A flow sensor suitable for use with split-flow ventilation--first preclinical data.

Volutrauma caused by artificial ventilation represents a major morbidity risk for premature infants. Our working group has recently developed an innovative "split-flow ventilation" system aiming at the reduction of tidal volumes (TVs). The main problem for the practical use of this system is the fact that conventional measurements of commercially available flow sensors are distorted by the split flow. In this study, we present the first preclinical data from testing an adapted flow sensor combination recognizing the split flow. A preterm infant test lung was conventionally ventilated, modified by insertion of a split-flow line. In addition to the customary flow sensor (FS-1), a second flow sensor (FS-2) was integrated into the split-flow line, and a third (FS-3) was placed at the exit of the test lung for reference measurements. The signals of all three flow sensors were read and processed by a computer. The program was set to graphically add up flow curves 1, 2, and 3 during one ventilation loop. After 10 runs, a mean curve of FS-1+2 was calculated and compared to the mean curve of FS-3. Furthermore, the mean TV of 10 runs measured by FS-1+2 was calculated and compared with the mean TV calculated by FS-3. The summation curve FS-1+2 proved identical to the reference curve FS-3. FS-1+2 yielded a TV of 6.6 +/- 0.01 mL (inspiratory) and 6.7 +/- 0.02 mL (expiratory). The corresponding values of FS-3 were 6.5 +/- 0.20 mL and 6.6 +/- 0.09 mL, respectively. According to our results, the presented flow sensor constellation allows exact flow measurements in the experimental setting and appears suitable for usage in a split-flow ventilation circuit under clinical conditions.