The role of prone positioning in patients with SARS-CoV-2-related respiratory failure in non

Abstract

Background:

Prone positioning (PP) is an established and commonly used lung recruitment method for intubated patients with severe acute respiratory distress syndrome, with potential benefits in clinical outcome. The role of PP outside the intensive care unit (ICU) setting is debated.

Objectives:

We aimed at assessing the role of PP in death and ICU admission in non-intubated patients with acute respiratory failure related to COronaVIrus Disease-19 (COVID-19) pneumonia.

Design:

This is a retrospective analysis of a collaborative multicenter database obtained by merging local non-interventional cohorts.

Methods:

Consecutive adult patients with COVID-19-related respiratory failure were included in a collaborative cohort and classified based on the severity of respiratory failure according to the partial arterial oxygen pressure to fraction of inspired oxygen ratio (PaO₂/FiO₂) and on clinical severity by the quick Sequential Organ Failure Assessment (qSOFA) score. The primary study outcome was the composite of in-hospital death or ICU admission within 30 days from hospitalization.

Results:

PP was used in 114 of 536 study patients (21.8%), more commonly in patients with lower PaO₂/FiO₂ or receiving non-invasive ventilation and less commonly in patients with known comorbidities. A primary study outcome event occurred in 163 patients (30.4%) and in-hospital death in 129 (24.1%). PP was not associated with death or ICU admission (HR 1.17, 95% CI 0.78–1.74) and not with death (HR 1.01, 95% CI 0.61–1.67) at multivariable analysis; PP was an independent predictor of ICU admission (HR 2.64, 95% CI 1.53–4.40). The lack of association between PP and death or ICU admission was confirmed at propensity score-matching analysis.

Conclusion:

PP is used in a non-negligible proportion of non-intubated patients with COVID-19-related severe respiratory failure and is not associated with death but with ICU admission. The role of PP in this setting merits further evaluation in randomized studies.

Keywords

ARDS COVID-19 hypoxemic respiratory failure non-invasive mechanical ventilation prone positioning SARS-CoV2

Introduction

The Severe Acute Respiratory Syndrome Coronavirus-2 (SARS-CoV-2)-related disease (COVID-19) is characterized by a wide spectrum of respiratory manifestations, ranging from no or flu-like syndromes to severe acute respiratory distress syndrome (ARDS).^1–4 It is estimated that prior to the introduction of vaccines, ARDS, as defined by the Berlin criteria, developed in about 42% of patients presenting with COVID-19 pneumonia, with rate as high as 61–81% in those with critical disease⁵ and was associated with a mortality rate nearing 50%.⁶ In vaccinated patients, these rates are probably lower but still not negligible.⁷

The severity of hypoxemia and the high mortality reported in patients with COVID-19 who required early intubation have led many clinicians to adopt lung recruitment methods to improve oxygenation, gas exchange, and, potentially, clinical outcome.⁸ Prone positioning (PP) is an established and commonly used lung recruitment method for intubated patients with severe non-COVID-19-related ARDS, with potential benefits in clinical outcome.⁹ Based on the pathophysiological rationale and the observed clinical benefits in patients with severe ARDS, the use of PP was extended to both intubated and not intubated COVID-19 patients in whom the presence of respiratory failure and the lack of effective treatment suggested a poor prognosis.¹⁰ In some patients, PP was used to avoid intubation and overcome intensive care unit (ICU) surge capacity during the COVID-19 outbreak. However, the clinical benefit of PP in non-intubated COVID-19 patients is controversial.¹¹

This multicenter non-intervention study aims at assessing the role of PP with respect to death or ICU admission at 30 days, in non-intubated patients with acute respiratory failure related to COVID-19 pneumonia. In addition, we aimed at assessing the use of PP in this setting.

Methods

Study design and population

Consecutive adult patients with acute respiratory failure related to COVID-19 pneumonia who had been admitted at the Department of Internal and Cardiovascular Medicine – COVID-19 of the University of Perugia or at the Department of Medicine at the University of Pisa from October 2020 to May 2021 were included in retrospective non-interventional cohorts; these cohorts were subsequently merged in a collaborative database. Inclusion criteria for the present study were as follows: (1) diagnosis of COVID-19 confirmed by real-time polymerase chain reaction (RT-PCR) testing, (2) pneumonia detected by chest X-ray or computed tomography (CT) or lung ultrasonography, and (3) new-onset respiratory failure.

Exclusion criteria were (1) contraindication to PP including any of the following: recent abdominal or thoracic surgery or trauma, facial, pelvic, or spine fracture, skeletal deformities such as severe kyphoscoliosis and severe limb contractures that could interfere with PP, and neurological issues making PP impossible or life-threatening; (2) need for intubation at hospital admission; (3) intubation for COVID-19 before or immediately at admission to the Emergency Department before inclusion in the study; (4) hemodynamic instability requiring vasopressor, and (5) pregnancy.

Decision on the use of PP, respiratory support, and drug therapy was left to the discretion of the attending physician. For the purpose of this study, non-invasive ventilation (NIV) included continuous positive airway pressure (CPAP), standard NIV, or a combination of the two methods. CPAP was mainly delivered by the use of helmets. NIV was initially set with the aim to provide a tidal volume on predicted body weight ratio varying between 6 and 8 ml/kg with a combination of positive end-expiratory pressure and inspired oxygen fraction (FiO₂) to get a peripheral oxygen saturation between 92% and 98%. NIV was delivered through full-face or oro-nasal facial mask connected via a double-tube circuit to a life-support ventilator (VIVO60 or VIVO65 Breas Medical, Sweden). High-flow nasal oxygen (HFNO) was not considered as NIV.

Patients were classified based on the severity of respiratory failure at admission. Severe acute hypoxemic respiratory failure was defined by a ratio of partial pressure of arterial oxygen (PaO₂) to fraction of inspired oxygen (FiO₂) (PaO₂/FiO₂) ⩽300 mmHg.

Diagnosis of COVID-19-related ARDS was made when patients met the Berlin criteria: (1) acute hypoxemic respiratory failure; (2) presentation within 1 week of worsening respiratory symptoms; (3) bilateral airspace disease on chest X-ray, computed tomography or ultrasound that is not fully explained by effusions, lobar or lung collapse, or nodules; and (4) cardiac failure not the primary cause of acute hypoxemic respiratory failure.⁴

Patients were classified according to clinical severity based on the quick Sequential (sepsis-related) Organ Failure Assessment (qSOFA) score.¹²

In addition, patients were classified as eligible or denied for ICU admission according to local assessment.

The study is reported according to the ‘Strengthening the Reporting of Observational Studies in Epidemiology (STROBE)’ statement guidelines for observational cohort studies (Additional file 1) and was approved by the local ethic review boards (Comitato Etico Regione Umbria, University of Perugia, protocol no. 20-05 SARS-CoV-2). According to local regulations, each patient provided informed consent for inclusion in the individual databases.

Study outcomes

The primary study outcome was the composite of in-hospital death or ICU admission occurring within 30 days from hospitalization. Decision concerning the need for ICU admission was left to the discretion of the attending physician. Secondary study outcomes were the individual components of the primary outcome as it follows death in the overall study population and ICU admission in patients not denied for ICU. Both secondary outcomes were assessed at 30 days from hospitalization.

Study procedures

Proning maneuver was performed by a trained staff of nurses and physicians in fully awake or consciously sedated patients. The duration of pronation lasted from a minimum of 4 to a maximum of 10 h and was repeated two to three times a day – depending on the duration of the individual PP periods – after a period of supine positioning. For patients who were uncompliant to awake PP or severely tachypneic (respiratory rate over 30 breaths per minute), sedation was applied. Conscious sedation was performed using dexmedetomidine, morphine, or benzodiazepine in multiple bolus or continuous infusions.

Data collection

For all included patients, the following were collected: age, sex, comorbidities (cardiovascular diseases, cerebrovascular diseases, history of chronic or obstructive pulmonary disease, systemic arterial hypertension, diabetes, chronic kidney disease), vital parameters (heart rate, respiratory rate), the PaO₂/FiO₂ ratio before, after, and during the prone position, partial oxygen pressure (PaO₂) and partial carbon dioxide pressure (PaCO₂) before oxygen therapy, clinical signs (dyspnea, hemoptysis), laboratory findings (hemoglobin, platelets, neutrophil–lymphocyte ratio, creatinine), and use of supplemental oxygen therapy or non-invasive respiratory support (NIV or CPAP) at admission and during observation.

Statistical analysis

This is a hypothesis-generating study as less data exist on the use of PP outside the ICU. Assuming an incidence of death or ICU admission within 30 days of 20% and the need to adjust the analysis for age, the severity of respiratory failure according to four categories (based on PaO₂/FiO₂: >300, 300–200, 200–100, <100 mmHg), and the clinical severity into two classes (based on qSOFA: <2 and ⩾2), we estimated a sample size of at least 500 patients to get powered information on the role of PP in the risk for the study outcome.

Frequency data are presented as proportions with 95% confidence intervals (CIs). Continuous data are shown as means with standard deviations (SDs). Student’s t-test and the chi-square test or Fisher’s exact test were used for comparisons of continuous and nominal variables, respectively.

Cox proportional regression model was used to assess the role of PP in the occurrence of death or ICU admission during the hospital stay and within 30 days from admission in patients with COVID-19-related respiratory failure. To adjust for other predictor of study outcome events, multivariable analyses were constructed from the set of significant (p < 0.10) univariable predictors at entry and of covariates of clinical interest. Age was considered as continuous variable; PaO₂/FiO₂ and qSOFA were considered as ordinal variables. PaO₂/FiO₂ was categorized according to the Berlin definition (PaO₂/FiO₂ >300, 300–200, 200–100, and ⩽100 mmHg);⁴ qSOFA was categorized into three risk categories (no item or one item = 0, two items or more = 1, unknown = 3). For both the variables, the reference group was the category with milder disease (PaO₂/FiO₂ >300 mmHg and qSOFA = 0). Multivariable analyses are reported as hazard ratio (HR) with 95% CIs and p values.

A propensity score-matching analysis was performed to create similar treatment groups (PP versus no PP) with respect to observed patients’ features. The propensity score was calculated on the characteristics significantly different between patients¹³ managed with or without PP. One patient treated with PP was matched to one patient not treated through optimal matching, determining matched samples with the smallest average within-pair absolute difference in propensity score.^14–16 The balance of measured covariates between matched cohort was assessed using standardized mean differences. Cox models with robust estimates were used to account for the clustering within matched sets. HRs and their 95% CI were reported for the study outcomes.¹⁷ A two-sided p-value less than 0.05 was considered statistically significant.

Analyses were performed with SPSS, version 26 and R 3.6.2 (package MatchIt).

Results

From October 2020 to May 2021, 583 patients with SARS-CoV-2-related respiratory failure were admitted at the study centers (Figure 1). After the exclusion of 47 patients because of the need for intubation at hospital admission, 536 patients were finally included in the study. Main patient features are reported in Table 1.

Figure 1.

Study flow diagram.

Table 1.

Characteristics of study patients overall and by use of PP.

Characteristic	Patients
	Overall study patients(n = 536)	PP
	Overall study patients(n = 536)	Yes(n = 114)	No(n = 422)
Age, M ± SD	69 ± 15 (17–98)	63.2 ± 13	70.4 ± 15
Male, n (%)	334 (62.3)	86 (75.4)	174 (41.2)
PaO₂/FiO₂ categories (mmHg)
PaO₂/ FiO₂, M ± SD	260 ± 95	225 ± 78	270 ± 97
>300	182 (34.6)	19 (16.7)	163 (38.7)
300–200	215 (40.1)	55 (48.2)	160 (38.0)
200–100	110 (20.5)	31 (27.2)	79 (18.8)
<100	28 (5.2)	9 (7.9)	19 (4.5)
Comorbidities, n (%)
COPD	63 (11.8)	7 (6.1)	56 (13.3)
Diabetes	104 (19.4)	17 (15.0)	87 (20.6)
Systemic hypertension	299 (55.8)	66 (57.9)	233 (55.2)
Cardiovascular disease	178 (33.2)	26 (22.8)	152 (36.0)
Cerebrovascular disease	54 (10.1)	6 (5.3)	48 (11.4)
Chronic kidney disease	49 (9.1)	5 (4.4)	44 (10.4)
Cancer	70 (13.1)	11 (9.6)	59 (14.0)
Denied for ICU	112 (20.8)	6 (5.2)	106 (25.1)
No. of comorbitidies, n (%)
0	153 (28.5)	43 (37.7)	110 (26.0)
1–2	236 (44.1)	48 (42.1)	188 (44.5)
>2	147 (27.4)	23 (20.2)	124 (29.4)
NIV, n (%)	187 (34.9)	81 (71.0)	106 (25.1)
qSOFA (n = 419), n (%)
0 items	232 (43.2)	49 (43.0)	183 (43.3)
1 item	162 (30.2)	127 (89.7)	35 (8.29)
2 items	20 (3.7)	19 (16.6)	1 (0.2)
3 items	4 (0.7)	2 (1.7)	2 (0.4)

COPD, chronic obstructive pulmonary disease; ICU, intensive care unit; NIV, non-invasive ventilation; PaO₂/FiO₂, ratio of partial pressure of arterial oxygen (PaO₂) to fraction of inspired oxygen (FiO₂); PP, prone positioning; qSOFA, quick sequential organ failure assessment; SD, standard deviation.

The mean age of study patients was 69 ± 15 years, ranging 17–98 years. Overall, 153 patients (28.5%) had no known comorbidities at hospital admission, 236 (44.1%) had one or two, and 147 patients (27.4%) had more than two known comorbidities at hospital admission.

At admission, PaO₂/FiO₂ was > 300 mmHg in 183 patients (34.0%), between 300 and 200 mmHg in 215 (40.2%), between 200 and 100 mmHg in 110 (20.6%), and <100 mmHg in 28 (5.2%). Blood gas parameters before the starting of oxygen therapy were available in 284 patients as the remaining were tested on supplemental oxygen. Before the starting of oxygen therapy median, PaO₂ was 63 (IQR = 56–76) mmHg and median PaCO₂ was 33 (IQR = 29–37) mmHg.

Overall, 373 patients received steroids treatment during the hospital stay (69.6%). The use of antiviral and immunomodulatory therapies was at the discretion of the attending physician.

NIV was used for the management of respiratory failure in 187 patients (34.9%) (Table 1).

High-flow nasal cannula (HFNC) was used in 46 patients; in 10 patients, it was the only strategy for respiratory support; in the remaining, it was used as a bridge to NIV.

Use of PP

Awake PP was used in 114 patients (21.3%), 86 males (75.4%) (Table 1). Patients managed by PP more commonly had severe respiratory failure in comparison with patients managed without PP (PaO₂/FiO₂ <300 mmHg in 83.3% versus 61.3%, respectively) and less commonly known comorbidities (no known comorbidities in 37.7% versus 22.5%, respectively). A lower proportion of patients denied for ICU admission received PP in comparison with patients not denied (5.2% versus 25.5%). NIV was used in 81 and 106 patients managed or not managed by PP (43.3% versus 25.1%, respectively).

Among patients aged 85 years or older, 69 had ARDS at admission (78.4%), and 47 died at 30 days; PP was used in five of these patients (5.7%).

Among patients treated by PP, 64 (79.7%) required sedation to maintain PP.

Death or ICU admission during the hospital stay and within 30 days from admission

In-hospital death or ICU admission at 30 days from hospitalization occurred in 163 patients (30.4%). Death occurred in 129 patients (24.1%).

In-hospital death or ICU admission at 30 days occurred in 39 patients who underwent PP (34.2%) and in 124 of those who did not undergo PP (29.4%) (HR 1.16, 95% CI 0.87–1.56).

Multivariable analysis for death or ICU admission at 30 days is reported in Table 2. PP was not associated with the risk of ICU admission or death at 30 days (HR 1.17, 95% CI 0.78–1.74), while increasing age, severity of respiratory failure, and the use of NIV were independent predictors of death or ICU admission. Concerning the severity of respiratory failure at admission, an increased risk of death or ICU admission was observed for patients with PaO₂/FiO₂ between 200 and 100 mmHg and for PaO₂/FiO₂ <100 mmHg (Table 2 and Figure 2).

Table 2.

Independent predictors of primary and secondary study outcome events at 30 days.

	HR	95% CI		p value
	HR	Lower limit	Upper limit	p value
In-hospital death or ICU admission
PP	1.17	0.78	1.74	0.440
Age per year	1.03	1.02	1.05	<0.001
Gender, male	0.83	0.59	1.18	0.318
Comorbitidies ⩾1	1.22	0.78	1.90	0.373
PaO₂/FiO₂ (mmHg)
>300	Ref.	–	–	–
300–200	1.38	0.85	2.24	0.191
200–100	2.59	1.57	4.28	<0.001
<100	4.08	2.17	7.67	<0.001
NIV	1.45	1.01	2.10	0.043
qSOFA
0	Ref.	–	–
1	1.60	0.88	2.93	0.122
Unknown	0.90	0.60	1.32	0.587
Denied for ICU	0.94	0.59	1.18	0.812
In-hospital death
PP	1.01	0.61	1.67	0.952
Age per year	1.07	1.05	1.09	<0.001
Gender, male	0.75	0.51	1.11	0.152
Comorbitidies ⩾1	1.42	0.80	2.50	0.225
PaO₂/FiO₂ (mmHg)
>300	Ref.	–	–	–
300–200	1.33	0.78	2.35	0.306
200–100	2.43	1.38	4.28	0.002
<100	3.96	1.93	8.13	<0.001
NIV	1.50	0.98	2.28	0.057
qSOFA
0	Ref.	–	–
1	1.94	1.07	3.51	0.028
Unknown	0.77	0.48	1.23	0.275
Denied for ICU	0.58	0.34	0.98	0.043
ICU admission in patients not denied for ICU (n = 424)
PP	2.62	1.53	4.40	<0.001
Age per year	0.99	0.97	1.01	0.803
Gender, male	1.07	1.07	0.58	0.814
Comorbitidies ⩾1	1.23	0.70	2.18	0.464
PaO₂/FiO₂ (mmHg)
>300	Ref.	–	–	–
300–200	1.04	0.48	2.29	0.904
200–100	1.47	0.63	3.41	0.363
<100	1.89	0.69	5.18	0.213
NIV	3.23	1.70	6.14	<0.001
qSOFA
0	Ref.	–	–
1	1.17	0.15	8.75	0.878
Unknown	1.07	0.59	1.95	0.812

CI, confidence interval; HR, hazard ratio; ICU, intensive care unit; NIV, non-invasive ventilation; PaO₂/FiO₂, ratio of partial pressure of arterial oxygen (PaO₂) to fraction of inspired oxygen (FiO₂); qSOFA, quick sequential organ failure assessment.

Figure 2.

Risk of death or ICU admission (a) by severity of respiratory failure and (b) by clinical severity.

Death occurred in 23 patients treated with PP (20.2%) and in 102 of those not treated with PP (24.2%), (HR 0.83, 95% CI 0.56–1.25). PP was not associated with the risk of death at 30 days (HR 1.01, 95% CI 0.61–1.67) at multivariable analysis. Increasing age, severity of respiratory failure, the use of NIV, and qSOFA ⩾2 were independent predictors of death at 30 days. Concerning the severity of respiratory failure at admission, an increased risk of death was observed for patients with PaO₂/FiO₂ between 200 and 100 mmHg and for PaO₂/FiO₂ <100 mmHg (Table 2).

Among 424 patients not denied for ICU admission, 60 were admitted to ICU. PP, severity of respiratory failure, and the use of NIV were independent predictors of ICU admission in these patients (Table 2).

Sensitivity analyses

The results of sensitivity analyses are reported in Table 3.

Table 3.

Multivariable analyses for predictors of study outcome events at 30 days in patient sub-groups.

	HR	95% CI		p value
	HR	Lower limit	Upper limit	p value
In-hospital death or ICU admission in patients in ARDS at admission (n = 353)
PP	1.07	0.70	1.65	0.745
Age per year	1.03	1.02	1.05	<0.001
PaO₂/FiO₂ (mmHg)
300–200	Ref.	–	–	–
200–100	1.94	1.33	2.82	<0.001
<100	2.97	1.76	4.99	<0.001
NIV	1.22	0.83	1.78	0.313
qSOFA
0	Ref.
1	1.61	0.86	3.04	0.137
Unknown	0.80	0.52	1.23	0.316
In-hospital death in patients in ARDS at admission (n = 353)
Proning	1.01	0.60	1.73	0.960
Age per year	1.07	1.05	1.09	<0.001
PaO₂/FiO₂ (mmHg)
300–200	Ref.	–	–	0.005
200–100	1.85	1.21	2.83	<0.001
<100	2.76	1.52	5.02
NIV	1.39	0.90	2.15	0.134
qSOFA
0	Ref.
1	1.99	1.07	3.70	0.030
Unknown	0.77	0.46	1.27	0.304
In-hospital death or ICU admission in patients with qSOFA available (n = 418)
Proning	1.21	0.77	1.90	0.400
Age per year	1.04	1.02	1.05	<0.001
PaO₂/FiO₂ (mmHg)
>300	Ref.	–	–	–
300–200	1.58	0.90	2.77	0.112
200–100	3.32	1.88	5.89	<0.001
<100	5.09	2.54	10.22	<0.001
qSOFA
0	Ref.	–	–
1	1.63	0.89	2.99	0.111
NIV	1.31	0.87	1.97	0.185
In-hospital death in patients with qSOFA available (n = 418)
Proning	1.22	0.71	2.09	0.462
Age per year	1.07	1.05	1.09	<0.001
PaO₂/FiO₂ (mmHg)
>300	Ref.	–	–	–
300–200	1.35	0.73	2.49	0.334
200–100	2.82	1.53	5.19	<0.001
<100	3.65	1.69	7.92	<0.001
qSOFA
0	Ref.	–	–	–
1	2.12	1.17	3.86	0.013
NIV	1.33	0.85	2.10	0.216
In-hospital death or ICU admission in patients not denied for ICU (n = 424)
PP	1.19	0.76	1.86	0.436
Age per year	1.03	1.01	1.05	<0.001
Gender, male	0.91	0.59	1.41	0.693
Comorbitidies ⩾1	1.32	0.78	2.24	0.297
PaO₂/FiO₂ (mmHg)
>300	Ref	–	–	–
300–200	1.06	0.60	1.90	0.820
200–100	2.59	1.44	4.65	0.001
<100	3.18	1.47	6.86	0.003
NIV	1.89	1.21	2.95	0.005
qSOFA
0	Ref.	–	–	–
1	3.31	1.49	7.37	0.003
Unknown	0.89	0.56	1.42	0.643
In-hospital death in patients not denied for ICU (n = 424)
PP	0.94	0.50	1.75	0.853
Age per year	1.10	1.07	1.14	<0.001
Gender, male	0.92	0.76	0.54	1.55
Comorbitidies ⩾1	1.71	0.77	3.81	0.184
PaO₂/FiO₂ (mmHg)
>300	Ref.	–	–	–
300–200	0.86	0.43	1.73	0.686
200–100	2.24	1.10	4.55	0.026
<100	4.15	1.63	10.55	0.003
NIV	1.71	1.00	2.90	0.046
qSOFA
0	Ref.	–	–
1	4.2	1.85	9.52	0.001
Unknown	0.64	0.34	1.20	0.172
In-hospital death in denied for ICU (n = 112)
PP	0.86	0.33	2.37	0.762
Gender, male	0.65	0.37	1.14	0.137
Comorbitidies ⩾1	1.11	0.46	2.67	0.814
PaO₂/FiO₂ (mmHg)
>300	Ref.	–	–	–
300–200	2.40	0.90	6.42	0.079
200–100	4.15	1.56	1.01	0.004
<100	6.25	1.96	11.85	0.002
NIV	1.12	0.53	1.95	0.864
qSOFA
0	Ref.	–	–
1	1.44	0.61	3.57	0.388
Unknown	0.76	0.36	1.61	0.481

ARDS, severe acute respiratory distress syndrome; CI, confidence interval; HR, hazard ratio; ICU, intensive care unit; NIV, non-invasive ventilation; PaO₂/FiO₂, ratio of partial pressure of arterial oxygen (PaO₂) to fraction of inspired oxygen (FiO₂); qSOFA, quick sequential organ failure assessment.

In 353 patients entering the study with ARDS, no association was observed between PP and the risk of death or ICU admission (HR 1.07, 95% CI 0.70–1.65). Mild ARDS (PaO₂/FiO₂ between 200 and 300 mmHg) was significantly associated with reduced risk of death or ICU admission (HR 0.33, 95% CI 0.20–0.56) (Figure 2).

When the analysis was restricted to 138 patients with moderate-to-severe ARDS (PaO₂/FiO₂ ⩽ 200 mmHg), no association was found with PP (HR 0.99, 95% CI 0.59–1.68) or NIV (HR 0.99, 95% CI 0.60–1.62) and the risk of death or ICU admission; the only independent risk factor for death or ICU admission was age (HR 1.03 per year, 95% CI 1.01–1.04).

In 187 patients receiving treatment with NIV, no association was found between PP (HR 1.43, 95% CI 0.92–2.22) and the risk of death or ICU admission.

When 418 patients with data on clinical severity available for the calculation of the qSOFA were considered, PP was not associated with the risk of death or ICU admission (HR 1.21, 95% CI 0.77–1.90); increasing age and severity of respiratory failure were confirmed to be independent predictors of death or ICU admission. Concerning clinical severity, a marginally significant increase in the risk of death or ICU admission was observed for patient having qSOFA score ⩾2 at admission (Figure 2).

Propensity score-matching analyses

Among patient characteristics reported in Table 1, age, sex, PaO₂/FiO₂ levels, number of comorbidities, and treatment with NIV differed between patients managed with or without PP. Propensity score matching was used to balance patients’ characteristics between the cohort including the 114 patients managed by PP and a control cohort including 114 matching patients managed without PP. The final propensity score-matching sample comprised 228 subjects.

The balance of measured covariates showed negligible difference between matched cohorts (standardized mean differences were all ⩽0.1).

In the propensity score-matching population, PP was not associated with death or ICU admission at 30 days (HR 1.33, 95% CI 0.83–2.14; p = 0.193) and not with death at 30 days (HR 1.26, 95% CI 0.70–2.20; p = 0.403). An association was confirmed between PP and ICU admission (HR 2.29, 95% CI 1.29–3.86; p = 0.004) (Table 4).

Table 4.

Characteristics of the patients after propensity score matching.

Variables	PP (n = 114)	No PP (n = 114)		Standardized difference
Age	63.29	63.22		0.054
Male, n (%)	24	27		0.061
P/F >300, n (%)	66	17		0.023
P/F 300–200, n (%)	48	47		0.017
P/F 200–100, n (%)	27	27		0.000
P/F <100, n (%)	7	7		0.000
Comorbitidies ⩾1, n (%)	62	64		0.036
NIV, n (%)	71	72		0.038
Primary and secondary outcomes after propensity score matching
	HR	95% CI		p value
	HR	Lower limit	Upper limit	p value
In-hospital death/ICU	1.33	0.83	2.14	0.193
In-hospital death	1.26	0.70	2.20	0.403
ICU admission	2.29	1.29	3.86	0.003

CI, confidence interval; HR, hazard ratio; ICU, intensive care unit; NIV, non-invasive ventilation; PP, prone positioning.

Discussion

Our study shows that in patients with SARS-CoV-2-related respiratory failure managed outside the ICU setting, awake PP was not associated with the risk of death or ICU admission at univariate analysis and neither after adjusting for severity of respiratory failure and clinical severity. These results were confirmed when the analysis was limited to patients with COVID-19-related ARDS, in patients not denied for ICU and after propensity score matching. Awake PP was an independent predictor of ICU admission. Awake PP is used mainly in the presence of severe respiratory failure and reduced bulk of comorbidities.

PP is an established evidence-based practice in patients with typical ARDS undergoing invasive mechanical ventilation¹⁸ as it has been shown to effectively reduce lung shunting, thanks to improvements in ventilatory homogeneity and relatively constant perfusion patterns.^19,20 PP does not appear to alter regional distribution of pulmonary blood flow, with perfusion predominating toward dorsal lung regions due to non-gravitational factors. However, a recent study in 20 critically ill COVID-19 patients assisted by NIV in ICU showed that PP improved oxygenation with the counterbalance of a greater thickening diaphragmatic fraction compared to supine position. These results suggest an increase in inspiratory effort during PP.²¹

In our study outside the ICU, awake PP was not associated with reduction in mortality or ICU admission in the overall population, not after balancing patients’ characteristics between patients managed with and without PP by propensity score matching. Recent randomized studies confirmed a clinical benefit of PP in intubated COVID-19 patients²² and in awake patients at early stages of respiratory failure requiring HFNO.²³ However, patients included in these studies were more than 10 years younger than patients included in our study.

It is well known that in patients with severe respiratory failure treated with NIV, PP may delay appropriate intubation. It is conceivable that awake PP should be proposed to selected patients, taking into account the severity of respiratory failure and patient tolerance with or without mild sedation, in order not to delay appropriate intubation and not to increase respiratory effort. In previous studies, PP reduced mortality in severe ARDS in the setting of mechanical ventilation, mainly with PP applied for at least 12 h daily.²⁴ Maintaining PP for 12 h in awake patients is problematic and may increase anxiety and work of breathing. In fact, in our study, a consistent proportion of patients required sedation to maintain PP. Whether sedation is effective in reducing the work of breathing and self-induced lung injury remains to be defined.²⁵ In addition, either improved oxygenation or sedation may contribute to delay protective intubation. Finally, it is conceivable that PP, if well tolerated, can be of help also in patients denied for ICU to overcome the critical phase of COVID-19 pneumonia and to relief dyspnea if well tolerated.²⁶ As for today, whether awake PP may reduce death or progression to ICU admission in the whole COVID-19 patients with respiratory failure remains controversial.

In our study outside the ICU, awake PP was used in about 20% of patients. The use of PP was more common in patients with severe respiratory failure. In this setting, PP can be not well tolerated due to the discomfort related to increased abdominal pressure, to devices for respiratory support and panic. Moreover, for older patients with spondylopathies, PP may produce pain. These issues are particularly relevant in awake patients. It is conceivable that in some patients, the benefit of improved oxygenation is overcome by the increase in respiratory effort. In a pragmatic randomized study, a multifaceted intervention to increase PP did not improve outcomes in non-critically ill patients with hypoxemia and COVID-19.²⁷

In our study age, the severity of respiratory failure and the use of NIV were associated with increased risk of death or ICU admission. While the role of age and severity of respiratory failure as predictors are intuitive and plausible, the role of NIV in the management of patients with COVID-19 pneumonia is controversial.²⁸ The use of positive end-expiratory pressure via NIV in the management of ARDS is beneficial by preventing alveolar derecruitment but may result in overdistension of previously well-ventilated alveoli.²⁹ In-hospital mortality in patients treated with NIV outside the ICU was as high as 36% in a meta-analysis of studies in COVID-19 patients.²⁸ Mortality may be related to the occurrence of barotrauma and ventilator-induced lung injury. With this purpose, the use of protective ventilation may result in clinical benefit by reducing ventilator-induced lung injury. However, NIV may delay appropriate intubation.

We considered ICU admission as a component of the primary outcome. This is common in studies in non-intubated patients with COVID-19.³⁰ However, ICU admission is an operator- and resource-dependent outcome. PP was associated with ICU admission in the overall study population and after propensity score-matching analysis. Whether these results reveal the selection of patients eligible to ICU admission for PP management remains undefined.

Our study has several limitations. PP was not uniformly applied, mainly concerning the duration, and this may limit the generalizability of our results. Data on daily duration of PP per patient, number of cycles, and sedation are not available. Second, the strategy for respiratory support was left at the discretion of the treating physicians. Third, we used qSOFA instead of a specific risk-scoring system for COVID-19. However, qSOFA is made of simple, rapid, and practical items that makes it widely used for risk stratification in critically ill patients.¹² Finally, device-related sores, prone-related pressure sores, ulcers, and medical device displacement commonly occurred, mainly in patients treated with PP, but these events were not systematically reported.

Our study has also some strengths, mainly related to the large sample size and the possibility to perform a propensity score-matching analysis. Moreover, our study included patients during the second COVID-19 wave, and both physicians and nurses had already been trained on proning and on sedation during the first COVID wave.

Conclusion

In conclusion, awake PP seems not to be associated with clinical benefit when used outside the ICU in patients with severe respiratory failure related to COVID-19 pneumonia. Randomized studies are required to definitively assess the clinical benefit of PP in awake and non-intubated patients with SARS-CoV-2-related respiratory failure.

Footnotes

Acknowledgements

None.

Declarations

ORCID iD

Mara Graziani

References

Burki

TK.

Coronavirus in China. Lancet Respir Med 2020; 8: 238.

Yuki

Fujiogi

Koutsogiannaki

COVID-19 pathophysiology: a review. Clin Immunol 2020; 215: 108427.

Zhu

Zhang

Wang

, et al. A novel coronavirus from patients with pneumonia in China, 2019. N Engl J Med 2020; 382: 727–733.

Ranieri

Rubenfeld

Thompson

, et al. Acute respiratory distress syndrome: the Berlin definition. JAMA 2012; 307: 2526–2533.

Chen

Cai

, et al. Risk factors associated with acute respiratory distress syndrome and death in patients with coronavirus disease 2019 pneumonia in Wuhan, China. JAMA Intern Med 2020; 180: 934–943.

McGoogan

JM.

Characteristics of and important lessons from the coronavirus disease 2019 (COVID-19) outbreak in China: summary of a report of 72 314 cases from the Chinese center for disease control and prevention. JAMA 2020; 323: 1239–1242.

Baden

El Sahly

Essink

, et al. COVE study group efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine. N Engl J Med 2021; 384: 403–416.

Gattinoni

Taccone

Carlesso

, et al. Prone position in acute respiratory distress syndrome. Rationale, indications, and limits. Am J Respir Crit Care Med 2013; 188: 1286–1293.

Cohen

Wasserstrum

Segev

, et al. Beneficial effect of awake prone position in hypoxaemic patients with COVID-19: case reports and literature review. Intern Med J 2020; 50: 997–1000.

10.

Koeckerling

Barker

Mudalige

, et al. Awake prone positioning in COVID-19. Thorax 2020; 75: 833–834.

11.

Gleissman

Forsgren

Andersson

, et al. Prone positioning in mechanically ventilated patients with severe acute respiratory distress syndrome and coronavirus disease 2019. Acta Anaesthesiol Scand 2021; 65: 360–363.

12.

Liu

Yao

Qiu

, et al. Predictive performance of SOFA and qSOFA for in-hospital mortality in severe novel coronavirus disease. Am J Emerg Med 2020; 38: 2074–2080.

13.

Austin

PC.

A critical appraisal of propensity-score matching in the medical literature between 1996 and 2003. Stat Med 2008; 27: 2037–2049.

14.

Austin

PC.

Statistical criteria for selecting the optimal number of untreated subjects matched to each treated subject when using many-to-one matching on the propensity score. Am J Epidemiol 2010; 172: 1092–1097.

15.

Hansen

Klopfer

SO.

Optimal full matching and related designs via network flows. J Comput Graph Stat 2006; 15: 609–627.

16.

Zhao

Luo

, et al. Propensity score matching with R: conventional methods and new features. Ann Transl Med 2021; 9: 812.

17.

Gayat

Resche-Rigon

Mary

, et al. Propensity score applied to survival data analysis through proportional hazards models: a Monte Carlo study. Pharm Stat 2012; 11: 222–229.

18.

Guérin

Reignier

Richard

, et al. PROSEVA study group. Prone positioning in severe acute respiratory distress syndrome. N Engl J Med 2013; 368: 2159–2168.

19.

Scholten

Beitler

Prisk

, et al. Treatment of ARDS with prone positioning. Chest 2017; 151: 215–224.

20.

Pappert

Rossaint

Slama

, et al. Influence of positioning on ventilation-perfusion relationships in severe adult respiratory distress syndrome. Chest 1994; 106: 1511–1516.

21.

Cammarota

Rossi

Vitali

, et al. Effect of awake prone position on diaphragmatic thickening fraction in patients assisted by noninvasive ventilation for hypoxemic acute respiratory failure related to novel coronavirus disease. Crit Care 2021; 25: 305.

22.

Rosén

von Oelreich

Fors

, et al. Awake prone positioning in patients with hypoxemic respiratory failure due to COVID-19: the PROFLO multicenter randomized clinical trial. Crit Care 2021; 25: 209.

23.

Ehrmann

Ibarra-Estrada

, et al. Awake prone positioning for COVID-19 acute hypoxaemic respiratory failure: a randomised, controlled, multinational, open-label meta-trial. Lancet Respir Med 2021: 1387–1395.

24.

Munshi

Del Sorbo

Adhikari

NKJ

, et al. Prone position for acute respiratory distress syndrome. A systematic review and meta-analysis. Ann Am Thorac Soc 2017; 14: S280–S288.

25.

MacIntyre

NR.

Physiologic effects of non invasive ventilation. Respir Care 2019; 64: 617–628.

26.

Alhatem

Spruijt

Heller

, et al. ‘Do-not-resuscitate (DNR)’ status determines mortality in patients with COVID-19. Clin Dermatol 2021; 39: 510–516.

27.

Fralick

Colacci

Munshi

, et al. Prone positioning of patients with moderate hypoxaemia due to covid-19: multicentre pragmatic randomised trial (COVID-PRONE). BMJ 2022; 376: e068585.

28.

Cammarota

Esposito

Azzolina

, et al. Noninvasive respiratory support outside the intensive care unit for acute respiratory failure related to coronavirus-19 disease: a systematic review and meta-analysis. Crit Care 2021; 25: 268.

29.

Shelhamer

Wesson

Solari

, et al. Prone positioningin moderate to severe acute respiratory distress syndrome due to COVID-19: a cohort study and analysis of physiology. J Intensive Care Med 2021; 36: 241–252.

30.

Maraziti

Becattini

Early variation of respiratory indexes to predict death or ICU admission in severe acute respiratory syndrome coronavirus-2-related respiratory failure. Respiration 2022; 101: 632–637.

The role of prone positioning in patients with SARS-CoV-2-related respiratory failure in non–intensive care unit

Abstract

Background:

Objectives:

Design:

Methods:

Results:

Conclusion:

Keywords

Introduction

Methods

Study design and population

Study outcomes

Study procedures

Data collection

Statistical analysis

Results

Use of PP

Death or ICU admission during the hospital stay and within 30 days from admission

Sensitivity analyses

Propensity score-matching analyses

Discussion

Conclusion

Footnotes

Acknowledgements

Declarations

ORCID iD

References