Objective: The purpose of this study was to investigate whether changes in breathing pattern, neuromuscular drive (P0.1), and the work involved in breathing might help to set the individual appropriate level of pressure support ventilation (PSV) in patients with acute respiratory failure (ARF) requiring ventilatory assistance.
Design: A prospective, interventional study.
Setting: An 8-bed multidisciplinary intensive care unit (ICU).
Patients: Ten patients with ARF due to adult respiratory distress syndrome (ARDS), sepsis or airway infection were included in the study. Chronic obstructive pulmonary disease (COPD) patients with acute exacerbation were excluded. None of these patients was in the weaning process.
Interventions: We found a level of pressure support able to generate a condition of near-relaxation in each patient, as evidenced by work of breathing (WOB) values close to 0 J/1. This level was called PS 100 and baseline physiological measurements, namely, breathing pattern, P0.1 and WOB were obtained. Pressure support was then reduced to 85%, 70%, and 50% of the initial value and the same set of measurements was obtained.
Measurements and results: Flow (V) was measured by a flow sensor (Varflex) positioned between the Y-piece of the breathing circuit and the endotracheal tube. Tidal volume was obtained by numerical integration of the flow signal. Airway pressure (P(aw)) was sampled through a catheter attached to the flow sensor. Esophageal pressure (Pes) was measured with a nasogastric tube incorporating an esophageal balloon. The esophageal balloon and flow and pressure sensors were connected to a portable monitor (CP 100 Bicore) that provided real-time display of flow, volume, P(aw) and Pes tracings and loops of Pes/V, P(aw)/V and V/V relationships. The breathing pattern was analyzed from the flow signal. Patient work of breathing (WOB) was calculated by integration of the area of the Pes/V loop. Respiratory drive (P0.1) was measured at the esophageal pressure change during the first 100 ms of a breath, by the quasi-occlusion technique. When pressure support was reduced, we found that the respiration rate significantly increased from PS 100 to PS 85, but varied negligibly with lower pressure support levels. Tidal volume behaved in a similar way, decreasing significantly from PS 100 to PS 85, but hardly changing at PS 70 and PS 50. In contrast, WOB and P0.1 increased progressively with decreasing pressure support levels. The changes in WOB were significant at each stage in the trial, whereas P0.1 increased significantly from PS 100 at other stages. Linear regression analysis revealed a highly positive, significant correlation between WOB and P0.1 at decreasing PSV levels (r = 0.87), whereas the correlation between WOB and ventilatory frequency was less significant (r = 0.53). No other correlation was found.
Conclusions: During pressure support ventilation, P0.1 may be a more sensitive parameter than the assessment of breathing pattern in setting the optimal level of pressure support in individual patients. Although P0.1 was measured with an esophageal balloon in the present study, non-invasive techniques can also be used.