Mechanical power normalized to aerated lung predicts noninvasive ventilation failure and death and contributes to the benefits of proning in COVID-19 hypoxemic respiratory failure

Giovanni Musso; Claudio Taliano; Marco De Iuliis; Elena Paschetta; Caterina Fonti; Andrea Ferraris; Marta Druetta; Ines Sarah Vianou; Francesca Ranghino; Federica Riedo; Davide Deangelis; Gloria Tirabassi

doi:10.1007/s13167-023-00325-5

. 2023 Jun 10;14(3):341–379. doi: 10.1007/s13167-023-00325-5

Mechanical power normalized to aerated lung predicts noninvasive ventilation failure and death and contributes to the benefits of proning in COVID-19 hypoxemic respiratory failure

Giovanni Musso ^1,^✉, Claudio Taliano ², Marco De Iuliis ³, Elena Paschetta ², Caterina Fonti ¹, Andrea Ferraris ³, Marta Druetta ², Ines Sarah Vianou ², Francesca Ranghino ², Federica Riedo ², Davide Deangelis ², Gloria Tirabassi ⁴

PMCID: PMC10256581 PMID: 37359998

Abstract

Background

Concern exists that noninvasive ventilation (NIV) may promote ventilation-induced lung injury(VILI) and worsen outcome in acute hypoxemic respiratory failure (AHRF). Different individual ventilatory variables have been proposed to predict clinical outcomes, with inconsistent results.

Mechanical power (MP), a measure of the energy transfer rate from the ventilator to the respiratory system during mechanical ventilation, might provide solutions for this issue in the framework of predictive, preventive and personalized medicine (PPPM). We explored (1) the impact of ventilator-delivered MP normalized to well-aerated lung (MP_WAL) on physio-anatomical and clinical responses to NIV in COVID-19-related AHRF and (2) the effect of prone position(PP) on MP_WAL.

Methods

We analyzed 216 noninvasively ventilated COVID-19 patients (108 patients receiving PP + NIV and 108 propensity score-matched patients receiving supine NIV) with moderate-to-severe(paO2/FiO2 ratio < 200) AHRF enrolled in the PRO-NIV controlled non-randomized study (ISRCTN23016116).

Quantification of differentially aerated lung volumes by lung ultrasonography (LUS) was validated against CT scans. Respiratory parameters were hourly recorded, ABG were performed 1 h after each postural change. Time-weighed average values of ventilatory variables, including MP_WAL, and gas exchange parameters (paO2/FiO2 ratio, dead space indices) were calculated for each ventilatory session. LUS and circulating biomarkers were assessed daily.

Results

Compared with supine position, PP was associated with a 34% MP_WAL reduction, attributable largely to an absolute MP reduction and secondly to an enhanced lung reaeration.

Patients receiving a high MP_WAL during the 1^st 24 h of NIV [MP_WAL(day 1)] had higher 28-d NIV failure (HR = 4.33,95%CI:3.09 − 5.98) and death (HR = 5.17,95%CI: 3.01 − 7.35) risks than those receiving a low MP_WAL(day 1).

In Cox multivariate analyses, MP_WAL(day 1) remained independently associated with 28-d NIV failure (HR = 1.68,95%CI:1.15–2.41) and death (HR = 1.69,95%CI:1.22–2.32).

MP_WAL(day 1) outperformed other power measures and ventilatory variables as predictor of 28-d NIV failure (AUROC = 0.89;95%CI:0.85–0.93) and death (AUROC = 0.89;95%CI:0.85–0.94).

MP_WAL(day 1) predicted also gas exchange, ultrasonographic and inflammatory biomarker responses, as markers of VILI, on linear multivariate analysis.

Conclusions

In the framework of PPPM, early bedside MP_WAL calculation may provide added value to predict response to NIV and guide subsequent therapeutic choices i.e. prone position adoption during NIV or upgrading to invasive ventilation, to reduce hazardous MP_WAL delivery, prevent VILI progression and improve clinical outcomes in COVID-19-related AHRF.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13167-023-00325-5.

Keywords: Predictive preventive personalized medicine (PPPM / 3PM), COVID-19, Hypoxemic respiratory failure, Individual outcomes, Differentially aerated lung volume, Predictive model, Noninvasive ventilation, Therapeutic choice, Lung ultrasound, Mechanical power, Well-aerated lung, Prone position

Introduction

The unmet need for accurate early predictors of treatment failure and death during noninvasive ventilation (NIV)

NIV is being increasingly used as first-line ventilatory support in de novo (i.e., not due to exacerbation of chronic lung disease or cardiac failure) acute hypoxemic respiratory failure (AHRF). However, concern has been raised that mechanical ventilation may exacerbate lung injury (i.e., ventilation-induced lung injury, VILI) and worsen outcome in spontaneously breathing patients with AHRF [1]. Different ventilatory variables have been proposed to contribute to VILI progression, including high tidal volume [2], minute ventilation [3] and inspiratory effort [4]: these parameters have been separately addressed in clinical studies, but data in noninvasively ventilated patients are scarce, inconsistent and inconclusive, at least in part due to the difficulty in controlling and monitoring time-varying individual ventilatory parameters in spontaneously breathing patients.

Potential contribution of Mechanical Power (MP) to clinical prediction in the framework of PPPM

Predictive, Preventive, and Personalized Medicine (PPPM) is an effective integrative approach, which has been promoted by the European Association for Predictive, Preventive and.

Personalized Medicine (EPMA, http://www.epmanet.eu/) [5]. It contains three aspects: individual predisposition prediction, targeted preventive measures and personalized treatment algorithms [6].

In recent years, a unifying patho-physiological theory, based on thermodynamic principles, has been proposed to underlie VILI: this theory attributes lung injury to the energy transfer from the ventilator to the pulmonary parenchyma, with energy dissipation within the lungs leading to heat production, cell integrity and extracellular matrix disruption, and inflammatory cell recruitment [7-9]. Consistently, mechanical power (MP), a measure of the energy transfer rate from the ventilator to the respiratory system, predicted mortality in invasively ventilated ARDS patients, irrespective of the combination of each ventilatory component [10-13].

Whether the energy delivered to the respiratory system during noninvasive ventilation (NIV) affects clinical outcomes in AHRF is unexplored. During COVID-19 pandemic, NIV intensity and duration were linked to an increased mortality in COVD-19-related AHRF [14-16], but no individual ventilatory parameter, including respiratory drive and inspiratory effort, was able to predict clinical outcomes [17-20]. Furthermore, clinical benefits from awake prone position (PP) during NIV have been reported by several, but not all studies, through still unclear mechanisms [21, 22].

Working hypothesis

We hypothesized the energy delivered by noninvasive ventilatory assistance to the respiratory system could contribute to physio-anatomical and clinical responses to NIV in Severe Acute Respiratory Syndrome Corona Virus 2 (SARS‐CoV‐2) pneumonia-related AHRF, and that the clinical benefits of awake PP in these patients could be at least in part mediated by a reduced MP delivery during prone NIV. We therefore investigated:

the effect of prone position (PP) on different measures of MP during NIV.
the contribution of MP measures delivered early during either supine or prone NIV to physio-anatomical and clinical outcomes in COVID-19 pneumonia

Methods

In this secondary analysis of the non-randomized, controlled Prone position in NonInvasive Ventilation (PRO-NIV) study(study ID: ISRCTN23016116) [23], we studied 216 SARS‐CoV‐2 pneumonia patients with acute (i.e. symptom onset < 14 days of hospital admission) moderate-to-severe AHRF(paO2/FiO2 ratio < 200 mmHg while on a FiO2 50% Venturi mask or a non-rebreather reservoir bag mask): 108 patients treated with NIV (CPAP or PSV) plus early PP and 108 matched controls treated with supine NIV) at HUMANITAS Gradenigo COVID Subintensive Care Unit between June 1st 2020 and June 30th 2021.

In both groups, NIV was initiated within 24 h of admission to Subintensive Care Unit and delivered continuously for ≥ 48 h or until discharge or death; full-face mask was the initial interface of choice. PP was initiated within 24 h of admission in the PP group. NIV and PP duration, equipment, settings, standard care, monitoring, treatment failure criteria were protocolized a priori before patients enrollment (see supplementary text, Supplementary Fig. 1, Supplementary Table 1) [23].

Measurements

Respiratory parameters

NIV duration, posture, ventilatory settings and parameters (spO2, RR, VTe, MV) were continuously monitored and recorded on an hourly basis on a predefined form.

ABGs were performed during NIV, every 24 h and ≥ 1 h after each postural change, after achievement of ventilatory stability (defined by a ≤ 10% variation in RR and VTe and air leaks < 10% for ≥ 30 min).

The following gas exchange parameters were calculated from ABG: paO2/FiO2 ratio and dead space indices (DSIs) [Ventilatory Ratio (VR) and corrected Minute Ventilation (MV_corr)]

The energy delivered by ventilatory assistance to respiratory system per time unit (mechanical power, MP) was estimated via Becher formula [24], which was already evaluated in noninvasively ventilated patients and was recommended to improve noninvasive ventilatory support monitoring in COVID-19 pneumonia [14, 25]. In this formula, ∆Pinsp was replaced by ∆Paw(airway pressure over PEEP) to reflect the energy imparted by ventilator during inspiration [1, 26].

Since NIV isa dynamic process, to reflect the power delivered during the entire ventilatory sessions, time weighted–average of hourly MP values during each NIV session (supine/prone) was calculated as the area under the parameter–versus–time plot (detailed in supplementary text) [27]. The same procedure was followed to calculate time weighted–average values of RR, VTe, and MV for each ventilator session and for each day.

For MV, we also planned to verify consistency between time-averaged values and those obtained at the time of ABGs, which were used to calculate DSIs.

The primary exposure variable of interest was MP normalized by the volume of well-aerated lung (MP_WAL), to reflect the “intensity” of the power, i.e., the volume of well-aerated lung exposed to energy load during mechanical ventilation, during the 1^st 24 h of NIV [MP_WAL(day 1)].

We also normalized MP by predicted body weight (MP_PBW) to account for individual lung size variation.

Lung imaging

All patients underwent a lung CT scan on admission: the nature and extent of parenchymal involvement were scored using a validated index [28] and the volume of well-aerated lung (WAL), of poorly aerated lung (PAL) and of non-aerated lung (NAL) were quantified on CT scans via a validated open-source software (3D Slicer ver.4.13.2) (see supplementary text) [29].

Lung ultrasound was performed daily from admission (day 0) to day 7 by three intensivists with expertise in lung and cardiac recording and interpretation (each operator having performed at least 50 supervised procedures and at least 200 non-supervised procedures) [30] using the same equipment (HM70A Samsung, Seoul, Korea), the same convex-array probe and the same setting.

The accuracy of ultrasound examinations in staging lung disease severity was preliminarily evaluated at baseline against the CT scan (double-blinded operators, LUS performed within 24 h of CT examination).

The severity and extent of parenchymal involvement of each of 6 lung regions (2 anterior, 2 lateral, 2 dorsal) were scored (range 0–3) [31] and recorded on a predefined form and the following indices were calculated (supplementary text):

regional and global lung ultrasound score (LUS);
regional and global LUS reaeration score, a validated index of lung recruitment (i.e., change from consolidated, non-aerated tissue to aerated tissue) [32, 33];
additionally, using software-based lung parenchyma segmentation and analysis function [29], each lung was divided into six areas to mirror as much as possible the regions explored by ultrasound, and regional and global LUS-derived WAL volume (i.e. lung volume with LUS score 0–1), PAL (i.e. lung volume with LUS score 2) and NAL (i.e. lung volume with LUS score 3) were derived from admitting LUS scan examinations as previously described in ARDS [34]. Global LUS-derived WAL volume, a predictor of COVID-19 pneumonia outcome [24], was then calculated from daily LUS scans through day 0–7(detailed in supplementary text).

The agreement between regional LUS score and regional CT classification was assessed with Cohen’s kappa coefficient, and the association between regional CT-derived gas/tissue content and regional LUS categories was assessed by simple linear regression and Spearman’s rank correlation (r_s).

Details on full derivations are provided in supplementary text.

The PEEP at which each LUS examination was made was recorded.

Circulating biomarkers: eighteen circulating inflammatory and procoagulative biomarkers were measured daily from admission to discharge or death [35].

Definitions and timepoints

We adopted the following definitions:

day 0 (baseline): the time of NIV initiation;
day 1: first 24 h after NIV initiation;
day 7: day 7 after NIV initiation;
timepoint sp0: supine position. In PP group, the session preceded the first PP session. ABG was performed ≥ 1 h after supine NIV initiation, after achieving ventilatory stability;
timepoint pp1: first PP session in PP group. ABG was performed ≥ 1 h after prone NIV initiation, after achieving ventilatory stability;
timepoint sp1: supine position; ABG was performed ≥ 1 h after resupination following the first PP session in the PP group and 24 h after NIV initiation in the supine group, after achieving ventilatory stability;

This schedule allowed comparing all groups in the supine position, after 24 h of NIV, while taking into account the effect of the first PP session (pp1) in the PP group.

Further definitions are provided in legends to Table 1 and in a study scheme (Supplementary Fig. 1).

Table 1.

Baseline, treatment-related parameters and clinical outcomes of supine and PP group; within each treatment group, patients were divided according to median values of Mechanical Power_WAL during the 1^st 24 h of NIV (MP_WAL at day 1) into High MP_wal (MP_wal ≥ 9.1 J/min/L) or Low MP_wal (MP_wal < 9.1 J/min/L)(n = 216)

Panel A: baseline parameters
Parameter	Supine				PP
	All (n = 108)	High MP_wal (n = 66)	Low MP_wal (n = 42)	P high vs. low	All (n = 108)	High MP_wal (n = 41)	Low MP_wal (n = 67)	P high vs. low
Age(yr)	67(62, 75)	68(62, 75)	66(57, 74)	0.592	66(56, 73)	69(65, 75)	65(56, 73)	0.839
Male sex(n, %)	73(68%)	41(62%)	22(52%)	0.812	75(69%)	29(70%)	46(61%)	0.631
White, non-hispanic(n, %)	98(91%)	60(90%)	38(90%)	0.892	97(90%)	38(93%)	59(88%)	0.713
Time from symptom onset to hospital admission(d)*	7(5, 9)	6(5, 8)	7(5, 9)	0.899	7(5, 9)	7(5, 9)	7(5, 9)	0.638
Time from hospital admission to NIV (hrs)	20(7, 56)	20(7, 56)	20(5, 62)	0.802	20(7, 64)	20(6, 61)	20(7, 64)	0.739
BMI(kg/m²)	27.8 (25.9, 31.2)	27.7 (25.4, 30.2)	28.1 (26.2, 30.7)	0.782	28.0 (26.3, 30.5)	27.9 (26.4, 30.1)	28.2 (26.3, 30.5)	0.631
Obesity(BMI ≥ 30 kg/m²) n(%)	36(33%)	22(33%)	14(33%)	0.639	38(35%)	13(32%)	25(37%)	0.412
Type 2 diabetes mellitus n(%)	26(24%)	15(23%)	11(26%)	0.713	28(26%)	13(31%)	15(22%)	0.492
Hypertension n(%)	60(56%)	37(56%)	23(52%)	0.639	56(52%)	22(54%)	34(51%)	0.579
Chronic lung disease
COPD	12(11%)	7(10%)	5(12%)	0.824	10(10%)	4(10%)	6(9%)	0.902
Asthma	3(3%)	1(1%)	2(5%)	0.538	3(3%)	1(2%)	2(3%)	0.727
Coronary artery disease n(%)	3(5%)	2(5%)	1(5%)	0.902	4(6%)	2(7%)	2(6%)	0.859
Chronic atrial fibrillation n(%)	5(4%)	3(4%)	2(4%)	0.572	6(6%)	3(4%)	3(4%)	0.711
Chronic kidney disease n(%)	9(8%)	5(8%)	4(10%)	0.713	10(9%)	4(10%)	6(9%)	0.631
Immunocompromised state	1(1%)	1(1%)	0(0%)	0.782	1(0%)	0(0%)	1(0%)	0.972
Smoking status:
former(%) current(%)	10(10%) 10(10%)	7(10%) 7(10%)	3(8%) 3(8%)	0.721 0.386	9(8%) 6(6%)	4(10%) 2(5%)	5(8%) 4(6%)	0.598 0.383
Do-not-Intubate (DNI) n(%)	10(9%)	6(9%)	4(10%)	0.739	8(8%)	3(7%)	5(8%)	0.811
ISARIC 4 C mortality score	14(11, 17)	14(11, 17)	13(10, 16)	0.593	14(10, 18)	14(12, 18)	13(10, 15)	9.739
SAPS II score	36(32, 39)	36(31, 39)	35(30, 40)	0.538	35(30, 41)	36(32, 41)	34(30, 39)	0.513
Lung CT scan volumes
WAL)(L) WAL(%) -Normally aerated (%) -Hyperynflated(%) PAL(L) PAL(%) NAL(L) NAL(%)	2.0(1.5, 2.5) 49(43, 62) 46(44,55) 3(1, 4) 1.9(1.6, 2.3) 44(39, 50) 0.4(0.3, 0.5) 8(6, 10)	2.0(1.6, 2.4) 49(44, 61) 50(45,55) 3(1, 4) 1.9(1.6, 2.3) 44(39, 50) 0.4(0.3, 0.5) 8(6, 10)	1.9(1.6, 2.5) 50(41, 61) 51(44, 56) 3(2, 5) 1.2(1.0, 1.4) 36(30, 41) 0.5(0.4, 0.8) 9(7, 12)	0.497 0.672 0.589 0.713 0.639 0.485 0.738 0.591	1.9(1.5, 2.4) 49(45, 58) 49(44, 55) 2(1, 4) 1.2(1.0, 1.4) 36(30, 41) 0.5(0.4, 0.7) 9(8, 13)	1.9(1.6, 2.3) 50(39, 60) 49(42, 55) 4(2, 4) 1.7(1.1, 2.1) 44(40, 48) 0.4(0.4, 0.6) 9(8, 12)	1.9(1.5, 2.4) 49(45, 58) 50(43, 54) 2(1, 4) 1.2(1.0, 1.4) 36(30, 41) 0.5(0.4, 0.7) 9(8, 13)	0.725 0.539 0.582 0.798 0.607 0.361 0.811 0.496
Lung global CT severity score	46(31, 63)	45(31, 63)	47(31, 61)	0.621	47(31, 62)	46(32, 65)	47(31, 62)	0.639
Lung global LUS severity score	25(22, 28)	25(22, 28)	24(23, 28)	0.838	25(22, 29)	25(23, 28)	25(22, 28)	0.729
C-reactive protein (mg/dL)	13.9(10.5, 18.5)	13.7(10.2,18.9)	14.3(10.9,19.0)	0.817	13.5(10.5,19.2)	13.2(10.6,18.9)	13.9(10.5,19.2)	0.813
Temperature(°C)	36.6 (36.1, 36.9)	36.6 (36.1, 36.9)	36.5 (36.0, 36.9)	0.713	36.6 (36.2, 37.0)	36.5 (36.1, 36.9)	36.6 (36.2, 37.0)	0.712
Sys BP (mmHg)	130 (120, 141)	130 (120, 141)	130 (118, 140)	0.630	133 (119, 142)	130 (116, 140)	133 (119, 142)	0.767
Dia BP (mmHg)	75(65, 80)	75(65, 80)	75(68, 89)	0.784	75(68, 89)	75(68, 89)	75(68, 89)	0.784
Heart Rate (beats/min)	85(71, 96)	85(71, 96)	81(70, 99)	0.837	83(75, 98)	81(70, 99)	83(75, 98)	0.698
NRS (Dyspnoea)#	3(2,5)	3(2,5)	4(2,6)	0.528	4(2,6)	3(2,6)	4(2,6)	0.612
WOB scale#	3(2, 3)	3(2, 3)	3(2, 3)	0.582	3(2, 3)	3(2, 3)	3(2, 3)	0.812
RR(breaths/min)#	26(24, 30)	27(24, 30)	25(22, 30)	0.219	26(24, 30)	27(24, 30)	25(22, 30)	0.227
PaO2/FiO2 ratio#	104(78, 137)	102(78, 137)	105(76, 140)	0.429	103(79, 136)	102(80, 139)	104(79, 136)	0.728
PaO2/FiO2 ratio category n(%)#
150–199 100–149 < 100	17(16%) 35 (32%) 56(52%)	10(15%) 22 (33%) 35(53%)	7(17%) 13(31%) 21(50%)	0.439 0.580 0.479	16(15%) 35 (32%) 57(53%)	6(15%) 15 (37%) 21(51%)	10(15%) 20 (30%) 36(54%)	0.913 0.294 0.683
paCO2 (mmHg)#	34(31, 37)	34(31, 37)	35(32, 39)	0.713	36(31, 40)	35(31, 38)	36(31, 40)	0.602
paCO2 > 45 mmHg# n(%)	1(1%)	1(1%)	0(0%)	0.912	0(0%)	0(0%)	0(0%)	0.983
Arterial pH#	7.47 (7.45, 7.49)	7.47 (7.45, 7.49)	7.46 (7.44, 7.49)	0.734	7.45 (7.42, 7.49)	7.47 (7.44, 7.49)	7.45 (7.42, 7.49)	0.691
Adiuvant therapies
Dexamethasone 6 mg/d	108 (100%)	66 (100%)	42(100%)	0.999	108(100%)	41(100%)	67(100%)	0.999
Antivirals
Remdesivir Lopinavir/ritonavir	1(1%) 0	1(1%) 0	1 0	0.999 0.999	1 0	0 0	(1%) 0	0.999 0.999
Hydroxychloroquine	0	0	0	0.999	0	0	0	0.999
Tocilizumab	3(3%)	2(6%)	1(2%)	0.712	0	0	0	0.999
Convalescent plasma	1(1%)	1(1%)	0	0.999	0	0	0	0.999
Enoxaparin Prophylactic dose Intermediate dose# Anticoagulant dose	104(96%) 49(45%) 13(/12%) 42 (39%)	63(96%) 30(45%) 8(/12%) 26 (39%)	40(96%) 18(42%) 8(19%) 16 (37%)	0.631 0.610 0.583 0.325	104(96%) 49(45%) 13(/12%) 42 (39%)	39(96%) 18(45%) 5(12%) 16(39%)	64(96%) 28(42%) 13(19%) 25 (37%)	0.731 0.615 0.592 0.435
Warfarin/DOACs	4(4%)	2(3%)	2(5%)	0.531	4(4%)	1(2%)	3(4%)	0.612
SARS-CoV-2 vaccine	0(0%)	0(0%)	0(0%)	0.999	0(0%)	0(0%)	0(0%)	0.999
Panel B: treatment-related parameters
	Supine				PP
	All (n = 108)	High MP_wal (n = 66)	Low MP_wal (n = 42)	P high vs. low	Low MP_wal (n = 67)	High MP_wal (n = 41)	Low MP_wal (n = 67)	P high vs. low
Duration of the 1^st PP session (hrs)	–	–	–	–	11.8(10.1, 13.9)	12.1(10.1, 14.0)	11.8(10.1, 13.9)	0.571
Daily hours of PP therapy at 28 days	–	–	–	–	16.0( 13.9, 18.1)	16.9(14.7, 18.9)	16.0( 13.9, 18.1)	0.439
Days of PP therapy	–	–	–	–	5(4, 7)	5(4, 7)	5(4, 7)	0.312
Patients completing ≥ 1 PP session/day for ≥ 2 consecutive days (n, %)	–	–	–	–	108(100%)	41(100%)	67(100%)	0.999
Dexmedetomidine infusion during NIV (n, %)	38(35%)	24(36%)	14(33%)	0.712	82(76%)	32 (78%)	50(75%)	0.813
Daily hours of NIV
Day 1 Day 0–7^§	24(24, 24) 21.2(20.0, 22.9)	24(24, 24) 21.6(20.4,22.9)	24(24, 24) 19.7(18.4,22.5)	0.999 0.513	24(24, 24) 20.6(19.6,22.8)	24(24, 24) 21.5(20.4,22.7)	24(24, 24) 20.2(18.6,22.8)	0.999 0.495
Days of NIV	8(6, 11)	8(6, 11)	7(6, 10)	0.316	7(6, 10)	7(6, 10)	7(6, 10)	0.316
Ventilatory mode(%):
CPAP/PSV
Day 0 Day 1 Day 7	52%/48% 52%/48% 58%/42%	57%/43% 57%/43% 59%/41%	58%/42% 58%/42% 61%/39%	0.417 0.417 0.327	52%/48% 52%/48% 60%/40%	56%/44% 56%/44% 59%/41%	57%/43% 57%/43% 71%/40%	0.649 0.649 0.472
Interface(%):
Full-face mask/helmet
Day 0 Day 1 day 7	100%/0% 100%/0% 96%/4%	100%/0% 100%/0% 94/6%	100%/0% 100%/0% 98%/2%	0.999 0.893 0.831	100%/0% 100%/0% 96%/4%	100%/0% 100%/0% 93%/7%	100%0% 100%/0% 97%/3%	0.999 0.999 0.749
PEEP (cmH2O)
Day 0 Day 1 Day 0–7^§	8(8, 9) 8(8, 9) 8(7, 10)	8(8, 9) 8(8, 9) 9(8, 10)	8 (7,8) 8(7,8) 8(7.8)	0.813 0.821 0.192	7 (7, 9) 7(7, 9) 7(6. 8)	8(7, 9) 8(7, 9) 8(7,9)	7 (6, 8) 7(6, 8) 7(6. 7)	0.612 0.297 0.103
Pressure Support(Ps) during PSV (mH2O)
Day 0 Day 1 Day 0–7^§	4(3, 5) 4(3, 5) 4(3, 5)	4(3, 5) 4(3, 5) 4(3, 5))	3(3, 4) 3(3, 4) 3(3, 5)	0.649 0.418 0.214	4(3, 5) 4(3, 5) 4(3, 5)	4(3, 5) 4(3, 5) 4(3, 5)	3(3, 4) 3(3, 4) 3(3, 5)	0.492 0.317 0.118
NRS (Dyspnoea)
Day 0 Day 1 Day 0–7^§	2(0,3) 2(0,4) 2(1, 3)	2(0,3) 2(0,4) 3(1, 4)	2(0,3) 2(0,3) 1(0, 2)	0.712 0.429 0.034	2(0,3) 2(0,3) 2(0, 4)	2(0,3) 2(0,4) 3(1, 4)	2(0,3) 2(0,3) 1(0, 2)	0.839 0.736 0.021
WOB scale
Day 0 Day 1 Day 0–7^§	2(1, 3) 2(1,3) 3(2, 4)	2(1, 3) 2(1,3) 3(2, 4)	2(1, 3) 2(1,3) 1(1, 2)	0.629 0.582 0.035	2(1, 3) 2(1,3) 1(1, 2)	2(1, 3) 2(1,3) 3(2,4)	2(1, 3) 2(1,3) 1(1, 2)	0.412 0.493 0.022
FiO2
Day 0 Day 1 Day 0–7^§	0.50(0.45, 0.70) 0.50(0.45, 0.70) 0.50(0.40, 0.60)	0.50(0.45,0.70) 0.50(0.45,0.70) 0.60(0.50,0.75)	0.50(0.45,0.70) 0.50(0.45,0.70) 0.40(0.35,0.50)	0.783 0.698 0.028	0.50(0.45,0.70) 0.50(0.45,0.70) 0.45(0.35,0.50)	0.50(0.50,0.70) 0.50(0.50,0.70) 0.50(0.50,0.70)	0.50(0.45,0.70) 0.50(0.40,0.70) 0.40(0.35,0.50)	0.712 0.512 0.031
VTe (ml/kg prBW)
Day 0 Day 1 Day 0–7^§	7.6 (7.1, 8.4) 7.6(7.2, 8.8) 7.7(7.3, 8.3)	8.0(7.6, 8.1) 7.8(7.2, 8.8) 8.0 (7.2, 8.8)	7.3(6.5, 8.4) 7.2(6.7, 8.6) 7.4(6.8, 8.6)	0.412 0.272 0.231	7.5(6.7, 8.5) 7.4(6.6, 8.5) 7.6 (6.8, 8.6)	7.7(6.9, 8.6) 7.6(7.1, 8.4) 7.9 (7.1, 8.6)	7.4(6.5, 8.5) 7.2(6.6, 8.4) 7.6 (6.8, 8.6)	0.595 0.384 0.611
RR(breaths/min)
Day 0 Day 1 Day 0–7^§	25(22, 30) 24(22, 26) 26(22, 28)	27(24, 30) 26(22, 28) 26(22, 28)	24(22, 28) 22(20, 24) 21(20, 25)	0.712 0.39 0.018	25(22, 30) 24(20, 27) 22(20, 25)	25(22, 30) 26(22, 28) 26(22, 27)	24(22, 26) 22(20, 25) 21(20, 25)	0.837 0.391 0.021
MV (L/min)
Day 0 Day 1 Day 0–7^§	14.2(11.1, 16.1) 13.6(12.6, 15.2) 14.0(11.4, 16.1)	14.2(11.1,16.1) 13.9(12.6,15.4) 14.4(11.9,16.1)	13.4(11.0,16.9) 13.1(10.6,15.3) 13.0(10.2,15.6)	0.817 0.534 0.159	13.6(11.1,16.7) 13.3(10.9,15.3) 13.9(12.1,15.9)	13.9(11.9,15.6) 13.8(11.9,15.7) 13.9(12.1,15.9)	13.6(11.1,16.7) 13.1(10.6,15.3) 11.9(11.1,15.9)	0.412 0.239 0.118
Time from NIV initiation to 1st PP session(hrs)	–	–	–	–	3(2, 4)	3(2, 4)	3(2, 4)	-
MPsp0 (J/min)	17.5(11.3,20.0)	29.7(20.5,28.7)	9.1(6.9,13.4)	< 0.001	18.3(15.4,24.9)	29.9(25.5,33.9)	13.1(10.1,16.9)	< 0.001
MP_PBW sp0 (J/min/kg)	0.21(0.13,0.26)	0.24(0.16,0.28)	0.13(0.10,0.27)	< 0.001	0.20(0.16,0.27)	0.25(0.17,0.29)	0.17(0.11,0.24)	< 0.001
MP_WALsp0 (J/min/L)	9.9(8.0, 13.2)	15.8(10.9,28.0)	6.1(4.4, 8.9)	< 0.001	9.8(8.1, 13.1)	17.1(11.2,21.7)	8.1(501, 11.9)	< 0.001
MP (J/min)
Day 1 Day 2 Day 1–7	16.7(10.5,19.2) 16.1(10.8,19.8) 16.6(10.1,19.5)	27.8(24.5,31.9) 28.1(24.9,32.2) 27.6(24.1,31.5)	8.7(6.3,10.2)8.8(6.6,10.1) 9.1(8.2, 10.4)	< 0.001 < 0.001 < 0.001	16.9(10.8,19.3) 17.2(10.4,21.1) 16.8(10.2,20.4)	27.5(25.1,31.9) 27.8(25.0,32.2) 27.4(24.2,33.1)	8.5(6.5,9.2) 8.9(6.8,9.6) 8.8(6.4, 9.4)	< 0.001 < 0.001 < 0.001
MP_PBW (J/min/kg)
Day 1 Day 2 Day 1–7	0.19(0.10, 0.23) 0.18(0.12, 0.23) 0.19(0.10, 0.20)	0.23(0.17,0.26) 0.23(0.17,0.26) 0.23(0.12,0.26)	0.12(0.10,0.29) 0.13(0.11,0.29) 0.13(0.11,0.21)	< 0.001 < 0.001 < 0.001	0.18(0.16,0.22) 0.19(0.15,0.22) 0.18(0.15,0.24)	0.23(0.17,0.27) 0.22(0.16,0.27) 0.23(0.20,0.29)	0.12(0.10,0.29) 0.12(0.12,0.29) 0.12(0.11,0.22)	< 0.001 < 0.001 < 0.001
MP_WAL (J/min/L)
Day 1 Day 2 Day 1–7	9.3(5.7,21.2) 9.0(5.1,20.6) 8.6(4.7,18.9) ‡	15.7(12.1,27.1) 15.6(12.1,29.0) 18.5(14.7,22.9) ‡	5.3(4.1, 7.4) 5.1(3.7, 7.1) 4.6(3.6, 6.2) ‡	< 0.001 < 0.001 < 0.001	8.9(5.5,20.0) 8.9(5.2, 19.4) 9.9(6.7,18.9) ‡	16.2(13.8,23.9) 16.0(12.9,22.1) 20.6(16.1,26.3) ‡	4.9(4.2, 7.9) 4.7(4.1, 7.9) 4.5(3.6,5.9) ‡	< 0.001 < 0.001 < 0.001
pH
Day 0 Day 1 Day 0–7^§	7.48(7.45, 7.50) 7.46(7.44, 7.48) 7.46(7.43, 7.49)	7.47(7.45,7.49) 7.46(7.44,7.47) 7.46(7.43,7.49)	7.47(7.42,7.48) 7.46(7.41,7.49) 7.43(7.40,7.46)	0.471 0.413 0.394	7.47(7.45,7.49) 7.46(7.41,7.47) 7.45(7.43,7.47)	7.46(7.42,7.50) 7.45(7.42,7.48) 7.44(7.40,7.48)	7.47(7.44,7.48) 7.45(7.41,7.47) 7.42(7.40,7.46)	0.498 0.479 0.309
paCO2 nnHg
Dat 0 Day 1 Day 0–7^§	35(33, 39) 35(35, 40) 36(35, 40)	35(33, 39) 35(35, 40) 36(35, 40)	35(34, 40) 34(32, 39) 35(32, 40)	0.391 0.437 0.562	35(34, 40) 34(32, 39) 35(32, 39)	35(33, 39) 35(34, 40) 35(34, 39)	35(34, 40) 34(32, 39) 37(32, 39)	0.713 0.316 0.513
PaO2/FiO2 ratio
sp0 sp1 sp0-7 ^§	147(116, 218) 155(135, 219) 178(142, 197) ‡	149(115, 220) 143(122,193) 150(127, 199)	142(110, 205) 168(130, 233) ‡ 185(168, 221) ‡	0.562 0.009 0.002	142(116, 192) 181(139, 219) ‡ 198(159,231) ‡‡	143(114, 192) 152(118, 180) 155(128, 192)	141(119, 193) 198(152, 232) ‡ 220(163, 273) ‡‡	0.629 0.001 < 0.001
O2 responders n (%) (paO2/FiO2 sp0-1 > 0)	42(39%)	16(34%)	26(62%)	0.009	71(66%)^∏	17(41%)	54(81%)	0.002
VR^≠
sp0 sp1 sp0-7^§	1.9(1.5, 2.4) 1.9(1.6, 2.3) 1.9(1.5, 2.4)	1.9(1.6, 2.3) 2.2(1.9, 2.3) ‡ 2.3(1.8, 2.4) ‡	2.0(1.4, 2.4) 1.6(1.2, 2.3) ‡ 1.6(1.0–2.2) ‡	0.735 0.010 0.008	1.9(1.5, 2.3) 1.6(1.4, 2.0) ‡ 1.5(1.1–1.9) ‡	1.9(1.5, 2.3) 1.9(1.6, 2.3) 1.9(1.6, 2.3)	1.9(1.6, 2.3) 1.5(1.3, 2.0) ‡ 1.4(0.9, 1.7) ‡‡	0.512 0.009 < 0.001
MVcorr (L/min)^≠
sp0 sp1 sp0-7^§	12.1(10.1, 13.7) 12.5(10.2, 14.9) 12.5(10.2, 15.1)	12.2(10.1,13.7) 12.7(10.2,14.9) ‡ 13.2(11.2,15.3) ‡	11.9(10.1,13.7) 10.9(9.3, 12.6) ‡ 10.9(9.0, 12.3) ‡	0.638 0.009 < 0.001	11.9(10.1,13.7 11.0(9.8, 12.6) ‡ 11.1(9.9, 12.3) ‡	12.3(11.4,13.4) 12.1(10.5,13.5) 12.2(10.4,14.0)	11.9(10.1,13.7) 10.8(9.3, 12.6) ‡ 10.5(8.9, 12.3)	0.638 0.009 < 0.001
CO2 responders n(%) (VR sp0-1 < 0)	32(30%)	11(16%)	21(50%) ≠	0.004	47/66(71%)^∏	12(29%)	47(71%) ≠	< 0.001
Ventral LUS
score Day 0 Day 1 Day 0–7^§	6(5, 9) 6(5,7) 6(5,7)	7(5, 9) 7(5,9) 7(5,9)	6(4, 8) 5(5, 7) 4(3, 7)	0.496 0.514 0.138	7(5, 8) 7(5, 8) 8(4, 7)	7(5, 8) 7(5, 8) 8(5, 7)	7(5, 8) 7(5, 8) 7(4, 7)	0.496 0.514 0.138
Lateral LUS
score Day 0 Day 1 Day 0–7^§	8(7, 9) 8(7, 9) 8(7, 10)	8(7, 9) 8(7, 9) 8(7, 10)	8(7,9) 7(6, 8) 7(6, 8)	0.839 0.010 0.028	8(7,9) 7(7, 8) 6(5, 7)	8(7, 9) 8(7, 9) 6(5, 10) ‡	8(7,9) 6(7, 8) ‡ 6(5, 7) ‡	0.839 0.010 < 0.001
Dorsal LUS score
Day 0 Day 1 Day 0–7^§	10(9, 11) 10(9, 11) 10(9, 12)	10(9, 11) 11(9, 12) 11(9, 13)	10(9, 11) 10(7, 11) 10(8, 12)	0.735 0.041 0.010	10(9, 11) 8(7, 10) ‡ 6(5, 8) ‡‡	10(9, 11) 10(9, 11) 8(6, 10) ‡	10(9, 11) 8(7, 10) ‡ 6(5, 8) ‡‡	0.735 0.009 < 0.001
Global LUS score
Day 0 Day 1 Day 0–7^§	24(22, 28) 24(20, 27) 24(20, 27)	25(22, 27) 26(21, 27) 26(21, 28)	24(22, 28) 22(18, 25) ‡ 21(18, 24) ‡	0.851 0.004 < 0.001	25(21, 28) 22(19, 26) ‡ 21(18, 24) ‡‡	25(22, 28) 25(22, 28) 22(20, 26) ‡	25(21, 28) 21(18, 24) ‡ 19(16, 22) ‡‡	0.719 0.006 < 0.001
Global LUS reaeration score
Day 1 Day 1–7^§	0(-4, 1) -1(-5, 1)	0(-7, -2) -6(-9, -4)	1(-1, 4) 5(2, 8)	0.021 0.008	4(3, 8) 8(4, 11)	2(-1, 5) 5(2, 8)	6(5, 10) 10(6, 14) ‡‡	0.009 < 0.001
Global LUS reaeration score ≥ 8 (n,%)
Day 1 Day 1–7^§	9(8%) 20(19%)	0(0%) 0(0%)	9(21%) 19(45%)‡	0.002 < 0.001	26(24%) 59(55%)‡‡	1(2%) 11(26%)	25(37%) 48(71%) ‡‡	0.008 < 0.001
WAL (%)
Day 0 Day 1 Day 0–7§	50(43, 59) 48(41, 60) 49(38, 62)	50(46, 61) 46(41, 57) 38(34, 56) ‡‡	50(41, 61) 50(42, 62) 59(49, 69) ‡‡	0.629 0.012 < 0.001	49(41, 59) 55(46, 62) ‡ 58(52,64) ‡‡	49(39, 60) 49(40, 60) 51(42, 62)	49(45, 59) 57(51, 64) ‡ 64(55, 70) ‡‡	0.798 0.009 < 0.001
Blood lactate (mmol/L)
day 0 day1 day 0–7 §	1.0 (0.9, 1.1) 1.1(1.0, 1.2) 1.2(1.0, 1.3)	1.0 (0.9, 1.1) 1.1(1.0, 1.2) 1.2(1.0, 1.3)	1.0 (0.9, 1.1) 1.1(1.0, 1.2) 1.1(1.0, 1.2)	0.734 0.513 0.317	1.0 (0.9, 1.2) 1.1(1.0, 1.2) 1.1(1.0, 1.2)	1.0 (0.9, 1.2) 1.1(1.0, 1.2) 1.2(1.1, 1.3)	1.0 (0.9, 1.2) 1.1(1.0, 1.2) 1.1(1.0, 1.2)	0.712 0.462 0.479
C-reactive protein (mg/dL)
day 0 day1 day 0–7^§	14.5(10.5, 19.0) 13.4(9.9, 18.9) 11.7(6.2, 14.1) ‡	14.7(10.2,18.9) 15.2(10.5,19.2) 13.9(10.2,16.4)	14.2(10.9,19.0) 11.5(7.5,14.5)‡ 7.4(2.1, 9.0) ‡‡	0.817 0.011 < 0.001	13.5(10.5,19.2) 9.8(7.5, 12.5) ‡ 4.5(3.1, 8.0) ‡‡	13.2(10.6,18.9) 12.1(8.8, 19.1) 9.8(7.3, 16.1)	13.9(10.5,19.2) 8.4(4.6, 12.3) ‡ 1.5(3.1, 1.0) ‡‡	0.813 < 0.001 < 0.001
Procalcitonin (ng/mL)††
day 0 day1 day 0–7 ^§	0.52(0.40, 0.62) 0.41(0.31, 0.69) 0.31(0.17, 0.45) ‡	0.54(0.40,0.62) 0.51(0.41,0.69) 0.51(0.17,0.65)	0.58(0.41,0.67) 0.35(0.15,0.55) ‡ 0.10(0.02,0.20) ‡‡	0.612 0.007 < 0.001	0.57(0.41,0.67) 0.35(0.15,0.55)‡ 0.25(0.12,0.37) ‡‡	0.56(0.40,0.65) 0.49(0.42,0.63) 0.46(0.39,0.61)	0.58(0.41,0.67) 0.15(0.05,0.25) ‡ 0.05(0.02,0.10) ‡‡	0.612 0.002 < 0.001
D-dimer (mg/L)
day 0 day 1 day 0–7^§	3.1(1.5, 3.9) 3.1(1.9, 4.0) 1.9(1.0, 4.9) ‡	3.1(1.7, 3.9) 3.5(1.9, 4.1) 3.9(2.0, 4.9)	2.9(1.9, 3.8) 2.5(1.8, 3.2) 0.9(0.3, 1.4) ‡	0.629 0.215 0.002	2.8(1.4, 3.4) 2.6(1.4, 3.2) 1.0(9.3, 1.9) ‡‡	2.9(1.7, 3.8) 3.0(1.8, 3.9) 2.9(1.5, 3.9)	2.7(1.4, 3.4) 2.4(1.4, 3.2) 0.7(0.3, 0.9) ‡‡	0.712 0.215 0.002
Panel C: Clinical outcomes
	Supine				PP
	All (n = 108)	High MP_wal (n = 66)	Low MP_wal (n = 42)	P high vs. low	All (n = 108)	High MP_wal (n = 41)	Low MP_wal (n = 67)	P high vs. low
NIV failure at 28 d	56 (52%)	47(70%)	9(21%) ≠	< 0.001	20(19%)^∏	17(42%)	3(4%) ≠	< 0.001
Days to NIV failure	7(6, 7)	7(6, 7)	8(7, 8)	0.713	8(7, 8)	7(6,7)	8(7, 8)	0.227
Death at 28 d	52(48%)	45(67%)	7(17%)	< 0.001	16(15%)^∏	15(37%)	1(2%) ≠	< 0.001
Days to death	12(9, 15)	12(9, 15)	15(14, 18)	0.213	15(14, 18)	13(11, 15)	15(14, 18)	0.139
ETI at 28 d	38/99 (38%)	34/62 (55%)	4/37(11%)	0.004	11/97 (11%)	9/36(25%)	2/61 (3%)**	0.011
Days to ETI	7(6, 7)	7(6, 7)	8(7, 8)	0.529	8(7, 8)	7(6,7)	8(7, 8)	0.239
Death at 60 d	52(48%)	45(67%)	7(17%)	< 0.001	17(15%)^∏	16(37%)	1(2%) ≠	< 0.001
Days in ICU	9(2, 16)	9(2, 16)	6(2, 10)	0.129	6(2, 10)	8(2, 12)	6(2, 10)	0.198

Open in a new tab

For each parameter median (IQR) is indicated, unless otherwise specified

The P values refer to comparison between low MP_WAL and high MP_WAL group within supine and PP treatment groups (statistically significant p-values are written in bold characters). Data are expressed as median (IQR) or as n(%) values, unless otherwise specified

^*P < 0.05 vs. High MP_WAL (day 1), within the same treatment group (PP or supine)

^** P < 0.01 vs. High MP_WAL (day 1), within the same treatment group (PP or supine)

≠ P < 0.001 vs. High MP_WAL (day 1), within the same treatment group (PP or supine)

ǁ P < 0.05 vs. supine group

^¶ P < 0.01 vs. supine group

^∏P < 0.001 vs. supine group

^† P < 0.01 vs. timepoint sp0

^‡ P < 0.01 vs. day 0

^‡‡ P < 0.001 vs. day 0

# on admission, while on a FiO2 50% Venturi mask or a non-rebreather reservoir bag mask

## prophylactic dose twice daily

^†† patients with documented infection (n = 4 within each group) were excluded from this analysis

Abbreviations: ABG: arterial blood gas analysis; DOAC: direct oral anticoagulants; ISARIC: International Severe Acute Respiratory Infection Consortium; MP: mechanical power; MP_PBW: mechanical power corrected for predicted body weight; MP_WAL: mechanical power corrected for well-aerated lung volume; NAL: non-aerated lung volume; NIV: noninvasive ventilation; NRS: numeric rating scale; PAL: poorly aerated lung volume; PP: prone position; RR: respiratory rate; SAPS: Simplified acute physiology score; WAL: well-aerated lung volume; WOB( work of breathin scale, as determined by Apigo scale

For each parameter median (IQR) is indicated, unless otherwise specified

The P values value refer to comparison between groups at baseline, at the end of follow-up and to comparison in changes during the follow-up, respectively (statistically significant p-values are written in bold characters). Data are expressed as mean ± SEM

Definitions:

Baseline: day of enrollment, before NIV initiation

^§Day 0: baseline, i.e. the time of NIV initiation in both treatment groups

^§Day 1: the first 24 h of NIV in both (PP and supine) treatment groups

^§Day 7: day 7 after NIV initiation in both (PP and supine) treatment groups

^§Day 0–7: mean value through day 0–7. Between-group comparison was made with repeated measures two-factor (within subject and between group) ANOVA, after log-transformation of non-normal variables

sp0: ≥ 1 h since (supine) NIV initiation on day 0

sp1: 24 h after NIV initiation in both (PP and supine) treatment groups

O2 response: change in paO2/FiO2 ratio during NIV in supine position at day 1 (timepoint sp1) as compared with timepoint sp0

A positive O2 response is defined by: pO2/FiO2sp1—pO2/FiO2sp0 > 0, or ∆paO2/FiO2sp0-1 > 0

CO2 response: an increase in CO2 clearance, defined by a reduction in dead space indices (VR and MV_corr, tested separately) at sp1 as compared with timepoint sp0

A positive CO2 response is defined by: VRsp1-VRsp0 < 0 (or ∆VRsp0-1 < 0) and by MV_corrsp1 – MV_corrsp0 < 0 (or ∆MV_corrsp0-1 < 0)]

This schedule allowed comparing gas exchange responses among groups every day, in the same (supine) position, after taking into account the effect of the overnight PP session in the PP groups

For VTe, RR, MV, MP measures, time weighted–average of hourly values during each session are presented

For MV, we verified consistency between time-average MV values and values obtained at the time of ABG (which were used to calculate Dead Space Indices): the comparison confirmed a good agreement between the two values, with p-value > 0.6 for comparison at all timepoints (not shown)

Ventral LUS (lung ultrasound) score: the score is the sum of individual scores for region L1, L2, R1, R2 (ranging 0–12)

Lateral LUS score: the score is the sum of individual scores for region L3, L4, R3, R4

Dorsal LUS score: the score is the sum of individual scores for region L5, L6, R5, R6

Global LUS score: the score is the sum of ventral, lateral and dorsal region scores (the sum of the 12 regions’ score, ranging 0–36)

LUS reaeration score, global: the sum of ventral, lateral and dorsal reaeration scores, calculated according to the method proposed and validated by Bouhemad et al

WAL(%): well-aerated lung volume as % total lung volume, resulting from the sum f regional WAL volumes assessed by LUS and validated against CT scan (see online supplement)

Outcomes

Primary outcomes were the occurrence of.

NIV failure within 28 days of enrolment, defined as intubation OR death;
death, censored at 28 days after enrollment.

Secondary outcomes were:

endotracheal intubation (ETI) at 28 days (after excluding patients with a do-not-intubate, DNI, order);
60-day death
O2-response: paO2/FiO2_sp1–paO2/FiO2_sp0 (or ∆paO2/FiO2_sp0-1).
CO2-response: ∆VR_sp0-1;
C-reactive protein (CRP) response: ∆CRP_0-1;
global LUS response at day 1: ∆global LUS_0-1;
global reaeration score at day 1: global reaeration score_1;
change in LUS- assessed WAL (%) at day 1 (∆WAL_0-1)

Statistical analyses

Sample size calculation and propensity score (PS)-matching of PP and supine group.

Sample size calculation and PS-matching of PP and supine groups for relevant baseline covariates are detailed in supplementary text.

Descriptive statistics

Data are given as median (IQR) or n (%) as appropriate.

We used chi-square test or Fisher´s exact test for categorical variables, T-Test for normally distributed and Kruskal–Wallis test for non-normal continuous variables.

Time change in continuous variables was assessed by computing the AUC with the trapezoid method [36].

Repeated measures two-factor (within subject and between group) ANOVA was used to compare continuous variables assessed at multiple timepoints (i.e., respiratory and biochemical parameters), after log-transformation of non-normal variables.

To explore the effect of early MP_WAL delivery on clinical outcomes, the whole cohort was split into 2 groups according to median MP_WAL values at day 1. The probability of 28-day NIV failure, death and ETI in low vs high MP_WAL group at day 1 was compared using Kaplan–Meier procedure and log-rank test. Data were analyzed on an intention-to-treat basis.

Beside categorizing patients into low/high MP_WAL, we explored dose–response relationship between early power delivery and clinical outcomes by dividing the entire cohort into quartiles of power measures at day 1. Comparison between quartiles was made by ANOVA with post-hoc comparison from the first quartile performed using the Tukey test. We planned to assess also reciprocal relationship between MP, MP_PBW and MP_WAL at day 1 by univariate analysis and Spearman correlation coefficient (r_s).

Multivariable Cox proportional regression analysis adjusting for imbalanced covariates between groups was used to assess the effect of confounders on 28-d NIV failure, death and ETI in the whole cohort, with the maximum number of covariates allowed in each model set at (event rate x N)/10, where N is the sample size [37]. The allocation assignment (PP or supine) was entered as a predefined covariate into the models. Calendar month of admission was forced into all models to account for unmeasured temporal disease trends during the pandemic.

Anticipating high collinearity between ventilatory variables, in all (Cox and linear) multivariable models we used a combination of backward procedure and exclusion of highly collinear variables through model-dependent Variance Inflation Factor(VIF) cut-off values to select covariates [38].

MP, MP_PBW and MP_WAL at day 1 were included into all backward multivariate models to assess relative robustness of the association of each power measure (MP, MP_PBW, MP_WAL) with outcomes.

We also explored the predictive performance of each power measure and of other ventilatory variables at day 1 for 28-day NIV failure and death and the optimal cut-offs using the area under receiver operating characteristic curve (AUROC) analysis and Youden index.

Comparison between AUROC curves was made by DeLong’s method.

In the physio-anatomical analysis, we explored dose–response relationship between the power delivered at day 1 by the ventilator to the respiratory system and gas exchange, ultrasonographic and circulating biomarker changes after the 1^st 24 h of NIV by univariable and multivariable regression analysis, after log transformation of skewed parameters; the best fit among four predictive models (linear, exponential, logarithmic, binomial) was searched using R² values.

Two-tailed p values < 0.05 were considered statistically significant (MedCalc 19.7, Ostend, Belgium).

Results

Characteristics at Inclusion

We analyzed 216 COVID-19 patients (108 treated with PP and 108 with supine NIV) (Table 1, Supplementary Fig. 2). Pre-post-matching SMDs, PS density and logit(PS) distribution plots revealed good balance between PP and supine groups in selected covariates (Supplementary Fig. 3, Supplementary Table 2).

Baseline demographics, pharmacological therapies, clinical-radiological features, and respiratory parameters of patients treated with PP or supine NIV, subgrouped according to median MP_WAL at day 1, were similar across groups; NRS (dyspnea), WOB score [39] and RR showed no evidence for increased respiratory drive, while lung CT scans showed largely predominant ground-glass opacities (Table 1).

LUS lung aeration assessment

The median (IQR) global LUS severity score was 25 (21, 30), closely correlating with global CT severity score (r_s = 0.83; p < 0.001: Supplementary Fig. 4).

Regional LUS categorization closely correlated with CT-assessed gas/tissue density, with each step increase of LUS score being associated with a significant increase in lung tissue density (Fig. 1 panel A): CT-LUS agreement was almost complete for WAL (k = 0.91; 95%CI: 0.85–0.97), substantial for PAL(k = 0.61; 95%CI: 0.57–0.66) and moderate (k = 0.55; 95%CI: 0.51–0.60) for NAL (Supplementary Table 3A–B), consistent with previous reports in ARDS [34].

Based on these findings, an equation was developed to predict WAL(%) from global LUS score (Fig. 1 panel B).

Treatment-related parameters and outcomes

No patient was lost to follow-up and there were no missing data for evaluated end-points.

Patients initiated NIV and PP within 24 h of admission to the Subintensive Care Unit. Over the initial 48 h of treatment, NIV was delivered continuously or until intubation and only brief interruptions were allowed for eventual adjustments and nursing care, lasting no more than few minutes; subsequently, daily breaks, lasting no more than 2 h, were allowed for meals and nursing care, depending on patient clinical condition and tolerance.

In PP and supine groups, ventilatory settings were similar among groups at baseline and remained unchanged from timepoints sp0 to sp1 (Table 1).

Daily hours of NIV and PP, the duration of the longest PP session, daily PP sessions and days of PP therapy are reported in Table 1.

Compared with supine group, PP therapy was associated with improved gas exchange and ultrasonographic parameters and clinical outcomes (Supplementary Fig. 5–6, Supplementary Table 4).

Effect of PP therapy on MP_WAL

PP therapy was associated with a significant MP_WAL reduction: MP_WAL decreased by a median 34% (IQR: 30–38%) during the first PP session (timepoint pp1) as compared with pre-proning values (timepoint sp0) (Fig. 2 panel A–D). MP_WAL reduction during 1^st PP session was driven mainly by a RR decrease and secondly by a modest (5%; 2–7%) WAL increase (Fig. 2A–D). The magnitude of WAL % increase during 1st PP session was consistent with recent reports [40-42].

Fig. 2 — Panel A–D: time course of MP_WAL, MP and their components during the initial 7 days of NIV in the PP (n = 108) and supine (n = 108) group. N at risk and median (IQR) values of PEEP, Pressure support (Ps) and well-aerated lung volume (WAL) are shown at the bottom of panel A. # P < 0.01 vs. baseline (same group); § P < 0.01 vs. baseline (both groups) † P < 0.001 vs. baseline (both groups) ǁ P < 0.01 between PP and supine groups ¶ P < 0.001 between PP and supine groups. In the box plots the middle line represents the median observed value, boxes represent the interquartile range (IQR), whiskers extend to the most extreme observed values with 1.5 times the interquartile range of the nearer quartile, and dots represent observed values outside that range. The connecting line connects the mean values within each box plot at different time points. For VTe, RR, MV, MP, MP_WAL time weighted–average of hourly values during each session (supine/prone) of NIV are presented. For MV, we verified consistency between time-average MV values and values obtained at the time of ABG (which were used to calculate Dead Space Indices): the comparison confirmed a good agreement between the two values, with p-value > 0.7 for comparison at all timepoints (not shown). The timepoints are indicated as follows: sp0: supine position timepoint 0 (baseline, after NIV initiation).In PP group, the session preceded the first PP session. ABGs were performed ≥ 1 h after supine NIV initiation, after achieving ventilatory stability; pp1: first PP session in PP group. ABGs were performed ≥ 1 h after prone NIV initiation, after achieving ventilatory stability; sp1: supine position timepoint 1 (after the initial 24 h of NIV); in the PP group this corresponded to the resupination after the 1^st PP session) on day 1. pp2: during the 2^nd PP session; sp2: supine position timepoint 2 (after the 2^nd day of NIV, in the PP group this also corresponded to the resupination after the 2^nd PP session on day 2; pp3 during the 3^rd PP session; sp3 supine position timepoint 3 (after the 3^rd day of NIV, in the PP group this also corresponded to the resupination after the 3^rd PP session on day 3. pp4 during the 4^th PP sessionx; sp4 supine position timepoint 4 (after the 4^th day of NIV, in the PP group this also corresponded to the resupination after the 4^th PP session on day 4. pp5 during the 5^th PP session; sp5 supine position timepoint 5 (after the 5^th day of NIV, in the PP group this also corresponded the 5^th PP session on day 5. pp6 during the 6^th PP session; sp6 supine position timepoint 6 (after the 6^th day of NIV, in the PP group this also corresponded to the resupination after the 6^th PP session on day 6. pp7 during the 7^th PP session; sp7 supine position timepoint 7 (after the 7^th day of NIV, in the PP group this also corresponded to the resupination after the 7^th PP session on day 7. This schedule allowed comparing gas exchange responses among group at day 1, after 24 h of NIV, in the same (supine) position, after taking into account the effect of the first overnight PP session (timepoint pp1) in the PP groups

Daily MP_WAL swings between prone and supine position subsided after day 5(timepoint pp5)(P supine vs. prone > 0.05), suggesting no additional effect on MP_WAL from further PP days (Fig. 2A).

Relationship of early (day 1) power measures with clinical outcomes

We first assessed the relationship between different power measures at day 1 and clinical outcomes (NIV failure, death) (Fig. 3A–C): we observed a progressive, linear, statistically significant increase in NIV failure and death rates across each MP_WAL quartile, but not across MP and MP_PBW quartiles, which rather showed a steep increase in the rate of clinical outcomes in the 3-4^th quartile as compared with the lower 2 quartiles.

The investigation of mutual relationship between MP, MP_PBW and MP_WAL at day 1 disclosed a the weak correlation between MP_WAL and the other power measures, indicating for any given value of MP delivered by NIV to the respiratory system, the energy per unit volume of WAL varied widely (Fig. 3D–E).

Early (day 1) MP_WAL is independently associated with clinical outcomes

We then categorized the whole study population based on median MP_WAL delivered during the initial 24 h of NIV (day 1) into high MP_WAL (i.e. MP_WAL ≥ 9.1 J/min/L) or low MP_WAL (i.e., MP_WAL < 9.1 J/min/L): patients receiving a low MP_WAL(day 1) showed lower 28-d NIV failure, intubation and death rates than those receiving a high MP_WAL(day 1) (Table 1, Fig. 4).

Fig. 4 — Cumulative incidence of Noninvasive Ventilation (NIV) failure (Panel A), death (Panel B) and endotracheal intubation (Panel C) at 28 days after enrollment in patients grouped according to high (≥ 9.1 J/min/L) vs low (< 9.1 J/min/L) MP_WAL at day 1. In the endotracheal intubation group, patients with a Do-Not-Intubate (DNI) disposition were excluded

In Cox proportional multivariable models, MP_WAL (day 1) predicted 28-d NIV failure (HR 1.682; 95%CI: 1.147–2.412, p = 0.003), intubation (HR 1.521; 95%CI: 1.193–2.314, p = 0.009) and death (HR 1.697; 95%CI: 1.217–2.313, p = 0.002) independently of baseline illness and lung disease severity and of other ventilatory variables (Table 2).

Table 2.

Cox multivariable analysis of predictors of 28-d NV failure, death and endotracheal intubation (ETI) (n = 216)

28-d NIV failure
SAPS II	1.027	0.983 to 1.272	0.293
PP therapy (1 = yes, 0 = No)	0.581	0.392 to 0.772	0.009
Baseline pO2/FiO2*	0.982	0.734 to 1.696	0.128
Baseline WAL (L)	1.044	0.891 to 1.986	0.319
RR (day 1)	1.186	0.892 to 1.539	0.146
VTe(day 1) (ml/kgPBW)	1.017	0.749 to 2.013	0.318
PEEP(day 1)(mmHg)	1.179	0.874 to 1.528	0.412
MP_WAL (day 1)(J/min/L)	1.682	1.147 to 2.412	0.005
28-d death
Parameter	HR	95%CI	P
Age	1.088	1.0426 to 1.198	0.028
ISARIC 4C score	1.026	0.947 to 1.192	0.334
PP therapy	0.629	0.271 to 0.812	0.010
Baseline paO2/FiO2*	0.999	0.894 to 1.166	0.296
Baseline WAL (L)	1.039	0.792 to 1.893	0.427
RR (day 1)	1.116	0.975 to 1.428	0.239
MP_WAL (day 1)	1.697	1.217 to 2.316	0.002
28-d endotracheal intubation
SAPS II	1.021	0.872 to 1.812	0.513
PP therapy (1 = yes, 0 = No)	0.598	0.283 to 0.781	0.007
Baseline pO2/FiO2*	0.992	0.621 to 1.769	0.389
Baseline WAL (L)	1.028	0.891 to 1.986	0.319
RR (day 1)	1.112	0.783 to 1.702	0.192
VTe(day 1) (ml/kgPBW)	1.132	0.649 to 2.913	0.413
MP_WAL (day 1)(J/min/L)	1.682	1.147 to 2.412	0.005

Open in a new tab

Statistically significant p-values are written in bold characters

ISARIC international severe acute respiratory infection consortium, NIV noninvasive ventilation, SAPS simplified acute physiology score, PEEP positive end-expiratory pressure, MP_WAL, mechanical power normalized to well-aerated lung volume, CI confidence interval, PBW predicted body weight

^*on admission, while on a FiO2 50% Venturi mask or a non-rebreather reservoir bag mask

Predictive performance of MP_WAL(day 1) for clinical outcomes

Each ventilatory variable at day 1 was regressed on NIV failure and mortality in the whole study population: MP_WAL(day 1) showed the highest AUROC for predicting 28-d NIV failure (AUROC 0.89; 95%CI: 0.85–0.93, p < 0.001) and death (AUROC 0.89; 95%CI: 0.85–0.94, p < 0.001), outperforming all other ventilatory parameters and WAL (p < 0.011 for both comparisons by DeLong’s test) (Table 3, Fig. 5). Additionally, MP_WAL (day 1) optimal thresholds for both NIV failure and death overlapped.

Table 3.

Discrimination of ventilator parameters measured during the 1st 24 h of NIV)(day 1) in predicting 28-d NIV failure and death in the whole study population (n = 216)

	NIV failure at 28 days				Death at 28 days
Parameter	Cut-off	Se(%)	Sp(%)	AUROC (95%CI)	Cut-off	Se(%)	Sp(%)	AUROC (95%CI)
MP_WAL (day 1) (J/min/L)	9.2	88	86	0.89 (0.85–0.93) *^#	9.3	89	87	0.89 (0.85–0.94) *^#
MP (day 1) (J/min)	16.7	78	77	0.80 (0.72–0.85) *	20.9	80	73	0.76 (0.72–0.83) *
MP_PBW (day 1) (J/min/kg)	0.147	76	77	0.79 (0.72–0.85) *	0.169	84	76	0.79 (0.73–0.86) *
VTe(day 1) (ml/kg)	7.9	64	72	0.66 (0.58–0.73)	9.6	65	78	0.67 (0.59–0.74)
RR(day 1) (breaths/min)	28	75	61	0.65 (0.58–0.72)	31	69	60	0.64 (0.56–0.70)
PEEP(day 1) (cmH2O)	9	36	74	0.54 (0.47–0.61)	10	46	78	0.63 (0.55–0.70)
Ps(day 1) (cmH2O)	4	65	50	0.58 (0.47–0.68)	6	74	53	0.64 (0.53–0.74)
WAL(day 1) (L)	1.7	66	72	0.70 (0.62–0.78)	1.0	81	45	0.63 (0.55–0.71)

Open in a new tab

^* p < 0.001 vs VTe, RR, PEEP, Ps, WAL at day 1 by DeLong test

^# p < 0.01 vs. MP or MP_PBW at day 1 by DeLong test

ISARIC international severe acute respiratory infection consortium, NIV noninvasive ventilation, SAPS simplified acute physiology score, PEEP positive end-expiratory pressure, MP_WAL mechanical power normalized to well-aerated lung volume, MP mechanical power, AUROC area under receiver operating characteristic curve, CI confidence interval, PBW predicted body weight

Fig. 5 — Time-course of different ventilatory (Panel **A-B**), gas exchange(Panel C-F), ultrasonographic (Panel **G-N**) and inflammatory (Panel **O-P**) parameters during the initial 7 days of NIV in the PP and supine groups, grouped according to low MP_WAL or high MP_WAL during the 1st 24 h of NIV (day 1) (n = 216). * P < 0.05 vs baseline (same group) # P < 0.01 vs. baseline (same group) § P < 0.01 vs. baseline (both groups) † P < 0.001 vs. baseline (both groups) ǁ P < 0.01 between groups ¶ P < 0.001 between groups # P < 0.001 ‡

Time course of ventilatory and other physio-anatomical variables during the initial 7 days of NIV

Over the initial 7 days there were missing data in the physiological parameters due to the occurrence of NIV failure or success and subsequent Unit discharge. Because these data were not missing at random but due to the consequence of treatment effect, we did not perform multiple imputation and excluded missing values from analyses.

Figure 6 shows time course of different ventilatory, ultrasonographic and inflammatory parameters in PP and supine patients receiving either high or low MP_WAL(day 1).

Within each treatment group (PP or supine), median daily MP_WAL remained relatively constant during initial days of NIV (Fig. 6A–B); gas exchange parameters (paO2/FiO2 ratio, VR, MVcorr), ultrasonographic parameters (global LUS score, global reaeration score, WAL %) and inflammatory biomarkers (CRP, procalcitonin) improved significantly over the initial 24 h of NIV (day 1, timepoint sp0-1) only in patients receiving a low MP_WAL, but not in those receiving a high MP_WAL (Table 1, Fig. 6C–P). The differences between low and high MP_WAL group observed at day 1 were magnified during the initial 7 days of NIV (Table 1, Fig. 6).

Impact of early (day 1) MP_WAL on physio-anatomical responses

We next explored potential mechanisms underlying the association of MP_WAL(day 1) with the observed clinical outcomes.

In the whole study population, early MP_WAL(day 1) correlated significantly with O2 response, CO2 response ultrasonographic responses and C-reactive protein changes at resupination (timepoint sp1) in univariate analysis (Fig. 7). The observed associations were confirmed by multivariable analysis, where MP_WAL(day 1) independently predicted ∆VR_sp0-1(β = 0.43, p = 0.011), ∆global LUS_sp0-1(β = 0.41, p = 0.018), global reaeration score_sp1(β = -0.49, p = 0.009), ∆WAL_sp0-1(β = 0.41, p = 0.018) and ∆C-reactive protein_sp0-1(β = 0.38, p = 0.029) (Table 4).

Fig. 7 — Dose–response relationship between Mechanical Power normalized to well-aerated lung volume during the 1^st 24 h of NIV [MP _WAL(day 1)] and physio-anatomical responses after the 1^st 24 h of NIV(∆sp0-1) in the whole study population (n = 216). PP patients are represented by black full dots, supine patients are represented by empty triangles. Green area denotes a positive response, red area a negative one

Table 4.

Multivariable linear regression analysis of predictors of CO2-response (by ventilatory ratio, VR), ultrasonographic response and C-reactive protein responses in the whole study population(n = 216)

O2-response (∆paO2_sp0-1)
Parameter	β	SE	P	VIF	R²	R²_adj
MP_WAL (day 1)	0.46	0.006	0.009	1.51	0.66	0.62
Duration of 1^st PP session (hours)	0.42	0.010	0.019	1.49	0.66	0.62
Not included in the model
paO2/FiO2_sp0	–0.29	0.312	0.312	1.47
PEEP (day 1)	0.18	0.192	0.287	1.31
VTe (day 1	–0.27	0.216	0.312	1.21
RR (day 1)	–0.15	0.189	0.193	1.77
CO2-response (∆VR_sp0-1)
Parameter	β	SE	P	VIF	R²	R²_adj
MP_WAL(day 1)	0.44	0.011	0.010	1.59	0.69	0.61
Duration of 1^st PP session (hours)	–0.41	0.008	0.028	1.45	0.69	0.61
Not included in the model
VR_sp0	–0.31	0.128	0.398	1.18
PEEP (day 1)	–0.13	0.139	0.151	1.01
RR(day 1)	0.29	0.146
Global LUS response: ∆ global LUS_sp0-1
Parameter	β	SE	P	VIF	R²	R² _adj
MP_WAL (day 1)	0.41	0.009	0.018	0.06	0.68	0.63
global LUS_sp0	-0.37	0002	0.009	1.89	0.68	0.63
Not included in the model
RR (day 1)	-0.19	0.318	0.573	1.47
PEEP (day 1)	-0.13	0.128	0.398	0.47
Global reaeration score_sp1
Parameter	β	SE	P	VIF	R²	R² _adj
Duration of 1^st PP session (hours)	0.51	0.004	0.012	1.72	0.63	0.60
MP_WAL (day 1)	–0.49	0006	0.009	1.59	0.63	0.60
Not included in the model
global LUS_sp0	0.29	0.612	0.138	1.29	0.65	0.64
PEEP (day 0)	0.16	0.316	0.479	1.13	0.65	0.64
Well-aerated volume response: ∆WAL(%)_sp0-1
Parameter	β	SE	P	VIF	R²	R² _adj
global LUS_sp0	0.51	0.004	0.012	1.72	0.65	0.63
MP_WAL (day 1)	–0.49	0006	0.010	1.59	0.65	0.63
Not included in the model
WAL (day 0)	0.29	0.612	0.138	1.29
PEEP (day 1)	0.16	0.316	0.479	1.13
C-reactive protein response: ∆ CRP_sp0-1
Parameter	β	SE	P	VIF	R²	R² _adj
CRP_sp0	–0.59	0.009	0.011	1.71	0.62	0.57
MP_WAL (day 1)	0.38	0.005	0.029	1.12	0.62	0.57
Not included in the model
Obesity	0.31	0.269	0.513	1.19
PEEP (day 0)	0.28	0.214	0.395	1.12

Open in a new tab

Statistically significant p-values are written in bold characters

Discussion

In the first report exploring MP_WAL in COVID-19 AHRF treated with prone or supine NIV, we found that.

PP was associated with a significant MP_WAL reduction, attributable mainly to an absolute MP reduction and secondly to a relief in the “intensity” of the power delivered to the aerated lung through enhanced lung reaeration (Fig. 2).
MP_WAL delivered during the first 24 h of NIV [MP_WAL(day 1)] outperformed other power measures and ventilatory variables in clinical and pathophysiological outcome prediction.
MP_WAL(day 1) independently predicted different markers of VILI, including gas exchange parameters, ultrasonographic and inflammatory biomarkers, in both supine and proning patients.

Diferent mechanical ventilatory variables have been proposed to affect clinical outcomes and to contribute to VILI, all of which have been separately addressed in experimental or clinical studies [1-3]. The MP represents the energy delivered to the respiratory system by mechanical ventilation, which results from a combination of ventilatory variables. Therefore, MP might have a higher predictive value for patient-centered outcomes, as different combinations of its components may yield a similar damage to the lung [4-8]. The ability to combine the effect of time-varying ventilatory variables seems particularly attractive in spontaneously breathing patients.

Our findings suggest MP delivered early after NIV initiation predicts both NIV failure and mortality. However, the power should be normalized to the size of for the size of well-inflated lung to get an optimal cut-off that could be used in clinical practice, as VILI originates from the interaction between the MP delivered to the respiratory system by the ventilation and the anatomical and pathophysiological characteristics of the latter. Notably, we demonstrated that bedside LUS could be used to accurately estimates WAL size, consistent with data from nonCOVID-19 ARDS [29].

Lastly, we showed that early PP may reduce MP_WAL while improving oxygenation and CO2 clearance during NIV. Therefore, early PP therapy may be a more “energetically” advantageous strategy to relieve hypoxemia than PEEP uptitration, which may overdistend the lung and increase MP and lung injury when it does not recruit atelectatic lung tissue [8, 35]. The association of PEEP levels and NIV duration with COVID-19 pneumonia severity and mortality is consistent with this view [11-13].

Our analysis may have several clinical implications for noninvasive AHRF management.

First, it suggests PP can effectively reduce MP delivery. Second, they suggest early MP_WAL calculation may help define the trade-offs between the benefits and risks of noninvasive ventilatory support, helping to titrate NIV to the least possible energy to achieve oxygenation and enhancing prompt NIV up-grade to invasive mechanical ventilation (IMV) to prevent VILI progression. Both these issues warrant evaluation in future ad hoc trials enrolling different AHRF etiologies.

Strengths and limitations

Strengths of our report are the thorough patient assessment, the hourly ventilatory data collection, which reflect actual power delivery during each ventilatory session more accurately than once daily data, and the separate assessment of supine and prone ventilatory modality. Limitations of our analysis include its post hoc nature and the absence of randomization in the original study. Furthermore, we focused on ventilator-delivered energy and did not assess the energy transferred from respiratory muscles to the lung during inspiration. However, consistent with COVID-19 literature. our patients had no clinical-radiological evidence for increased inspiratory effort or reduced lung compliance [14-17], and Pes swings, an index of inspiratory effort, are unrelated to disease course and unaffected by PP [43], while MPWAL predicted physio-anatomical and clinical outcomes in our patient population.

Conclusions and expert recommendations

Concern has been raised that mechanical ventilation may promote VILI and worsen outcome in spontaneously breathing patients with AHRF. Different ventilatory variables have been proposed to contribute to VILI progression, but no effective predictive, preventive and patient-centered model has been developed [44-48].

Based on our findings we suggest early quantitation of well-aeraled lung volume before NIV initiation in each patient with AHRF candidate to NIV, through CT scan analysis or at least through lung ultrasound, which in our series was at least as accurate for WAL quantitation as free open source software.

After NIV initiation, the value of MP delivered to the patient by NIV could be estimated from routine ventilatory parameters using available formulas and adjusted for individual WAL volume to estimate MP_WAL: if patient’s MP_WAL exceeds suggested safey threshold (9.1 J/min/L), VILI could be prevented by implementing PP, which reduced MP_WAL by an average of 34% in our series.

If PP fails to reduce MP_WAL below safety threshold, NIV upgrade to invasive ventilation should be considered to achieve complete ventilatory control and reduce MP_WAL and the risk of VILI.

The added value of this approach in noninvasively ventilated AHRF patients is to provide an accurate tool for early prediction of the risk of VILI and adverse outcomes(i.e., MP_WAL at day 1), an effective mean to prevent or reduce the risk of VILI (i.e., prone position adoption). Furthermore, this approach should be personalized, i.e. the MP delivered by the ventilator to patient’s respiratory system should be adjusted for individual WAL size, which varies widely across AHRF patients and could be calculated by lung imaging techniques.

This integrated approach, which showed accuracy in our study, needs prospective validation in independent cohorts of patients with AHRF of diverse etiologies.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary file1 (DOC 102 KB)^{(101.5KB, doc)}

Supplementary file2 (DOCX 1866 KB)^{(1.8MB, docx)}

Abbreviations

ABG: Arterial blood gas
AHRF: Acute hypoxemic respiratory failure
CPAP: Continuous positive airway pressure
DSI: Dead space index
ETI: Endotracheal intubation
ISARIC: International Severe Acute Respiratory Infection Consortium
LUS: Lung ultrasound score
MV: Minute ventilation
MVcorr: Corrected MV
NAL: Non-aerated lung volume
NIV: Noninvasive ventilation
NRS: Numeric rating scale
PBW: Predicted body weight
PEEP: Positive end-expiratory pressure
PP: Prone position
PRO-NIV: Prone position in NonInvasive Ventilation
Ps: Pressure support
PS: Propensity score
PSV: Pressure support ventilation
RR: Respiratory rate
SAPS: Simplified acute physiology score
MP: Mechanical power
MP_WAL: Mechanical power normalized to well-aerated lung
PAL: Poorly aerated lung volume
SMD: Standardized mean difference
VILI: Ventilation-induced lung injury
VR: Ventilatory ratio
VTe: Expiratory tidal volume
WAL: Well-aerated lung volume
WOB: Work of breathing

Appendix

PPPM Innovation highlights

Working hypothesis in the framework of PPPM

Concern has been raised that mechanical ventilation may promote lung injury (so called ventilation-induced lung injury, VILI) and worsen outcome in spontaneously breathing patients with acute hypoxemic respiratory failure (AHRF).

Different ventilatory variables have been proposed to contribute to VILI progression, but data in noninvasively ventilated patients are scarce, inconsistent and inconclusive and no valuable outcome predictor currently exists.

Mechanical power (MP), a measure of the energy transfer rate from the ventilator to the respiratory system, predicted mortality in invasively ventilated ARDS patients, but its role in noninvasively ventilated AHRF is unexplored.

Innovation towards the

predictive approach,

Mechanical Power normalized to the volume of well-aerated lung (MPWAL) delivered during the initial 24 h of noninvasive ventilation (NIV) independently predicts NIV failure intubation and death and affects physio-anatomical markers of VILI (i.e., gas exchange, lung reaeration by lung ultrasound, and inflammatory biomarkers), outperforming all other ventilatory variables in outcome prediction.

4.
targeted prevention

We found prone position during noninvasive ventilation is an effective tool to reduce mechanical power delivered during NIV.

5.
personalization of medical services

Bedside MPWAL calculation during the 1^st day of NIV may guide NIV delivery, help define the tradeoffs between the benefits of noninvasive support and the risks of VILI, and enhance prompt escalation to invasive mechanical ventilation to minimize lung injury progression.

How does the presented innovation go beyond

Early bedside MP_WAL calculation is a useful parameter that can predict NIV outcome, suggest the need to adopt prone position to reduce MP delivery during NIV, and help decide when to stop NIV and upgrade treatment to invasive ventilation to more effectively preventi VILI progression.

Authors’ contributions

Giovanni Musso: conceived and designed the study, analyzed the results, drafted the article, gave final approval.

Claudio Taliano: acquired data, critically analyzed the results, contributed to the draft of the article, gave final approval.

Marco De Iuliis: acquired data, critically analyzed the results, contributed to the draft of the article, gave final approval.

Elena Paschetta: acquired data, critically analyzed the results, contributed to the draft of the article, gave final approval.

Caterina Fonti: acquired data, critically analyzed the results, contributed to the draft of the article, gave final approval.

Andrea Ferraris: acquired data, critically analyzed the results, contributed to the draft of the article, gave final approval.

Marta Druetta: acquired data, critically analyzed the results, contributed to the draft of the article, gave final approval.

Ines Sarah Vianou: acquired data, critically analyzed the results, contributed to the draft of the article, gave final approval.

Francesca Ranghino: acquired data, critically analyzed the results, contributed to the draft of the article, gave final approval.

Federica Riedo: acquired data, critically analyzed the results, contributed to the draft of the article, gave final approval.

Davide Deangelis: acquired data, critically analyzed the results, contributed to the draft of the article, gave final approval.

Gloria Tirabassi: acquired data, critically analyzed the results, contributed to the draft of the article, gave final approval.

Data availability

Individual, de-identified participant data will be shared upon request to the corresponding author by email till up to 2 years after publication of the article.

Declarations

Ethics approval

Ethics approval was granted by the Comitato Etico Interaziendale A.O.U. Città della Salute e della Scienza di Torino (prot. N. 0046392) on Dec 15th, 2020 as an ancillary study to the PRO-NIV study (ISRCTN23016116) and this study was performed in accordance with Helsinki declaration. All patients gave informed consent to participate.

Consent for publication

All authors gave their consent for publication.

Competing interests

no author has any present or past conflict of interest or financial and/or non-financial competing interest to disclose

Footnotes

Publisher's note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1.Gattinoni L, Marini JJ, Collino F, Maiolo G, Rapetti F, Tonetti T, et al. The future of mechanical ventilation: lessons from the present and the past. Crit Care. 2017;21:183. doi: 10.1186/s13054-017-1750-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Frat JP, Ragot S, Coudroy R, Constantin JM, Girault C, Prat G, et al. Predictors of intubation in patients with acute hypoxemic respiratory failure treated with a noninvasive oxygenation strategy. Crit Care Med. 2018;46:208–215. doi: 10.1097/CCM.0000000000002818. [DOI] [PubMed] [Google Scholar]
3.Carteaux G, Millan-Guilarte T, De Prost N, Razazi K, Abid S, Thille AW, Schortgen F, et al. Failure of noninvasive ventilation for de novo acute hypoxemic respiratory failure: role of tidal volume. Crit Care Med. 2016;44:282–290. doi: 10.1097/CCM.0000000000001379. [DOI] [PubMed] [Google Scholar]
4.Tonelli R, Fantini R, Tabbì L, Castaniere I, Pisani L, Pellegrino MR, et al. Early inspiratory effort assessment by esophageal manometry predicts noninvasive ventilation outcome in de novo respiratory failure. A pilot study. Am J Respir Crit Care Med. 2020;202:558–567. doi: 10.1164/rccm.201912-2512OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Golubnitschaja O, Baban B, Boniolo G, Wang W, Bubnov R, Kapalla M, et al. Medicine in the early twenty-first century: paradigm and anticipation - EPMA position paper 2016. EPMA J. 2016;7:23. doi: 10.1186/s13167-016-0072-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Golubnitschaja O, Costigliola V. Predictive, preventive and personalised medicine as the medicine of the future: Anticipatory scientific innovation and advanced medical services. In: Nadin M (ed). Anticipation and medicine. Cham, Springer, Switzerland; 2017. pp. 69–85.
7.Tonetti T, Cressoni M, Collino F, Maiolo G, Rapetti F, Quintel M, Gattinoni L. Volutrauma, atelectrauma, and mechanical power. Critical Care Med. 2017;45:e327–e328. doi: 10.1097/CCM.0000000000002193. [DOI] [PubMed] [Google Scholar]
8.Xie Y, Zheng H, Mou Z, Wang Y, Li X. High Expression of CXCL10/CXCR3 in Ventilator-Induced Lung Injury Caused by High Mechanical Power. Biomed Res Int. 2022;2022:6803154. doi: 10.1155/2022/6803154. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Xie Y, Wang Y, Liu K, Li X. Correlation analysis between mechanical power, transforming growth factor-(1, and connective tissue growth factor levels in acute respiratory distress syndrome patients and their clinical significance in pulmonary structural remodeling. Medicine (Baltimore) 2019;98:e16531. doi: 10.1097/MD.0000000000016531. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Guérin C, Papazian L, Reignier J, Ayzac L, Loundou A, Forel JM, et al. Effect of driving pressure on mortality in ARDS patients during lung protective mechanical ventilation in two randomized controlled trials. Crit Care. 2016;20:384. doi: 10.1186/s13054-016-1556-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Serpa Neto A, Deliberato RO, Johnson A, Bos LD, Amorim P, Pereira SM, et al. Mechanical power of ventilation is associated with mortality in critically ill patients: an analysis of patients in two observational cohorts. Intensive Care Med. 2018;44:1914–1922. doi: 10.1007/s00134-018-5375-6. [DOI] [PubMed] [Google Scholar]
12.Coppola S, Caccioppola A, Froio S, Formenti P, De Giorgis V, Galanti V, et al. Effect of mechanical power on intensive care mortality in ARDS patients. Crit Care. 2020;24:246. doi: 10.1186/s13054-020-02963-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Zhu Y, Peng W, Zhen S, Jiang X. Mechanical power normalized to predicted body weight is associated with mortality in critically ill patients: a cohort study. BMC Anesthesiol. 2021;21:278. doi: 10.1186/s12871-021-01497-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Wu MA, Lopez G, Nebuloni M, Ottolina D, Montomoli J, Carsana L, et al. Lung histopathologic clusters in severe COVID-19: a link between clinical picture and tissue damage. Crit Care. 2021;25:423. doi: 10.1186/s13054-021-03846-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.COVID-ICU group, for the REVA network, COVID-ICU investigators. Benefits and risks of noninvasive oxygenation strategy in COVID-19: a multicenter, prospective cohort study (COVID-ICU) in 137 hospitals. Crit Care. 2021;25:421. [DOI] [PMC free article] [PubMed]
16.Le Terrier C, Sigaud F, Lebbah S, et al. Early prone positioning in acute respiratory distress syndrome related to COVID-19: a propensity score analysis from the multicentric cohort COVIDICU network-the ProneCOVID study. Crit Care. 2022;26:71. doi: 10.1186/s13054-022-03949-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Tonelli R, Busani S, Tabbì L, Fantini R, Castaniere I, Biagioni E, et al. Inspiratory Effort and Lung Mechanics in Spontaneously Breathing Patients with Acute Respiratory Failure due to COVID-19: A Matched Control Study. Am J Respir Crit Care Med. 2021;204:725–728. doi: 10.1164/rccm.202104-1029LE. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Alhazzani W, Møller MH, Arabi YM, Loeb M, Gong MN, Fan E, et al. Surviving sepsis campaign: guidelines on the management of critically Ill adults with coronavirus disease 2019 (COVID-19) Crit Care Med. 2020;48:e440–e469. doi: 10.1097/CCM.0000000000004363. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Chiumello D, Chiodaroli E, Coppola S, CappioBorlino S, Granata C, Pitimada M, Wendel Garcia PD. Awake prone position reduces work of breathing in patients with COVID-19 ARDS supported by CPAP. Ann Intensive Care. 2021;11:179. doi: 10.1186/s13613-021-00967-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Coppola S, Chiumello D, Busana M, Giola E, Palermo P, Pozzi T, et al. Role of total lung stress on the progression of early COVID-19 pneumonia. Intensive Care Med. 2021;47:1130–1139. doi: 10.1007/s00134-021-06519-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Rosén J, von Oelreich E, Fors D, JonssonFagerlund M, Taxbro K, Skorup P, 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. doi: 10.1186/s13054-021-03602-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Kaur R, Vines DL, Mirza S, Elshafei A, Jackson JA, Harnois LJ, et al. Early versus late awake prone positioning in non-intubated patients with COVID-19. Crit Care. 2021;25:340. doi: 10.1186/s13054-021-03761-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Musso G, Taliano C, Molinaro F, Fonti C, Veliaj D, Torti D, et al. Early prolonged prone position in noninvasively ventilated patients with SARS-CoV-2-related moderate-to-severe hypoxemic respiratory failure: clinical outcomes and mechanisms for treatment response in the PRO-NIV study. Crit Care. 2022;26:118. doi: 10.1186/s13054-022-03937-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Chiumello D, Gotti M, Guanziroli M, Formenti P, Umbrello M, Pasticci I, et al. Bedside calculation of mechanical power during volume- and pressure-controlled mechanical ventilation. Crit Care. 2020;24:417. doi: 10.1186/s13054-020-03116-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Pelosi P, Tonelli R, Torregiani C, Baratella E, Confalonieri M, Battaglini D, et al. Different Methods to Improve the Monitoring of Noninvasive Respiratory Support of Patients with Severe Pneumonia/ARDS Due to COVID-19: An Update. J Clin Med. 2022;11:1704. doi: 10.3390/jcm11061704. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Musso G, Taliano C, Paschetta E, De Iuliis M, Fonti C, Vianou IS, et al. Mechanical power delivered by noninvasive ventilation contributes to physio-anatomical and clinical responses to early vs. late proning in COVID-19 pneumonia. Critical Care Med. 2023;in press. 10.1097/CCM.0000000000005915 [DOI] [PubMed]
27.Coopersmith CM, Antonelli M, Bauer SR, et al. The Surviving Sepsis Campaign: Research Priorities for Coronavirus Disease 2019 in Critical Illness. Crit Care Med. 2021;49:598–622. doi: 10.1097/CCM.0000000000004895. [DOI] [PubMed] [Google Scholar]
28.Salaffi F, Carotti M, Tardella M, Borgheresi A, Agostini A, Minorati D, et al. The role of a chest computed tomography severity score in coronavirus disease 2019 pneumonia. Medicine (Baltimore) 2020;99:e22433. doi: 10.1097/MD.0000000000022433. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Colombi D, Bodini FC, Petrini M, Maffi G, Morelli N, Milanese G, et al. Well-aerated lung on admitting chest CT to predict adverse outcome in COVID-19 pneumonia. Radiology. 2020;296:E86–E96. doi: 10.1148/radiol.2020201433. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Arbelot C, Neto FLD, Gao Y. Lung ultrasound in emergency and critically Ill patients number of supervised exams to reach basic competence. Anesthesiology. 2020;132:899–907. doi: 10.1097/ALN.0000000000003096. [DOI] [PubMed] [Google Scholar]
31.Steinberg I, Pasticci I, Busana M, et al. Lung ultrasound and electrical impedance tomography during ventilator-induced lung injury. Crit Care Med. 2022;50:e630–e637. doi: 10.1097/CCM.0000000000005479. [DOI] [PubMed] [Google Scholar]
32.Chiumello D, Mongodi S, Algieri I, Vergani GL, Orlando A, Via G, et al. Assessment of lung aeration and recruitment by CT scan and ultrasound in acute respiratory distress syndrome patients. Crit Care Med. 2018;46:1761–1768. doi: 10.1097/CCM.0000000000003340. [DOI] [PubMed] [Google Scholar]
33.Rousset D, Sarton B, Riu B, Bataille B, Silva S; of the PLUS study group. Bedside ultrasound monitoring of prone position induced lung inflation. Intensive Care Med. 2021;47:626–628 [DOI] [PMC free article] [PubMed]
34.Soummer A, Perbet S, Brisson H, Arbelot C, Constantin JM, Lu Q, et al; Lung Ultrasound Study Group. Ultrasound assessment of lung aeration loss during a successful weaning trial predicts postextubation distress*. Crit Care Med. 2012;40:2064–72. [DOI] [PubMed]
35.Zhao X, Wang K, Zuo P, et al. Early decrease in blood platelet count is associated with poor prognosis in COVID-19 patients—indications for predictive, preventive, and personalized medical approach. EPMA J. 2020;11:139–145. doi: 10.1007/s13167-020-00208-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Matthews JN, Altman DG, Campbell MJ, Royston P. Analysis of serial measurements in medical research. BMJ. 1990;300:230–235. doi: 10.1136/bmj.300.6719.230. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Long JS Regression Models for categorical and limited dependent variables. Sage Publications. 1997, Thousand Oaks, CA, US.
38.Craney TA, Surles JG, Values M-D. Qual Eng. 2002;14(3):391–403. doi: 10.1081/QEN-120001878. [DOI] [Google Scholar]
39.Apigo M, Schechtman J, Dhliwayo N, Al Tameemi M, Gazmuri RJ. Development of a work of breathing scale and monitoring need of intubation in COVID-19 pneumonia. Crit Care. 2020;24:477. doi: 10.1186/s13054-020-03176-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Camporota L, Sanderson B, Chiumello D, Terzi N, Argaud L, Rimmelé T, et al. Prone position in COVID-19 and -COVID-19 acute respiratory distress syndrome: An international multicenter observational comparative study. Crit Care Med. 2022;1(50):633–643. doi: 10.1097/CCM.0000000000005354. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Fossali T, Pavlovsky B, Ottolina D, Colombo R, Basile MC, Castelli A, et al. Effects of prone position on lung recruitment and ventilation-perfusion matching in patients with COVID-19 acute respiratory distress syndrome: A combined CT scan/electrical impedance tomography study. Crit Care Med. 2022;50:723–732. doi: 10.1097/CCM.0000000000005450. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Cardinale M, Boussen S, Cungi PJ, Esnault P, Mathais Q, Bordes J, et al. Lung-dependent areas collapse, monitored by electrical impedance tomography, may predict the oxygenation response to prone ventilation in COVID-19 acute respiratory distress syndrome. Crit Care Med. 2022;50:1093–1102. doi: 10.1097/CCM.0000000000005487. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Terzi N, Bayat S, Noury N, Turbil E, Habre W, Argaud L, et al. Comparison of pleural and esophageal pressure in supine and prone positions in a porcine model of acute respiratory distress syndrome. J Appl Physiol. 2020;128:1617–1625. doi: 10.1152/japplphysiol.00251.2020. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Chaari L, Golubnitschaja O. Covid-19 pandemic by the “real-time” monitoring: the Tunisian case and lessons for global epidemics in the context of 3PM strategies. EPMA J. 2020;11:133–138. doi: 10.1007/s13167-020-00207-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Skladany L, Koller T, AdamcovaSelcanova S, et al. Challenging management of severe chronic disorders in acute pandemic situation: Chronic liver disease under COVID-19 pandemic as the proof-of-principle model to orchestrate the measures in 3PM context. EPMA J. 2021;12:1–14. doi: 10.1007/s13167-021-00231-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Radanliev P, De Roure D, Walton R, et al. COVID-19 what have we learned? The rise of social machines and connected devices in pandemic management following the concepts of predictive, preventive and personalized medicine. EPMA J. 2020;11:311–332. doi: 10.1007/s13167-020-00218-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Saffaran S, Das A, Laffey JG, Hardman JG, Yehya N, Bates DG. Utility of driving pressure and mechanical power to guide protective ventilator settings in two cohorts of adult and pediatric patients with acute respiratory distress syndrome: A computational investigation. Crit Care Med. 2020;48:1001–1008. doi: 10.1097/CCM.0000000000004372. [DOI] [PubMed] [Google Scholar]
48.Marini JJ, Gattinoni L. Time course of evolving ventilator-induced lung injury: the "shrinking baby lung". Crit Care Med. 2020;48:1203–1209. doi: 10.1097/CCM.0000000000004416. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary file1 (DOC 102 KB)^{(101.5KB, doc)}

Supplementary file2 (DOCX 1866 KB)^{(1.8MB, docx)}

Data Availability Statement

Individual, de-identified participant data will be shared upon request to the corresponding author by email till up to 2 years after publication of the article.

[CR1] 1.Gattinoni L, Marini JJ, Collino F, Maiolo G, Rapetti F, Tonetti T, et al. The future of mechanical ventilation: lessons from the present and the past. Crit Care. 2017;21:183. doi: 10.1186/s13054-017-1750-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR2] 2.Frat JP, Ragot S, Coudroy R, Constantin JM, Girault C, Prat G, et al. Predictors of intubation in patients with acute hypoxemic respiratory failure treated with a noninvasive oxygenation strategy. Crit Care Med. 2018;46:208–215. doi: 10.1097/CCM.0000000000002818. [DOI] [PubMed] [Google Scholar]

[CR3] 3.Carteaux G, Millan-Guilarte T, De Prost N, Razazi K, Abid S, Thille AW, Schortgen F, et al. Failure of noninvasive ventilation for de novo acute hypoxemic respiratory failure: role of tidal volume. Crit Care Med. 2016;44:282–290. doi: 10.1097/CCM.0000000000001379. [DOI] [PubMed] [Google Scholar]

[CR4] 4.Tonelli R, Fantini R, Tabbì L, Castaniere I, Pisani L, Pellegrino MR, et al. Early inspiratory effort assessment by esophageal manometry predicts noninvasive ventilation outcome in de novo respiratory failure. A pilot study. Am J Respir Crit Care Med. 2020;202:558–567. doi: 10.1164/rccm.201912-2512OC. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR5] 5.Golubnitschaja O, Baban B, Boniolo G, Wang W, Bubnov R, Kapalla M, et al. Medicine in the early twenty-first century: paradigm and anticipation - EPMA position paper 2016. EPMA J. 2016;7:23. doi: 10.1186/s13167-016-0072-4. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR6] 6.Golubnitschaja O, Costigliola V. Predictive, preventive and personalised medicine as the medicine of the future: Anticipatory scientific innovation and advanced medical services. In: Nadin M (ed). Anticipation and medicine. Cham, Springer, Switzerland; 2017. pp. 69–85.

[CR7] 7.Tonetti T, Cressoni M, Collino F, Maiolo G, Rapetti F, Quintel M, Gattinoni L. Volutrauma, atelectrauma, and mechanical power. Critical Care Med. 2017;45:e327–e328. doi: 10.1097/CCM.0000000000002193. [DOI] [PubMed] [Google Scholar]

[CR8] 8.Xie Y, Zheng H, Mou Z, Wang Y, Li X. High Expression of CXCL10/CXCR3 in Ventilator-Induced Lung Injury Caused by High Mechanical Power. Biomed Res Int. 2022;2022:6803154. doi: 10.1155/2022/6803154. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR9] 9.Xie Y, Wang Y, Liu K, Li X. Correlation analysis between mechanical power, transforming growth factor-(1, and connective tissue growth factor levels in acute respiratory distress syndrome patients and their clinical significance in pulmonary structural remodeling. Medicine (Baltimore) 2019;98:e16531. doi: 10.1097/MD.0000000000016531. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR10] 10.Guérin C, Papazian L, Reignier J, Ayzac L, Loundou A, Forel JM, et al. Effect of driving pressure on mortality in ARDS patients during lung protective mechanical ventilation in two randomized controlled trials. Crit Care. 2016;20:384. doi: 10.1186/s13054-016-1556-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR11] 11.Serpa Neto A, Deliberato RO, Johnson A, Bos LD, Amorim P, Pereira SM, et al. Mechanical power of ventilation is associated with mortality in critically ill patients: an analysis of patients in two observational cohorts. Intensive Care Med. 2018;44:1914–1922. doi: 10.1007/s00134-018-5375-6. [DOI] [PubMed] [Google Scholar]

[CR12] 12.Coppola S, Caccioppola A, Froio S, Formenti P, De Giorgis V, Galanti V, et al. Effect of mechanical power on intensive care mortality in ARDS patients. Crit Care. 2020;24:246. doi: 10.1186/s13054-020-02963-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR13] 13.Zhu Y, Peng W, Zhen S, Jiang X. Mechanical power normalized to predicted body weight is associated with mortality in critically ill patients: a cohort study. BMC Anesthesiol. 2021;21:278. doi: 10.1186/s12871-021-01497-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR14] 14.Wu MA, Lopez G, Nebuloni M, Ottolina D, Montomoli J, Carsana L, et al. Lung histopathologic clusters in severe COVID-19: a link between clinical picture and tissue damage. Crit Care. 2021;25:423. doi: 10.1186/s13054-021-03846-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR15] 15.COVID-ICU group, for the REVA network, COVID-ICU investigators. Benefits and risks of noninvasive oxygenation strategy in COVID-19: a multicenter, prospective cohort study (COVID-ICU) in 137 hospitals. Crit Care. 2021;25:421. [DOI] [PMC free article] [PubMed]

[CR16] 16.Le Terrier C, Sigaud F, Lebbah S, et al. Early prone positioning in acute respiratory distress syndrome related to COVID-19: a propensity score analysis from the multicentric cohort COVIDICU network-the ProneCOVID study. Crit Care. 2022;26:71. doi: 10.1186/s13054-022-03949-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR17] 17.Tonelli R, Busani S, Tabbì L, Fantini R, Castaniere I, Biagioni E, et al. Inspiratory Effort and Lung Mechanics in Spontaneously Breathing Patients with Acute Respiratory Failure due to COVID-19: A Matched Control Study. Am J Respir Crit Care Med. 2021;204:725–728. doi: 10.1164/rccm.202104-1029LE. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR18] 18.Alhazzani W, Møller MH, Arabi YM, Loeb M, Gong MN, Fan E, et al. Surviving sepsis campaign: guidelines on the management of critically Ill adults with coronavirus disease 2019 (COVID-19) Crit Care Med. 2020;48:e440–e469. doi: 10.1097/CCM.0000000000004363. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR19] 19.Chiumello D, Chiodaroli E, Coppola S, CappioBorlino S, Granata C, Pitimada M, Wendel Garcia PD. Awake prone position reduces work of breathing in patients with COVID-19 ARDS supported by CPAP. Ann Intensive Care. 2021;11:179. doi: 10.1186/s13613-021-00967-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Coppola S, Chiumello D, Busana M, Giola E, Palermo P, Pozzi T, et al. Role of total lung stress on the progression of early COVID-19 pneumonia. Intensive Care Med. 2021;47:1130–1139. doi: 10.1007/s00134-021-06519-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Rosén J, von Oelreich E, Fors D, JonssonFagerlund M, Taxbro K, Skorup P, 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. doi: 10.1186/s13054-021-03602-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Kaur R, Vines DL, Mirza S, Elshafei A, Jackson JA, Harnois LJ, et al. Early versus late awake prone positioning in non-intubated patients with COVID-19. Crit Care. 2021;25:340. doi: 10.1186/s13054-021-03761-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR23] 23.Musso G, Taliano C, Molinaro F, Fonti C, Veliaj D, Torti D, et al. Early prolonged prone position in noninvasively ventilated patients with SARS-CoV-2-related moderate-to-severe hypoxemic respiratory failure: clinical outcomes and mechanisms for treatment response in the PRO-NIV study. Crit Care. 2022;26:118. doi: 10.1186/s13054-022-03937-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR24] 24.Chiumello D, Gotti M, Guanziroli M, Formenti P, Umbrello M, Pasticci I, et al. Bedside calculation of mechanical power during volume- and pressure-controlled mechanical ventilation. Crit Care. 2020;24:417. doi: 10.1186/s13054-020-03116-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Pelosi P, Tonelli R, Torregiani C, Baratella E, Confalonieri M, Battaglini D, et al. Different Methods to Improve the Monitoring of Noninvasive Respiratory Support of Patients with Severe Pneumonia/ARDS Due to COVID-19: An Update. J Clin Med. 2022;11:1704. doi: 10.3390/jcm11061704. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR26] 26.Musso G, Taliano C, Paschetta E, De Iuliis M, Fonti C, Vianou IS, et al. Mechanical power delivered by noninvasive ventilation contributes to physio-anatomical and clinical responses to early vs. late proning in COVID-19 pneumonia. Critical Care Med. 2023;in press. 10.1097/CCM.0000000000005915 [DOI] [PubMed]

[CR27] 27.Coopersmith CM, Antonelli M, Bauer SR, et al. The Surviving Sepsis Campaign: Research Priorities for Coronavirus Disease 2019 in Critical Illness. Crit Care Med. 2021;49:598–622. doi: 10.1097/CCM.0000000000004895. [DOI] [PubMed] [Google Scholar]

[CR28] 28.Salaffi F, Carotti M, Tardella M, Borgheresi A, Agostini A, Minorati D, et al. The role of a chest computed tomography severity score in coronavirus disease 2019 pneumonia. Medicine (Baltimore) 2020;99:e22433. doi: 10.1097/MD.0000000000022433. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR29] 29.Colombi D, Bodini FC, Petrini M, Maffi G, Morelli N, Milanese G, et al. Well-aerated lung on admitting chest CT to predict adverse outcome in COVID-19 pneumonia. Radiology. 2020;296:E86–E96. doi: 10.1148/radiol.2020201433. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR30] 30.Arbelot C, Neto FLD, Gao Y. Lung ultrasound in emergency and critically Ill patients number of supervised exams to reach basic competence. Anesthesiology. 2020;132:899–907. doi: 10.1097/ALN.0000000000003096. [DOI] [PubMed] [Google Scholar]

[CR31] 31.Steinberg I, Pasticci I, Busana M, et al. Lung ultrasound and electrical impedance tomography during ventilator-induced lung injury. Crit Care Med. 2022;50:e630–e637. doi: 10.1097/CCM.0000000000005479. [DOI] [PubMed] [Google Scholar]

[CR32] 32.Chiumello D, Mongodi S, Algieri I, Vergani GL, Orlando A, Via G, et al. Assessment of lung aeration and recruitment by CT scan and ultrasound in acute respiratory distress syndrome patients. Crit Care Med. 2018;46:1761–1768. doi: 10.1097/CCM.0000000000003340. [DOI] [PubMed] [Google Scholar]

[CR33] 33.Rousset D, Sarton B, Riu B, Bataille B, Silva S; of the PLUS study group. Bedside ultrasound monitoring of prone position induced lung inflation. Intensive Care Med. 2021;47:626–628 [DOI] [PMC free article] [PubMed]

[CR34] 34.Soummer A, Perbet S, Brisson H, Arbelot C, Constantin JM, Lu Q, et al; Lung Ultrasound Study Group. Ultrasound assessment of lung aeration loss during a successful weaning trial predicts postextubation distress*. Crit Care Med. 2012;40:2064–72. [DOI] [PubMed]

[CR35] 35.Zhao X, Wang K, Zuo P, et al. Early decrease in blood platelet count is associated with poor prognosis in COVID-19 patients—indications for predictive, preventive, and personalized medical approach. EPMA J. 2020;11:139–145. doi: 10.1007/s13167-020-00208-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR36] 36.Matthews JN, Altman DG, Campbell MJ, Royston P. Analysis of serial measurements in medical research. BMJ. 1990;300:230–235. doi: 10.1136/bmj.300.6719.230. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR37] 37.Long JS Regression Models for categorical and limited dependent variables. Sage Publications. 1997, Thousand Oaks, CA, US.

[CR38] 38.Craney TA, Surles JG, Values M-D. Qual Eng. 2002;14(3):391–403. doi: 10.1081/QEN-120001878. [DOI] [Google Scholar]

[CR39] 39.Apigo M, Schechtman J, Dhliwayo N, Al Tameemi M, Gazmuri RJ. Development of a work of breathing scale and monitoring need of intubation in COVID-19 pneumonia. Crit Care. 2020;24:477. doi: 10.1186/s13054-020-03176-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR40] 40.Camporota L, Sanderson B, Chiumello D, Terzi N, Argaud L, Rimmelé T, et al. Prone position in COVID-19 and -COVID-19 acute respiratory distress syndrome: An international multicenter observational comparative study. Crit Care Med. 2022;1(50):633–643. doi: 10.1097/CCM.0000000000005354. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR41] 41.Fossali T, Pavlovsky B, Ottolina D, Colombo R, Basile MC, Castelli A, et al. Effects of prone position on lung recruitment and ventilation-perfusion matching in patients with COVID-19 acute respiratory distress syndrome: A combined CT scan/electrical impedance tomography study. Crit Care Med. 2022;50:723–732. doi: 10.1097/CCM.0000000000005450. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Cardinale M, Boussen S, Cungi PJ, Esnault P, Mathais Q, Bordes J, et al. Lung-dependent areas collapse, monitored by electrical impedance tomography, may predict the oxygenation response to prone ventilation in COVID-19 acute respiratory distress syndrome. Crit Care Med. 2022;50:1093–1102. doi: 10.1097/CCM.0000000000005487. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR43] 43.Terzi N, Bayat S, Noury N, Turbil E, Habre W, Argaud L, et al. Comparison of pleural and esophageal pressure in supine and prone positions in a porcine model of acute respiratory distress syndrome. J Appl Physiol. 2020;128:1617–1625. doi: 10.1152/japplphysiol.00251.2020. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Chaari L, Golubnitschaja O. Covid-19 pandemic by the “real-time” monitoring: the Tunisian case and lessons for global epidemics in the context of 3PM strategies. EPMA J. 2020;11:133–138. doi: 10.1007/s13167-020-00207-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR45] 45.Skladany L, Koller T, AdamcovaSelcanova S, et al. Challenging management of severe chronic disorders in acute pandemic situation: Chronic liver disease under COVID-19 pandemic as the proof-of-principle model to orchestrate the measures in 3PM context. EPMA J. 2021;12:1–14. doi: 10.1007/s13167-021-00231-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR46] 46.Radanliev P, De Roure D, Walton R, et al. COVID-19 what have we learned? The rise of social machines and connected devices in pandemic management following the concepts of predictive, preventive and personalized medicine. EPMA J. 2020;11:311–332. doi: 10.1007/s13167-020-00218-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR47] 47.Saffaran S, Das A, Laffey JG, Hardman JG, Yehya N, Bates DG. Utility of driving pressure and mechanical power to guide protective ventilator settings in two cohorts of adult and pediatric patients with acute respiratory distress syndrome: A computational investigation. Crit Care Med. 2020;48:1001–1008. doi: 10.1097/CCM.0000000000004372. [DOI] [PubMed] [Google Scholar]

[CR48] 48.Marini JJ, Gattinoni L. Time course of evolving ventilator-induced lung injury: the "shrinking baby lung". Crit Care Med. 2020;48:1203–1209. doi: 10.1097/CCM.0000000000004416. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Mechanical power normalized to aerated lung predicts noninvasive ventilation failure and death and contributes to the benefits of proning in COVID-19 hypoxemic respiratory failure

Giovanni Musso

Claudio Taliano

Marco De Iuliis

Elena Paschetta

Caterina Fonti

Andrea Ferraris

Marta Druetta

Ines Sarah Vianou

Francesca Ranghino

Federica Riedo

Davide Deangelis

Gloria Tirabassi

Abstract

Background

Methods

Results

Conclusions

Supplementary Information

Introduction

The unmet need for accurate early predictors of treatment failure and death during noninvasive ventilation (NIV)

Potential contribution of Mechanical Power (MP) to clinical prediction in the framework of PPPM

Working hypothesis

Methods

Measurements

Respiratory parameters

Lung imaging

Definitions and timepoints

Table 1.

Outcomes

Statistical analyses

Sample size calculation and propensity score (PS)-matching of PP and supine group.

Descriptive statistics

Results

Characteristics at Inclusion

LUS lung aeration assessment

Fig. 1.

Treatment-related parameters and outcomes

Effect of PP therapy on MPWAL

Fig. 2.

Relationship of early (day 1) power measures with clinical outcomes

Fig. 3.

Early (day 1) MPWAL is independently associated with clinical outcomes

Fig. 4.

Table 2.

Predictive performance of MPWAL(day 1) for clinical outcomes

Table 3.

Fig. 5.

Time course of ventilatory and other physio-anatomical variables during the initial 7 days of NIV

Fig. 6.

Impact of early (day 1) MPWAL on physio-anatomical responses

Fig. 7.

Table 4.

Discussion

Strengths and limitations

Conclusions and expert recommendations

Electronic supplementary material

Abbreviations

Appendix

PPPM Innovation highlights

Working hypothesis in the framework of PPPM

Innovation towards the

How does the presented innovation go beyond

Authors’ contributions

Data availability

Declarations

Ethics approval

Consent for publication

Competing interests

Footnotes

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Effect of PP therapy on MP_WAL

Early (day 1) MP_WAL is independently associated with clinical outcomes

Predictive performance of MP_WAL(day 1) for clinical outcomes

Impact of early (day 1) MP_WAL on physio-anatomical responses