Accuracy of intra-oesophageal dynamic pressure probes, and their validity in determining dynamic intra-pleural pressure

Abstract

Oesophageal pressure (Res) is often measured to estimate pleural pressure (Ppl) for the calculation of respiratory system elastance and resistance. High-fidelity dynamic Ppl estimation requires that Ppl waveforms be transmitted unchanged both across the Ppl-Pes tissue barrier and across Pes catheter-manometers. In this study the frequency responses of liquid- and gas-filled catheter manometers used in clinical practice were examined in detail using the in vitro sinewave technique. The assumption that fluid-filled catheter-manometer frequency responses fit a second order system was tested by comparing second order curvefits to measured curves, An acute lung injury (ALI) model of human respiratory disease was developed in monkeys. In health and A L direct Ppl and Pes were measured simultaneously to determine the Ppl-Pes tissue barrier amplitude frequency response. The relevant bandwidth of dynamic Pes waveforms was determined. It is found that liquid-filled feeding catheters measure dynamic Pes within a 5% error up to a maximum respiratory rate (FRR) of 82 breaths/min and are suitable for use only in subjects with low-frequency respiratory mechanics. FRR differences exist between French Gauge sizes and differing catheter brands: French Gauge size is a poor predictor of Pes measurement suitability. Infant air-balloon catheters' FRR is up to 148 breaths/min. They possess superior frequency response characteristics compared to matching liquid-filled catheters, have lower frequency response variability within catheter samples, and are suited to dynamic Pes measurements during high-frequency respiratory mechanics. ii Frequency responses of fluid-filled feeding catheters employed in Pes catheter-manometers do not adequately fit second order systems, casting doubt on the validity of applying second order mathematical models to predict cathetermanometer behaviour from stap-responses. Decreased dynamic lung compliance, reduced alveolar gas exchange and diffuse alveolar capillary leak similar to that of comparable humans evolves following oleic-acid administration in monkeys; the model is suitable for evaluation of pulmonary mechanics and gas exchange during ALI. The Ppl-Pes tissue barrier has a uniform frequency response within the bandwidth of conventional Pes waveforms in healthy or diseased lungs, and does not attenuate Ppl-Pes waveform transmission between 1 - 40 Hz. At Pes frequencies higher than conventional regions of clinical interest the Ppl-Pes barrier resonates, is pressure amplitude dependent at low pressure offsets, and altered by ALI. During conventional ventilation for ALI, Pes-manometers require a uniform frequency response up to 8.5 Hz to achieve a s 5% in vivo Pes waveform measurement error. These findings advance the accuracy of pulmonary function studies in high frequency respiratory mechanics, such as conventional infant ventilation or high frequency ventilation.