Effect of leak and breathing pattern on the accuracy of tidal volume estimation by commercial home ventilators: a bench study.

Luján, Manel; Sogo, Ana; Pomares, Xavier; Monsó, Eduard; Sales, Bernat; Blanch, Lluís

BACKGROUND New home ventilators are able to provide clinicians data of interest through built-in software. Monitoring of tidal volume (VT) is a key point in the assessment of the efficacy of home mechanical ventilation. OBJECTIVE To assess the reliability of the VT provided by 5 ventilators in a bench test. METHODS Five commercial ventilators from 4 different manufacturers were tested in pressure support mode with the help of a breathing simulator under different conditions of mechanical respiratory pattern, inflation pressure, and intentional leakage. Values provided by the built-in software of each ventilator were compared breath to breath with the VT monitored through an external pneumotachograph. Ten breaths for each condition were compared for every tested situation. RESULTS All tested ventilators underestimated VT (ranges of -21.7 mL to -83.5 mL, which corresponded to -3.6% to -14.7% of the externally measured VT). A direct relationship between leak and underestimation was found in 4 ventilators, with higher underestimations of the VT when the leakage increased, ranging between -2.27% and -5.42% for each 10 L/min increase in the leakage. A ventilator that included an algorithm that computes the pressure loss through the tube as a function of the flow exiting the ventilator had the minimal effect of leaks on the estimation of VT (0.3%). In 3 ventilators the underestimation was also influenced by mechanical pattern (lower underestimation with restrictive, and higher with obstructive). CONCLUSIONS The inclusion of algorithms that calculate the pressure loss as a function of the flow exiting the ventilator in commercial models may increase the reliability of VT estimation.