Dangerous pressurization and inappropriate alarms during water occlusion of the expiratory circuit of commonly used infant ventilators

Abstract

Background Non-invasive continuous positive airways pressure is commonly a primary respiratory therapy delivered via multi-purpose ventilators in premature newborns. Expiratory limb occlusion due to water accumulation or 'rainout' from gas humidification is a frequent issue. A case of expiratory limb occlusion due to rainout causing unexpected and excessive repetitive airway pressurisation in a Draeger VN500 prompted a systematic bench test examination of currently available ventilators. Objective To assess neonatal ventilator response to partial or complete expiratory limb occlusion when set to non-invasive continuous positive airway pressure mode. Design Seven commercially available neonatal ventilators connected to a test lung using a standard infant humidifier circuit with partial and/or complete expiratory limb occlusion were examined in a bench test study. Each ventilator was set to deliver 6 cmH2 O in non-invasive mode and respiratory mechanics data for 75%, 80% and 100% occlusion were collected. Results Several ventilators responded inappropriately with complete occlusion by cyclical pressurisation/depressurisation to peak pressures of between 194 and 646cmH2 O at rates varying between 2 to 77 inflations per minute. Tidal volumes varied between 101 and 243mL. Alarm responses varied from 'specific' (tube occluded) to 'ambiguous' (Safety valve open). Carefusion Avea responded by continuing to provide the set distending pressure and displaying an appropriate alarm message. Draeger Babylog 8000 did not alarm with partial occlusions and incorrectly displayed airways pressure at 61cmH2 O compared to the measured values of 13cmH2 O. Conclusions This study found a potential for significant adverse ventilator response due to complete or near complete expiratory limb occlusion in CPAP mode.