Awake prone position reduces work of breathing in patients with COVID-19 ARDS supported by CPAP

Abstract

Background: The use of awake prone position concomitant to non-invasive mechanical ventilation in acute respiratory distress syndrome (ARDS) secondary to COVID-19 has shown to improve gas exchange, whereas its effect on the work of breathing remain unclear. The objective of this study was to evaluate the effects of awake prone position during helmet continuous positive airway pressure (CPAP) ventilation on inspiratory effort, gas exchange and comfort of breathing.

Methods: Forty consecutive patients presenting with ARDS due to COVID-19 were prospectively enrolled. Gas exchange, esophageal pressure swing (ΔPes), dynamic transpulmonary pressure (dTPP), modified pressure time product (mPTP), work of breathing (WOB) and comfort of breathing, were recorded on supine position and after 3 h on prone position.

Results: The median applied PEEP with helmet CPAP was 10 [8-10] cmH₂O. The PaO₂/FiO₂ was higher in prone compared to supine position (Supine: 166 [136-224] mmHg, Prone: 314 [232-398] mmHg, p < 0.001). Respiratory rate and minute ventilation decreased from supine to prone position from 20 [17-24] to 17 [15-19] b/min (p < 0.001) and from 8.6 [7.3-10.6] to 7.7 [6.6-8.6] L/min (p < 0.001), respectively. Prone position did not reduce ΔPes (Supine: - 7 [- 9 to - 5] cmH₂O, Prone: - 6 [- 9 to - 5] cmH₂O, p = 0.31) and dTPP (Supine: 17 [14-19] cmH₂O, Prone: 16 [14-18] cmH₂O, p = 0.34). Conversely, mPTP and WOB decreased from 152 [104-197] to 118 [90-150] cmH₂O/min (p < 0.001) and from 146 [120-185] to 114 [95-151] cmH₂O L/min (p < 0.001), respectively. Twenty-six (65%) patients experienced a reduction in WOB of more than 10%. The overall sensation of dyspnea was lower in prone position (p = 0.005).

Conclusions: Awake prone position with helmet CPAP enables a reduction in the work of breathing and an improvement in oxygenation in COVID-19-associated ARDS.

Keywords: ARDS; Awake prone position; COVID-19; Continuous positive airway pressure; Helmet CPAP; Work of breathing.

Fig. 1

Lung Computed Tomography. Lung Computed Tomography images (at the level of the carina), extracted from lung CT of three different patients, representative of the ARDS severity of the study population

Fig. 2

A Change in PaO₂/ FiO₂ Ratio from supine to prone position. Individual patient measurements are represented by dots (blue representing supine and red prone position, respectively), thin lines connect pairs of individual patient measurements (supine—prone), and thick horizontal lines display the median. Significance levels: P value ≥ 0.05—NS, < 0.05—*, < 0.01—**, < 0.001—***. B Correlation between the change in PaO₂/ FiO₂ Ratio from supine to prone position and its respective value in supine position. The scatter-plot represents individual patient measurement-pairs, the black line displays the fitted linear regression, and the shaded gray area depicts its 95% Confidence Interval. ρ—Pearson correlation coefficient, p—P value

Fig. 3

A Change in Work of Breathing from supine to prone position. Individual patient measurements are represented by dots (blue representing supine and red prone position, respectively), thin lines connect pairs of individual patient measurements (supine—prone), and thick horizontal lines display the median. Significance levels: P value ≥ 0.05—NS, < 0.05—*, < 0.01—**, < 0.001—***. B Correlation between the change in Work of Breathing from supine to prone position and its respective value in supine position. The scatter-plot represents individual patient measurement-pairs, the black line displays the fitted linear regression, and the shaded gray area depicts its 95% Confidence Interval. ρ—Pearson correlation coefficient, p—P value

Fig. 4

Changes in Esophageal Pressure Swing (A), dynamic Transpulmonary Pressure (B) and modified Pressure–Time Product (C) between supine and prone position. Individual patient measurements are represented by dots (blue representing supine and red prone position, respectively), thin lines connect pairs of individual patient measurements (supine—prone), and thick horizontal lines display the median. Significance levels: P value ≥ 0.05—NS, < 0.05—*, < 0.01—**, < 0.001—***