PURPOSE: Detection of intrinsic positive pressure ventilation (auto-PEEP)by auscultation of the expiratory valve of the ventilator and listening for persistence of airflow at end-expiration.
METHODS: A dual lung model( Michigan instruments 5600i )was set up to simulate auto-PEEP on a ventilator(Siemans model 300 ventilator).This was achieved by using the following parameters-compliance 60 liters/cm H20, frequency 16 breaths per minute, inspiratory time one second/expiratory time two seconds,tidal volume 600cc,flow 60 litres per minute. This correlated to auto-PEEP of 2cmH20 as measured by end expiratory hold maneuver and confirmed by use of pressure manometer.To replicate spontaneous breathing pattern the lung model was simultaneously attached to a second ventilator ( Siemans model 900c) with higher respiratory rate 20 breaths per minute. This results in inspiration being triggered by either Siemens model 300 ventilator ( controlled breathing) or the second lung bellow on the lung model ( spontaneous effort) which is attached to the Siemens 900c. In this case the presence or absence of PEEPi was confirmed by the flow waveform on the graphics monitior.To eliminate auto peep the respiratory rate was reduced to 12 breaths per minute. Each participant was asked if they could hear flow at the end of expiration (i.e auto-peep) by ausculating the expiratory valve with the bell of the stethoscope.
RESULTS: A total of 280 observation were made by the 70 examinees In total PEEPi was present in 153 observations- this was correctly detected in 141 cases (92%) and was not detected in 12 cases (8%). In total PEEPi was absent 127 observations- this was correctly detected in 121 cases ( 95% ) and was falsely detected in 6 cases (5%).
CONCLUSION: We conclude that this method is a very good at detecting and excluding auto-PEEP.
CLINICAL IMPLICATIONS: Many ventilators do not have on line waveform tracing capability. Particularly, in circumstances where ventilators are not routinely situated and only used in emergency circumstances.This technique could replace clinical assessment whose utility is compromised by poor negative predictive value.
DISCLOSURE: Brian Casserly, None.