Figure 1.
Event categories and breathing effort according to each category and mode of ventilation. (A) Recording on pressure-control ventilation illustrating different event categories. From top to bottom flow, airway pressure, esophageal pressure (Peso), and muscular pressure (Pmus) tracings are shown. The start and end of each respiratory cycle (breath) are defined by the zero crossing of flow (vertical dashed lines). An inspiratory effort is defined by the positive deflection in Pmus (horizontal black lines). The start and end of each effort are marked with blue and red dots, respectively. Reverse-triggering (RevT) events are seen on breaths #1, #2, #4, #6, and #8, interposed with mandatory passive events (#3, #5, and #7) in a 1:2 ratio. The first RevT event (breaths #1 and #2) results in breath stacking (BS) (star). A series of mandatory passive events is seen from breaths #9–#12, where Pmus is zero (flat line). A patient-triggered (PatT) event is seen at the end (breath #13). Instantaneous calculation of Pmus is represented with vertical dashed lines on breath #6 in the Peso tracing (i.e., difference between chest wall recoil pressure [Pcw] and Peso). Pcw is the expected increase in Peso during a hypothetical passive inflation with the same Vt, calculated as the product of chest wall elastance and Vt (see similarity between Pcw in breath #6 and Peso in breaths #5 and #7, where inflation is passive but Vt is smaller). The black vertical arrow in the Pmus tracing (breath #2) illustrates the maximum positive deflection in Pmus during each inspiratory effort (Pmus swing), and the gray area (breath #4) illustrates the pressure–time product. (B) Pmus swings according to the type of event and mode of ventilation in patients with acute hypoxemic respiratory failure. Median and interquartile range of Pmus swing in healthy subjects breathing at rest without inspiratory assistance are displayed as gray horizontal full and dashed lines, respectively (comparison with healthy subjects at rest: RevT without BS and PaT on pressure support, P > 0.05; PatT in A/C, ineffective efforts in assist-control mode [A/C] and pressure support mode [PSV], P < 0.05, PaT during proportional-assist ventilation with adjustable gain factors [PAV+], P < 0.01; passive, RevT, and PaT with BS on A/C, P < 0.001). Events with BS are seen in red, showing a higher Pmus swing versus corresponding events without BS. Numbers of events were as follows: A/C mandatory breath (Mand.) passive, n = 10,017; A/C Mand. RevT without BS, n = 1,387; A/C Mand. RevT with BS, n = 455; A/C ineffective effort (IE), n = 27; A/C PatT without BS, n = 1,334; A/C PatT with BS, n = 58; PSV IE, n = 23; PSV PatT, n = 3,212; and PAV+ PatT, n = 2,572. Twelve breathing efforts with Pmus swing higher than 60 cm H2O were removed to improve data visualization. (C) Pressure–time product per minute for recordings during different modes of ventilation and containing different types of events in patients with acute hypoxemic respiratory failure. The reference median and interquartile range of PTPmus per minute for recordings of healthy subjects breathing at rest without inspiratory assistance are displayed as gray horizontal full and dashed lines, respectively (comparison with healthy subjects at rest: A/C with RevT, PSV, and PAV+, P > 0.05; A/C with only PaT, P < 0.05). Numbers of recordings (breathing efforts) were as follows: A/C with RevT, n = 18 (n = 2,333); A/C only PatT, n = 4 (n = 887); PSV PatT, n = 21 (n = 3,029); and PAV+ PatT, n = 24 (n = 2,564). †Sixty-three efforts out of 10,017 mandatory passive events were erroneously detected by the algorithm because of an artifact in the Peso signal in passive breaths (rate of false positives < 1%), resulting in measurement of effort (Pmus swing). Paw = airway pressure; PTPmus = pressure–time product measured according to the algorithm; PTPmus[mean] = mean pressure–time product measured according to the algorithm for one recording.