals.

Patient-triggered synchronized ventilation requires reliable and early detection of the infant's inspiratory effort. Several trigger methods have been developed that frequently lack the sensitivity to detect inspiration in small preterm infants (trigger failure), or show a high rate of breaths triggered by artifacts in the respiratory signal (autotrigger). The purpose of this study was to determine the effectiveness of the following trigger signals: abdominal movement sensed by a newly developed induction technique, chestwall motion detected by changes in transthoracic impedance, and tidal volume measured by anemometry at the endotracheal tube connector. Ten preterm infants (birth weight, 580-1,424 g; median weight, 943 g; study weight, 535-1,415 g; median weight, 838 g; gestation age, 26-32 weeks, median gestational age, 28 weeks, study age, 1-50 days, median study age, 11 days) were included in the study. A Sechrist SAVI ventilator was triggered by one of three signals: chestwall or abdominal movement, or tidal volume generated by the infants. Response time between beginning of inspiratory flow, the occurrence of the trigger signal (signal delay), and the onset of the triggered breath (trigger delay) were determined for each of the three signals. The signal response time was -13.5 msec (95% CI, -33 to -2 msec) for the abdominal movement signal, indicating that it started before inspiratory flow; 0.0 msec for the volume signal; and 44.0 msec (95% CI, 29-73 msec) for the chestwall signal (P < 0.002); this long delay was secondary to chestwall distortion and a subsequent delay in outward ribcage movement in many infants. The trigger delay for the abdominal signal was 90.0 msec (95% CI, 55-104 msec), 135.5 msec (95% CI: 82-186 msec) for the volume signal, and 176.5 msec (95% CI: 165-232 msec) for the chestwall signal, indicating that there was a difference in the rise time of signal voltage between the three methods (P < 0.01). The rate of autotriggered breaths was 3.2% (95% CI, 0.3-9.3%) when using the abdominal signal, 0.55% (95% CI, 0.0-2.1%) for the tidal volume signal, and 11.25% (95% CI, 0.5-27.8%) for the chestwall signal (P < 0.05). The incidence of trigger failure was low with all three signals and was not significantly different between the techniques. In summary, the chestwall signal had a long trigger delay and was highly susceptible to false triggering. It is, therefore, not a reliable trigger signal for synchronized mechanical ventilation in preterm infants. In contrast, tidal volume and abdominal movement signals had an acceptable trigger delay and a low rate of autotriggering, making them useful clinical trigger signals.