Awake Prone Position During Noninvasive Ventilation in COVID-19: Doubts Need Addressal

To the Editor:

The study by Oliveira et al¹ is important for the reinforcement of the concept prone position and COVID-19, especially prone position for hypoxemic, non-intubated patients with COVID-19. We consider that this study has some aspects that are not well clarified that we believe are interesting to be able to analyze the impact on mortality.

First, from a methodology point of view, definitions and exclusion criteria need clarification in some aspects. The type of ventilation showed no statistically significant association and represented an independent factor. There is no clear definition of “type of ventilation,” and there is no obvious differentiation between oxygenation systems (high-flow nasal cannula [HFNC], CPAP, oxygen masks, nasal cannula) and noninvasive ventilation [NIV]), as if they were considered the same thing. The excluded/did not tolerate/nonresponders to prone subgroup needs more clarifications. In our experience, for example, we also use the prone position accompanied by mild sedation that increases adherence and the possibility of having a favorable result² especially in patients who do not tolerate. The prone position maneuver itself: During the maneuver, 33 (80.4%) subjects were treated with reservoir masks, 15 (10.5%) with HFNC, 2 (4.5%) with NIV, and 1 (4.5%) with a nasal cannula. The choices of initial respiratory treatment modalities are not clear. We think that a computed tomography radiologic description of the kind of parenchymal alterations is important for the choice of initial strategy of treatment. In the consolidation pattern,³ the response is unpredictable, especially if the thickened areas are peripheral and posterior. The pronation may interrupt the process of progressive basilar atelectasis and rapid deterioration.⁴ It should be noted that the response to oxygenation during prone position is highly variable but seems time dependent, and the improvement in oxygenation may be seen even after few days of initiation of prone position.⁵ Voggenreiter et al⁶ reported that among subjects with acute lung injury and ARDS with each cycle of proning there was a significant increase in oxygenation (approximately 4 mm Hg). These results suggest that initial nonresponders may show response on subsequent attempts. However, the authors of the current study¹ used only the first trial of prone position for a median duration of 120 min for classification of nonresponders, which might have wrongly classified a significant number of subjects as nonresponders.

Data collection and proning protocol: It’s not explained why data are collected only after the first session of pronation maneuver. We also use continue monitoring for the few but, nevertheless, present episodes of pneumomediastinum,⁷ even spontaneous ones, monitoring with particular attention to painful symptoms and sudden tachycardia. Moreover, readers are not provided any information regarding the proning protocol followed in the study. The inclusion of the details of the proning protocol might be useful for clinicians for day-to-day practice as well as for researchers for validation. Unfortunately, an approved protocol for awake proning for management of hypoxemia is lacking in the literature. Recently, a protocol was published by Bamford et al⁸ that advocates 30 min to 2 h lying fully prone (bed flat), 30 min to 2 h lying on right side (bed flat), 30 min to 2 h sitting up (30–60°) by adjusting head of the bed, 30 min to 2 h lying on left side (bed flat), and 30 min to 2 h lying prone and again continue to repeat the cycle.

Second, regarding the nasal high flow-predictive factor, ROX index was not a predictive factor for the success of the maneuver. The authors consider that the ROX index before the maneuver in responders and nonresponders was 3.8 (3.3–5.4) and 3.4 (2.8–4.4) as not different. What does that mean in this observation? Can it be related to time from the onset of disease/hypoxemia within that prone position should be initiated? As shown by Nakos et al,⁵ initiation of prone position within 3 d resulted in improvement in the P_aO2/F_IO2, whereas initiation after 3 d did not improve P_aO2/F_IO2. There is no clarity as to the optimal period from the onset of disease/hypoxemia within that prone position should be initiated. Third, regarding the hospital mortality observed, the causes are not adequately clarified if they are directly associated with a “protective” effect of the prone or it is simply a maneuver that avoids intubation due to hypoxemia without affecting hospital mortality. This aspect should be better clarified by the authors.

A recent review⁹ of 8 studies that brought together 199 subjects with COVID-19 respiratory failure, not treated with mechanical ventilation but placed in the prone position, highlighted a substantial ineffectiveness of this maneuver in these types of patients, contrary to what was demonstrated in this paper. Therefore, further studies are needed to verify its effectiveness in this subtype of patients.

• Copyright © 2022 by Daedalus Enterprises