Effects of awake prone positioning in non-intubated spontaneously breathing COVID-19 patients requiring high flow oxygen therapy in high dependency unit (HDU): An observational study

The pandemic of coronavirus disease 2019 (COVID-19) threatens to overload intensive care units (ICUs) due to large numbers of patients requiring respiratory support [1]. Patients with coronavirus disease 2019 are at greater risk of developing acute respiratory distress syndrome (ARDS) [2]. To reduce the burden on the ICU and prevent ICU transfers, early awake prone positioning was done in patients with COVID-19 requiring supplementary oxygen support [3]. Nowadays prone positioning is mostly used in the ICU, particularly for mechanically ventilated patients with ARDS [3, 4]. It also helps in preventing ventilator-induced lung injury [5, 6]. Prone positioning increases oxygenation by recruiting dorsal lung regions and draining airway secretions, and it improves gas exchange and survival in ARDS. 3 The overall lung ventilation from dorsal to ventral areas is more homogenous in prone position, thus improving oxygenation. The reports of application of prone position in spontaneously breathing non-intubated patients is limited to few case reports. Theoretically, mechanisms of improving oxygenation in non-intubated spontaneously breathing patients should be same as for mechanically ventilated patients [7]. In this observational study we describe 26 patients with COVID-19 pneumonia requiring oxygen supplementation who underwent awake prone positioning. We describe the effects of early awake prone positioning on oxygenation, breathing patterns, and haemodynamics.

This observational study was conducted in the High Dependency Unit (HDU) of the Chest Disease Hospital associated with Government Medical College, Srinagar, Jammu, and Kashmir. The study was conducted on spontaneously breathing non-intubated patients with COVID-19 with hypoxemic respiratory failure requiring supplemental oxygen therapy after obtaining clearance from the Ethical Committee and informed consent from patients.

The 26 patients (12 males and 14 females) non-intubated spontaneously breathing with SPO₂ <94% on 0.4 FiO₂ were treated with prone positioning. Patients’ characteristics are summarised in Table 1 and Figure 1. Mean age was 51.5 ± 14.89. Mean HDU length of stay was 3 days. One patient required intubation and was shifted to intensive care unit (ICU), 25 patients were discharged from HDU. Among 26 patients, 25 tolerated prone positioning sessions during the study period. Mean hours of prone positioning were 19.4 ± 2.06 hrs (range 16-22 hours).

Figure 1

Outcomes

1

Oxygen saturation (SPO₂): Effect of prone positioning on SPO₂ is shown in Table 3 and Figure 3. Mean SPO₂ increased from 85.73 ± 6.13% (pre prone positioning) to 88.58 ± 3.89 % (during prone positioning) and it further increased to 91.19 ± 3.18 % after completion of various prone positioning sessions. There was significant increase in SPO₂ after various prone positioning sessions with P-value of 0.003 (pre prone vs. during prone) and 0.001 (pre prone vs. post prone).

Figure 2

Figure 3

Table 2

Respiratory and haemodynamic parameters on admission

Parameter		Mean	SD	Range
Respiratory parameters	SPo2	83.6	6.13	60-90
	Pao2	50.4	6.68	33-62
	RR	30.6	2.82	24-36
Haemodynamic parameters	MAP	78.4	9.57	65-108
	HR	109.2	12.01	90-130

Table 3

Respiratory and haemodynamic parameters during different study periods

Parameter	Pre		Prone		Post		P-value
	Mean	SD	Mean	SD	Mean	SD	Pre vs. Prone	Pre vs. Post
SPo2	85.73	6.86	88.58	3.89	91.19	3.18	0.003*	<0.001*
Pao2	53.15	6.00	58.38	6.81	64.23	6.96	<0.001*	<0.001*
RR	28.58	3.89	26.27	4.98	23.15	5.83	0.002*	<0.001*
MAP	76.88	8.02	78.54	10.77	77.54	6.53	0.304	0.672
HR	108.31	12.77	102.92	12.96	94.81	15.21	0.004*	<0.001*

Statistically Significant Difference (P-value<0.05); P-value by Repeated Measures ANOVA

2

Partial pressure of oxygen (PaO₂): Effect of prone positioning on PaO₂ is shown in Table 3 and Figure 4. Mean PaO₂ increased from 53.15 ± 6.0 mmHg (pre prone positioning) to 58.38 ± 6.8 mmHg (during prone positioning) and it further increased to mean value of 64.23 ± 6.96 mmHg after completion of various prone positioning sessions. There was significant increase in values of PaO₂ after completion of various prone positioning sessions, with P-value of <0.001.

Figure 4

3

Respiratory rate: Effects of prone positioning on respiratory rate is shown in Table 3 and Figure 5. Mean respiratory rate decreased from 28.58 ± 3.89 breaths per minute (pre prone positioning) to 26.27 ± 4.98 breaths per minute (during prone positioning) and it further decreased to mean value of 23.15 ± 5.83 breaths per minute after completion of various prone positioning sessions. There was a significant decrease in respiratory rate of patients after completion of various prone positioning sessions, with P-value of 0.002 (pre vs. prone) and <0.001 (pre vs. post).

Figure 5

4

Haemodynamic parameters: There was not any significant change in mean arterial pressure during various study periods. There was significant decrease in heart rate (Table 3).

Table 4

Respiratory and haemodynamic parameters at discharge

Parameter		Mean	SD	Range
	SPo2	92.6	2.38	88-96
Respiratory parameters	Pao2	67.3	4.82	56-78
	RR	21.6	3.46	17-30
Haemodynamic parameters	MAP	77.9	8.45	69-108
	HR	91.4	7.31	70-110

Prone positioning in acute respiratory failure has been proven to be an effective therapy in improving oxygenation by decreasing ventilation perfusion mismatch, [8] and promoting recruitment of nonaerated dorsal regions of the lung [9, 10]. Furthermore, prone positioning also reduces ventilator-induced lung injury (VILI) [5, 6]. Recent evidence implies that patients with severe ARDS have increased chances of survival after prone positioning [3, 4]. Most of these observational studies are consistent with mechanically ventilated patients. Data available for prone positioning in awake spontaneously breathing non-intubated patients is scarce [7]. Autopsy findings of critically ill COVID-19 patients in Wuhan, China, showed that exudation, infiltration of macrophages, and fibrosis of lung parenchyma were the main pathological features. The recent clinical data showed that symptoms of COVID-19 ARDS are atypical [11]. This is the reason why these patients partially respond to lung recruitment manoeuvres and higher positive end-expiratory pressure (PEEP) strategies when mechanically ventilated.

Therefore, application of prone positioning for mobilisation of copious secretions with forward drainage towards the central airways for clearance might improve the lung mechanics [12].

In our observational study of 26 patients in HDU we found that after multiple sessions of prone positioning in awake non-intubated spontaneously breathing patients there is a significant improvement in oxygenation of these patients and decrease in work of breathing. Results of our study were consistent with a study done by Scaravilli et al. [7] which described the application of prone positioning in 15 awake spontaneously breathing non-intubated patients with hypoxemic acute respiratory failure. We also observed that patient comfort levels increased after multiple prone positioning sessions. Our results are also consistent with the study done by Feltracco et al. [13, 14] in a cohort of 5 patients using non-invasive ventilation (NIV). Valter et al. [15] also documented application of prone positioning in 4 patients with acute respiratory failure and observed improvement in oxygenation during prone positioning.

In our observational study we demonstrated that early awake prone positioning in spontaneously breathing non-intubated COVID 19 patients with hypoxemic acute respiratory failure is feasible, safe, and associated with a significant benefit for oxygenation.