Multifaceted bench comparative evaluation of latest intensive care unit ventilators.

BACKGROUND Independent bench studies using specific ventilation scenarios allow testing of the performance of ventilators in conditions similar to clinical settings. The aims of this study were to determine the accuracy of the latest generation ventilators to deliver chosen parameters in various typical conditions and to provide clinicians with a comprehensive report on their performance. METHODS Thirteen modern intensive care unit ventilators were evaluated on the ASL5000 test lung with and without leakage for: (i) accuracy to deliver exact tidal volume (VT) and PEEP in assist-control ventilation (ACV); (ii) performance of trigger and pressurization in pressure support ventilation (PSV); and (iii) quality of non-invasive ventilation algorithms. RESULTS In ACV, only six ventilators delivered an accurate VT and nine an accurate PEEP. Eleven devices failed to compensate VT and four the PEEP in leakage conditions. Inspiratory delays differed significantly among ventilators in invasive PSV (range 75-149 ms, P=0.03) and non-invasive PSV (range 78-165 ms, P<0.001). The percentage of the ideal curve (concomitantly evaluating the pressurization speed and the levels of pressure reached) also differed significantly (range 57-86% for invasive PSV, P=0.04; and 60-90% for non-invasive PSV, P<0.001). Non-invasive ventilation algorithms efficiently prevented the decrease in pressurization capacities and PEEP levels induced by leaks in, respectively, 10 and 12 out of the 13 ventilators. CONCLUSIONS We observed real heterogeneity of performance amongst the latest generation of intensive care unit ventilators. Although non-invasive ventilation algorithms appear to maintain adequate pressurization efficiently in the case of leakage, basic functions, such as delivered VT in ACV and pressurization in PSV, are often less reliable than the values displayed by the device suggest.