Potentially modifiable ventilation factors associated with outcome in neurocritical care vs. non-neurocritical care patients: Rational and protocol for a patient-level analysis of PRoVENT, PRoVENT-iMiC and ENIO (PRIME)

Serafini, S.C.; Cinotti, R.; Asehnoune, K.; Battaglini, D.; Robba, C.; Neto, A.S.; Pisani, L.; Mazzinari, G.; Tschernko, E.M.; Schultz, M.J.

doi:10.1016/j.redare.2025.501690

Article information

Abstract

Full Text

Bibliography

Download PDF

Statistics

Figures (2)

Tables (3)

Table 1. Inclusion and exclusion criteria across the 3 original studies.

Table 2. Data collected across the three 3 studies.

Table 3. Outcomes in the 3 original studies.

Show moreShow less

Abstract

Introduction

Ventilator settings and ventilation variables and parameters vary between neurocritical care and non-neurocritical care patients. We aim to compare ventilation management in neurocritical care patients versus non–neurocritical care patients under invasive mechanical ventilation support, and to determine which factors related to ventilatory management have an independent association with outcome in neurocritical patients.

Methods and analysis

We meta-analyze harmonized individual patient data from three observational studies (‘PRactice of VENTilation in critically ill patients without ARDS’ [PRoVENT], ‘PRactice of VENTilation in critically ill patients in Middle–income Countries’ [PRoVENT–iMiC] and ‘Extubation strategies and in neuro–intensive care unit patients and associations with outcomes’ [ENIO]), pooled into a database named ‘PRIME’. The primary endpoint is all cause ICU mortality. Secondary endpoints are key ventilator settings and ventilation variables and parameters. To identify potentially modifiable and non–modifiable factors contributing to ICU mortality, a multivariable model will be built using demographic factors, comorbidities, illness severities, and respiratory and laboratorial variables. In analyses examining the impact of ventilatory variables on outcome, we will estimate the relative risk of ICU mortality for neurocritical and non-neurocritical care patients by dividing the study population based on key ventilator variables and parameters.

Ethics and dissemination

This meta-analysis will address a clinically significant research question by comparing neurocritical care with non–neurocritical care patients. As this is a meta-analysis, additional ethical committee approval is not required. Findings will be disseminated to the scientific community through abstracts and original articles in peer–reviewed journals. Furthermore, the PRIME database will be made accessible for further post–hoc analyses.

Registration

PROVENT, PROVENT–iMiC and ENIO, and the pooled database PRIME are registered at clinicaltrials.gov (NCT01868321 for PRoVENT, NCT03188770 for PRoVENT–iMiC, and NCT03400904 for ENIO, and for PRIME is pending).

Keywords:

ICU

Neurocritical patients

Neurocritical care

Ventilation

Ventilatory support

Modifiable factors

Mortality

ICU mortality

Meta-analysis

Resumen

Introducción

Los ajustes ventilatorios y las variables y parámetros de ventilación varían entre los pacientes neurocríticos y los no neurocríticos. Nuestro objetivo es comparar el manejo ventilatorio en los pacientes neurocríticos y no neurocríticos con soporte mecánico ventilatorio invasivo, así como determinar qué factores relacionados con el manejo ventilatorio tienen una asociación independiente con el resultado en los pacientes neurocríticos.

Métodos y análisis

Metaanalizamos los datos individuales armonizados de pacientes de tres estudios observacionales (‘PRactice of VENTilation in critically ill patients without ARDS’ [PRoVENT], ‘PRactice of VENTilation in critically ill patients in Middle–income Countries’ [PRoVENT–iMiC] y ‘Extubation strategies and in neuro–intensive care unit patients and associations with outcomes’ [ENIO]), agrupados en una base de datos denominada ‘PRIME’. El criterio de valoración primario es la mortalidad en la UCI por cualquier motivo. Los criterios de valoración secundarios son los ajustes de ventilación y las variables y parámetros de ventilación clave. Para identificar los factores potencialmente modificables y no modificables que contribuyen a la mortalidad en la UCI, se construirá un modelo multivariable utilizando factores demográficos, comorbilidades, severidades de la enfermedad y variables respiratorias y de laboratorio. En los análisis que examinan el impacto de las variables ventilatorias en el resultado, calcularemos el riesgo relativo de la mortalidad en la UCI para los pacientes neurocríticos y no neurocríticos, dividiendo la población de estudio sobre la base de las variables y parámetros de ventilación claves.

Ética y divulgación

Este metaanálisis abordará una cuestión de investigación clínicamente significativa comparando los pacientes neurocríticos y no neurocríticos. Al tratarse de un metaanálisis, no se requiere la aprobación adicional del comité de ética. Los hallazgos serán distribuidos a la comunidad científica mediante resúmenes y artículos originales en publicaciones revisadas por pares. Además, se hará accesible la base de datos PRIME para análisis futuros post hoc.

Registro

PROVENT, PROVENT–iMiC y ENIO, así como la base de datos agrupada PRIME están registrados en clinicaltrials.gov (NCT01868321 para PRoVENT, NCT03188770 para PRoVENT–iMiC y NCT03400904 para ENIO, estando pendiente para PRIME).

Palabras clave:

UCI

Pacientes neurocríticos

Cuidados neurocríticos

Ventilación

Soporte ventilatorio

Factores modificables

Mortalidad

Mortalidad en la UCI

Metaanálisis

Full Text

Background

In recent years, there has been a concerted effort to tailor respiratory support to lung conditions in order to mitigate the potential harm caused by positive pressure ventilation.1 This led to the concept of lung–protective ventilation, consisting of ventilation with low volumes and low pressures to mitigate ventilator–induced lung injury (VILI).2 Recent studies have found that driving pressure (ΔP)3,4 and mechanical power (MP) of ventilation is associated with important clinical outcomes,4–6 albeit mostly in patients with acute respiratory distress syndrome (ARDS). Most critically ill patients receiving invasive ventilation, including most neurocritical care patients,7 however, do not have ARDS.

Mechanical ventilation principles in neurocritical care patients differ from those of non–neurocritical care patients in various ways.8–11 For instance, the lungs of neurocritical care patients are seldom injured, resulting in more uniform aeration and generally higher lung compliance.12–15 Additionally, target cumulative fluid balances may vary between neurocritical care and non–neurocritical care patients, resulting in differences in the amount of extravascular lung water.16–18 In neurocritical care patients, the respiratory rate is often more strictly regulated19, and tidal volume are generally higher20 to maintain arterial carbon dioxide levels within brain–protective safety limits.21,22 Moreover, high positive end–expiratory pressure (PEEP) is typically avoided in neurocritical care patients because it can potentially increase intracranial pressure (ICP),12,23 and the oxygen fraction (FiO2) is titrated more carefully, or even liberally, to prevent hypoxemia and hyperoxemia, as both are associated with poor outcomes.24,25 Alone or together, these differences may result in a different intensity in neurocritical care patients.12 Due to these conditions, brain injury patients are often excluded from major studies on lung–protective ventilation. Consequently, the number of studies focusing on differences in ventilation management between neurocritical care and non–neurocritical care patients remains extremely limited.

To address this issue, we harmonized individual patient data from 3 prospective, observational studies: PRactice of VENTilation in critically ill patients without ARDS (PRoVENT),26 PRactice of VENTilation in critically ill patients in Middle–income Countries (PRoVENT–iMiC),27 and Extubation strategies and in neuro-intensive care unit patients and associations with outcomes (ENIO)28 into a pooled dataset named ‘PRIME’. In this article, we outline the structure of this database and present a detailed plan for an analysis that focuses on comparing ventilator setting, ventilation variables and parameters, and potentially modifiable outcome-related factors between neurocritical care and non–neurocritical care patients. We hypothesize that ventilation management in neurocritical care patients differs from that of non–neurocritical care patients. Additionally, we hypothesize that certain, potentially modifiable ventilation factors are independently associated with outcome in neurocritical care patients.

MethodsStudy design

PRIME is a harmonized, pooled dataset of individual data from patients included in 3 prospective observational studies focusing on ventilation management in critically ill patients. The protocols of each study were approved by a central Institutional Review Board, and also by local ethics committees when so required. Written informed consent was obtained whenever required by local legislation. PROVENT, PROVENT–iMiC, ENIO, and the pooled database PRIME are registered at clinicaltrials.gov (NCT01868321 for PRoVENT, NCT03188770 for PRoVENT–iMiC, and NCT03400904 for ENIO. Registration pending for PRIME). The creation of the pooled database, including the harmonization of collected data, did not require additional ethical approval or individual patient informed consent. PRIME was built under the responsibility of the PROtective VEntilation (PROVE) network.29 Each analysis of PRIME will follow the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement.30

Patients

Both PROVENT and PROVENT–iMiC included consecutive patients aged 18 years or older that received invasive ventilation in an ICU. Reasons for intubation were respiratory failure, anaesthesia for surgery, cardiac arrest, depressed consciousness due to traumatic brain injury, subarachnoid haemorrhage, intra–cranial bleeding, stroke, and central nervous system infection. ENIO also included patients aged 18 years or older that received invasive ventilation in an ICU, but limited inclusion to patients with acute brain injury (ABI) with a baseline Glasgow Coma Score (GCS) of 12 or lower before endotracheal intubation. Investigators in PRoVENT and PRoVENT–iMiC enrolled consecutive patients over a fixed period of 4 weeks. Investigators in ENIO screened and included patients for a period of at least 6 months. Both PRoVENT and PRoVENT–iMiC excluded patients in whom ventilation was started before the study recruitment week, patients receiving only non-invasive ventilation, and patients that were transferred to the ICU from another hospital under ventilation. A detailed description of inclusion and exclusion criteria of PROVENT, PROVENT–iMiC and ENIO is provided in Table 1. The data of all patients included in PROVENT, PROVENT–iMiC and ENIO were harmonized and pooled in the PRIME database.

Table 1.

Inclusion and exclusion criteria across the 3 original studies.

Study	Eligibility criteria	Exclusion criteria
PRoVENT	• Aged 18 years or older • Admitted to an ICU • Receiving invasive ventilation, even if initiated outside the hospital, in the emergency room, clinical ward, or operating room, or if it started in the ICU after admission	• Patients in whom ventilation started before the study recruitment week • Patients receiving only NIV • Those transferred from another hospital under ventilation
PRoVENT–iMiC	• Admitted to an ICU • Receiving invasive ventilation, even if initiated outside the hospital, in the emergency room, hospital ward, or operating room	• Patients under 18 years of age • Patients receiving only NIV • Those transferred from another hospital under ventilation
ENIO	• Patients aged 18 years or older • Admitted to the ICU with a baseline GCS of 12 or lower before endotracheal intubation • Required invasive mechanical ventilation for 24 h or more • Underwent an attempt to wean the patient from the ventilator, defined as either an extubation trial or tracheostomy	• Pregnant individuals • Those with spinal cord injuries above T4 • Individuals resuscitated postcardiac arrest • Patients with Guillain–Barré syndrome, motor neuron disease, muscular dystrophy, and myasthenia gravis • Patients who died before extubation • Withdrawal of life-sustaining treatment within 24 h of ICU admission • End-of-life extubation. • Major respiratory co-morbidities (e.g., chronic oxygen at home, chronic obstructive pulmonary disease grade III or IV of the Gold classification) • Major chest trauma ≥ 3 • Patients who underwent tracheostomy prior to ICU admission

Abbreviations: ARDS, Acute Respiratory Distress Syndrome; ENIO, Extubation in neurocritical care patients: the ENIO international prospective study; GCS, Glasgow Coma Scale; PRoVENT Practice of VENTilation in critically ill patients without ARDS study; PRoVENT–iMiC, Practice of VENTilation in critically ill patients in middle income countries study; NIV, Non-Invasive Ventilation.

All patients in the PRIME database were eligible for participation in the initial analysis presented here, except those with ARDS at the start ventilation, as defined by the Berlin criteria.31 No additional exclusion criteria were applied.

Data collected in the original studies

In PROVENT, PROVENT–iMiC and ENIO the following data were collected:

•
Baseline and demographic variables at admission––age, sex, actual body weight, height, body mass index (BMI), type of admission (medical, surgical or trauma), comorbidities including chronic obstructive pulmonary disease (COPD), diabetes mellitus, chronic kidney disease, active cancer, heart failure, and chronic liver failure, and reason for invasive ventilation;
•
Ventilation characteristics––ventilation mode, PEEP, VT, RR, peak pressure (Ppeak) or plateau pressure (Pplat), FiO2, blood gas analysis data when available, arterial pH, partial pressure of arterial blood (PaO2), and partial pressure of carbon dioxide (PaCO2). From the first calendar day, defined as 0, the patients received invasive ventilation in the ICU until day 7 (PRoVENT), and (PRoVENT–iMiC), and at days 1, 3, 7 after ICU admission (ENIO);
•
Outcomes––proportion of patients in the ICU who were at risk of ARDS, (defined using the Lung Injury Prediction Score [LIPS], where patients were stratified into ARDS risk groups based on their LIPS), ventilation management, development of ARDS and other pulmonary complications (such as pneumonia, pneumothorax, pleural effusion, atelectasis, and cardiogenic pulmonary oedema), duration of ventilation, number of ventilator-free days, length of stay in the ICU and hospital, and mortality in the ICU, hospital, and at 90 days (PRoVENT); ventilator settings and additional ventilation variables and parameters, proportion of patients at risk for or presenting ARDS, incidence of pulmonary complications, length of stay in the ICU, and all-cause ICU mortality (PRoVENT–iMiC); PRoVENT and PRoVENT iMic differed in that PRoVENT reviewed medical records for ventilation complications from day 1 to day 7 as well as at ICU and hospital discharge, whereas PRoVENT iMic only examined the period from day 1 to day 7. Extubation failure, re-intubation, tracheostomy, duration of non-invasive mechanical ventilation (mask and high-flow nasal oxygen cannula), ICU length of stay, in-ICU mortality, and in-hospital mortality (ENIO).

Harmonization and merging

Reasons for invasive ventilation were harmonized and merged into PRIME as shown in Table 2, and included the categories ‘COPD exacerbation’, ‘unplanned postoperative ventilation’, ‘chronic heart failure’, ‘community–acquired pneumonia’, ‘nosocomial pneumonia’, ‘sepsis (other than pneumonia)’, ‘metabolic/hepatic encephalitis’, ‘intoxication’, and neurologic conditions like ‘traumatic brain injury’, ‘subarachnoid haemorrhage’, ‘intra–cranial bleeding’, ‘stroke’, and central ‘nervous system infection’. Ventilation days were harmonized and merged as ‘day 0’ for the first observation of ventilatory variables in patients receiving invasive mechanical ventilation in the ICU, ‘day 2’ for the evaluation after 2 days, and ‘day 6’, for the last assessment. ICU mortality was defined as all-cause death occurring in the ICU (Table 3).

Table 2.

Data collected across the three 3 studies.

	Reason for intubation	Data collection frequency
PRoVENT	Depressed level of consciousness sub-divided into:	Day 0 defined as first calendar day patient received invasive ventilation; Extended dataset collected daily until day 7 and at discharge from the ICU. Any day patient received mechanical ventilation, irrespective of duration or type of ventilation
	• TBI • SAH • ICH • AIS • Central nervous system infection (such as: brain abscess, empyema, meningitis, encephalitis)
	Other reasons sub-divided into:
	• COPD exacerbation • unplanned postoperative ventilation • chronic heart failure • community acquired pneumonia • nosocomial pneumonia • sepsis (other than pneumonia) • metabolic/hepatic encephalitis • intoxication
	Other reasons (not inherent to the brain injury)
PRoVENT-iMiC	Depressed level of consciousness sub-divided into:	Baseline and demographic variables collected on the day of admission; Extended dataset collected daily until day 7 and at discharge from the ICU
	• TBI • SAH • ICH • AIS • Central nervous system infection (such as: brain abscess, empyema, meningitis, encephalitis)
	Other reasons sub-divided into:
	• COPD exacerbation • unplanned postoperative ventilation • chronic heart failure • community acquired pneumonia • nosocomial pneumonia • sepsis (other than pneumonia) • metabolic/hepatic encephalitis • intoxication
	Other reasons (not inherent to the brain injury)
ENIO	Neurocritical care patients with GCS < 12:	Demographic and baseline data were collected at the time of enrolment. Ventilatory parameters, arterial blood gas values and respiratory data collected at day 1, day 3, and day 7 after ICU admission.
	• TBI • SAH • ICH • AIS • Central nervous system infection (such as: brain abscess, empyema, meningitis, encephalitis)
	Other reasons (inherent to the brain injury, such as tumour)
PRIME	Population: Patients with ABI: TBI, SAH, ICH, AIS, central nervous system infection and Patients without ABI: all other reasons	The initial observations of ventilatory parameters (labelled as Day 0) occur on the admission day in PRoVENT and PRoVENT-iMiC, and the following day in ENIO. Subsequent evaluations take place after 2 days in the ICU (labelled as Day 2), with the second observation on day 2 in PRoVENT and PRoVENT-iMiC, and day 3 in ENIO. The third assessment, performed after 6 days in the ICU (labelled as Day 6), is recorded as day 6 in PRoVENT and PRoVENT-iMiC, and day 7 in ENIO

Abbreviations: ABI, Acute Brain Injury; AIS, Acute Ischemic Stroke; ARDS, Acute Respiratory Distress Syndrome; COPD, Chronic Obstructive Pulmonary Disease; ENIO, Extubation in neurocritical care patients: the ENIO international prospective study; ICU, Intensive Care Unit; ICH, Intracranial Hemorrhage; GCS, Glasgow Coma Scale; NIV, Non-invasive Ventilation; PRoVENT Practice of VENTilation in critically ill patients without ARDS study; PRoVENT–iMiC, Practice of VENTilation in critically ill patients in middle income countries study; SAH, Subarachnoid Aneurysmal Hemorrhage.

Table 3.

Outcomes in the 3 original studies.

	Risk of ARDS or ARDS at admission	Ventilation management	Occurrence of ARDS	Duration of ventilation	Ventilator–free days and alive at day 28	Length of stay in ICU	Length of stay in hospital	all cause ICU–mortality	Extubation failure and re-intubation	Tracheostomy
PRoVENT	X	X	X	X	X	X	X	X		X
PRoVENT-iMiC	X	X	X			X		X		X
ENIO			X	X		X		X	X	X
PRIME			X			X		X		X

Abbreviations: ARDS, Acute Respiratory Distress Syndrome; ENIO, Extubation in neurocritical care patients: the ENIO international prospective study; ICU, Intensive Care Unit; PRoVENT Practice of VENTilation in critically ill patients without ARDS study; PRoVENT–iMiC, Practice of VENTilation in critically ill patients in middle income countries study.

When one study collected a variable differently––for example, categorizing patients based on the Glasgow Coma Scale into groups such as < 7, between 8 and 12, and > 13, the same categorization method was adopted across all studies. During pooling, any variables not found in at least 2 datasets and data dictionaries were excluded.

Endpoints

The primary endpoint will be all-cause ICU mortality. Secondary endpoints are key ventilator settings and ventilation variables and parameters.

Calculations and definitions

VT is expressed in ml/kg predicted body weight (PBW), wherein PBW is calculated as follows:

in men, PBW = 50 + 0.91 × (height (cm) – 152.4 (cm))

in women, PBW = 45.5 + 0.91 × (height (cm) – 152.4 (cm))

For RR, expressed in breath/min, we use the total RR as reported in the original studies.

Dynamic ΔP is expressed in cm H2O, and calculated as follows:

ΔP = Pplat (cm H2O) – PEEP (cm H2O) (in volume–controlled ventilation)

ΔP = Ppeak (cm H2O) – PEEP (cm H2O) (in pressure–controlled ventilation)

Dynamic CRS is expressed in ml/cm H2O is calculated as follows:

CRS = VT (ml)/ΔP (cm H2O)

Dynamic MP is expressed in J/min, and calculated as follows:11

MP = 0.098 × VT × RR × (Pmax–0.5 × ΔP)

Sample size

No formal sample size calculation was performed. The sample size for this initial analysis of PRIME was based on the number of available patients in the PRIME database. Anticipating a 15% dropout rate due to the application of exclusion criteria for this current analysis, we aimed to include at least 3270 patients.

Statistical analysis

Continuous data will be presented as median with IQR, and categorical variables as frequencies and proportions. Differences in baseline characteristics between patient groups will be analysed using the Pearson Chi–squared or Fisher exact tests for categorical variables and with the Wilcoxon rank–sum test for continuous variables.

In this initial analysis of PRIME, patients will be categorized as either neurocritical care or non–neurocritical care based on the reason for invasive ventilation at admission. Baseline demographics and patient characteristics will be presented in a table for all patients combined and separately for neurocritical care patients versus non–neurocritical care patients. Ventilator settings and ventilation variables and parameters will be shown in another table. Key ventilator settings and ventilation variables and parameters will be presented in cumulative distribution graphs, wherein vertical dotted lines denote a cut-off value for each variable or parameter. Ventilator settings and ventilation variables and parameters will also be compared over time for days 0, 2 and 6.

To identify potentially modifiable and non–modifiable factors contributing to ICU mortality, a multivariable model will be built using demographic factors, comorbidities, illness severities, and respiratory and laboratorial variables at day 0.

In analyses examining the impact of ventilatory variables on outcome, we will estimate the relative risk (RR) of ICU mortality for neurocritical and non-neurocritical care patients by dividing the study population based on the median of MP, ΔP, PEEP and VT as assessed on day 0 of ventilation.

The odds ratio for all-cause ICU mortality of MP will be shown by curves showing the odds ratios corresponding to increments of 1 standard deviation of MP. These curves will be stratified based on neurocritical and non-neurocritical care and adjusted for the variables included in the final model.

Data will be analysed using R 4.0.3 statistical software (www.r-project.org); P < 0.05 will be considered statistically significant. Where appropriate, statistical uncertainty will be expressed by 95% confidence intervals.

Discussion

This is the initial pre-planned analysis of PRIME, a pooled database that harmonized and merged individual patient data from 3 previous global studies of invasive ventilation in critically ill patients. The primary aim is to compare key elements of ventilation management between neurocritical care and non–neurocritical care patients. The secondary aim is to determine which potentially modifiable and non–modifiable factors are associated with ICU mortality in the neurocritical group. The insights gained from this analysis may significantly influence the design of future clinical studies in neurocritical care patients. Specifically, conducting comprehensive sensitivity analyses based on factors including the type of brain injury, patient age, severity of lung conditions, and decisions regarding withdrawal of care could provide valuable insights for further refining study protocols and enhancing patient care strategies.

The newly created database and this initial analysis of PRIME have some strengths. It utilizes a large, pooled database (Fig. 1), allowing for sophisticated and comprehensive analyses that enhance the statistical power and reliability of the findings. The global participation (Fig. 2), which includes data from various types of hospitals in both resource–rich and resource–limited settings, increases the generalizability of the results and the substantial sample size surpasses that of any other database we have reviewed on this topic, suggests its suitability for our research objectives. Additionally, the use of robust statistical methods, including multivariable models, ensures the reliability and validity of the results. Adhering strictly to a predefined statistical analysis plan minimizes the risk of deviations from the initial research hypotheses, thereby maintaining the integrity and credibility of the study.

Figure 1.

Map-chart of the present individual patient data meta-analysis.

Abbreviations: ENIO, Extubation in neurocritical care patients: the ENIO international prospective Study; PRoVENT, Practice of VENTilation in critically ill patients without ARDS study; PRoVENT–iMiC, Practice of VENTilation in critically ill patients in middle income countries study.

Figure 2.

Inclusion flow outline of the present individual patient data meta-analysis.

Abbreviations: ARDS, acute respiratory distress syndrome; ENIO, Extubation in neurocritical care patients: the ENIO international prospective Study; PRoVENT, Practice of VENTilation in critically ill patients without ARDS study; PRoVENT–iMiC, Practice of VENTilation in critically ill patients in middle income countries study.

We limited the endpoint to all-cause ICU mortality, since follow–up differed among the 3 studies: while PRoVENT had a follow–up period of 90 days, follow–up in PRoVENT–iMiC stopped at ICU discharge, and follow–up in ENIO continued till hospital discharge, but stopped for some outcomes if length of ICU stay exceeded 28 days. ICU mortality was chosen as the sole endpoint of this first analysis, as this endpoint was collected and reported consistently in all 3 studies.

The newly created database and this initial analysis of PRIME have some limitations. The 3 parent studies were conducted separately and in different years, which may introduce variability in data collection methods, patient populations, and clinical practices. We cannot adjust for temporal effects. As a posthoc analysis of previously collected observational data, the findings can only be interpreted as hypothesis–generating and need to be confirmed through prospective studies to establish causality. Furthermore, the heterogeneity in mechanical ventilation modes used in clinical practice across the 3 studies necessitates the adoption of surrogate formulas to compute ΔP, CRS and MP. Although we believe the use of dynamic driving pressure and mechanical power should replace the static measures, the inspiratory and expiratory hold manoeuvres required to obtain exact plateau and intrinsic PEEP values are time-consuming and not always compatible with clinical activity. This variability may affect the consistency and comparability of the results.

Despite these limitations, this analysis provides valuable insights that could guide future research and clinical practice. The large sample size, worldwide participation, and rigorous statistical methods employed in this study offer a robust foundation for identifying potentially modifiable factors that may improve the outcome of neurocritical care patients. These findings could have significant clinical implications, paving the way for future studies aimed at improving patient outcomes in this critical area of healthcare.

Following this initial analysis of PRIME, the database will be available for post–hoc analysis. For this, interested investigators will have to provide a detailed analysis plan, containing a testable hypothesis, clearly described endpoints and an analysis plan. After approval, investigators will be provided with a mock database that will allow them to write an analysis script in R. PROVE network investigators will then run the script on the PRIME database and provide the investigators with the results. Any publication from an additional analysis must follow the rules for publication established by the PROVE network.32

Funding

We received no funding for this meta–analysis.

Conflict of interests

The authors declare that they have no conflict of interests.

References

[1]

N.P. Juffermans, P.R.M. Rocco, J.G. Laffey.

Protective ventilation.

Intensive Care Med, 48 (2022), pp. 1629-1631

http://dx.doi.org/10.1007/s00134-022-06820-z | Medline

[2]

A.S. Slutsky, V.M. Ranieri.

Ventilator-induced lung injury.

N Engl J Med, 369 (2013), pp. 2126-2136

http://dx.doi.org/10.1056/NEJMra1208707 | Medline

[3]

M.B.P. Amato, M.O. Meade, A.S. Slutsky, et al.

Driving pressure and survival in the acute respiratory distress syndrome.

N Engl J Med, 372 (2015), pp. 747-755

http://dx.doi.org/10.1056/NEJMsa1410639 | Medline

[4]

M. Urner, P. Jüni, L.P. Rojas-Saunero, et al.

Limiting dynamic driving pressure in patients requiring mechanical ventilation.

Crit Care Med, 51 (2023), pp. 861-871

http://dx.doi.org/10.1097/CCM.0000000000005844

[5]

A.S. Neto, S.N.T. Hemmes, C.S.V. Barbas, et al.

Association between driving pressure and development of postoperative pulmonary complications in patients undergoing mechanical ventilation for general anaesthesia: a meta-analysis of individual patient data.

Lancet Respir Med, 4 (2016), pp. 272-280

http://dx.doi.org/10.1016/S2213-2600(16)00057-6 | Medline

[6]

E.L.V. Costa, A.S. Slutsky, L.J. Brochard, et al.

Ventilatory variables and mechanical power in patients with acute respiratory distress syndrome.

Am J Respir Crit Care Med, 204 (2021), pp. 303-311

http://dx.doi.org/10.1164/rccm.202009-3467OC | Medline

[7]

G. Bellani, J.G. Laffey, T. Pham, et al.

Epidemiology, patterns of care, and mortality for patients with acute respiratory distress syndrome in intensive care units in 50 countries.

JAMA, 315 (2016), pp. 788

http://dx.doi.org/10.1001/jama.2016.0291 | Medline

[8]

M. Urner, P. Jüni, B. Hansen, M.S. Wettstein, N.D. Ferguson, E. Fan.

Time-varying intensity of mechanical ventilation and mortality in patients with acute respiratory failure: a registry-based, prospective cohort study.

Lancet Respir Med, 8 (2020), pp. 905-913

http://dx.doi.org/10.1016/S2213-2600(20)30325-8 | Medline

[9]

F. Vasques, E. Duscio, I. Pasticci, et al.

Is the mechanical power the final word on ventilator-induced lung injury?—we are not sure.

Ann Transl Med, 6 (2018), pp. 395

http://dx.doi.org/10.21037/atm.2018.08.17 | Medline

[10]

L. Gattinoni, J.J. Marini, F. Collino, et al.

The future of mechanical ventilation: lessons from the present and the past.

Crit Care, 21 (2017), pp. 183

http://dx.doi.org/10.1186/s13054-017-1750-x | Medline

[11]

L. Gattinoni, T. Tonetti, M. Cressoni, et al.

Ventilator-related causes of lung injury: the mechanical power.

Intensive Care Med, 42 (2016), pp. 1567-1575

http://dx.doi.org/10.1007/s00134-016-4505-2 | Medline

[12]

S. Wahlster, M. Sharma, S. Taran, et al.

Utilization of mechanical power and associations with clinical outcomes in brain injured patients: a secondary analysis of the extubation strategies in neuro-intensive care unit patients and associations with outcome (ENIO) trial.

Crit Care, 27 (2023), pp. 156

http://dx.doi.org/10.1186/s13054-023-04410-z | Medline

[13]

N. Matin, K. Sarhadi, C.P. Crooks, et al.

Brain-lung crosstalk: management of concomitant severe acute brain injury and acute respiratory distress syndrome.

Curr Treat Options Neurol, 24 (2022), pp. 383-408

http://dx.doi.org/10.1007/s11940-022-00726-3 | Medline

[14]

S. Mrozek, J.M. Constantin, T. Geeraerts.

Brain-lung crosstalk: implications for neurocritical care patients.

World J Crit Care Med, 4 (2015), pp. 163

http://dx.doi.org/10.5492/wjccm.v4.i3.163 | Medline

[15]

M. Ziaka, A. Exadaktylos.

Brain–lung interactions and mechanical ventilation in patients with isolated brain injury.

Crit Care, 25 (2021), pp. 358

http://dx.doi.org/10.1186/s13054-021-03778-0 | Medline

[16]

C. Robba, M. Van Der Jagt, F. Taccone, G. Citerio, A. Messina.

Fluid balance and hemodynamic monitoring of traumatic brain injured patients: an international survey.

J Neurosurg Anesthesiol, 36 (2024), pp. 177-179

http://dx.doi.org/10.1097/ANA.0000000000000925 | Medline

[17]

M.L.N.G. Malbrain, T. Langer, D. Annane, et al.

Intravenous fluid therapy in the perioperative and critical care setting: executive summary of the International Fluid Academy (IFA).

Ann Intensive Care, 10 (2020), pp. 64

http://dx.doi.org/10.1186/s13613-020-00679-3 | Medline

[18]

M.L.N.G. Malbrain, N. Van Regenmortel, B. Saugel, et al.

Principles of fluid management and stewardship in septic shock: it is time to consider the four D’s and the four phases of fluid therapy.

Ann Intensive Care, 8 (2018), pp. 66

http://dx.doi.org/10.1186/s13613-018-0402-x | Medline

[19]

R. Cinotti, S. Taran, R.D. Stevens.

Setting the ventilator in acute brain injury.

Intensive Care Med, 50 (2024), pp. 1513-1515

http://dx.doi.org/10.1007/s00134-024-07476-7 | Medline

[20]

E.E. Tejerina, P. Pelosi, C. Robba, et al.

Evolution over time of ventilatory management and outcome of patients with neurologic disease.

Crit Care Med, 49 (2021), pp. 1095-1106

http://dx.doi.org/10.1097/CCM.0000000000004921 | Medline

[21]

F.S. Taccone, A.L. De Oliveira Manoel, C. Robba, J.L. Vincent.

Use a “GHOST-CAP” in acute brain injury.

Crit Care, 24 (2020), pp. 89

http://dx.doi.org/10.1186/s13054-020-2825-7 | Medline

[22]

C. Robba, D. Battaglini, A. Abbas, et al.

Clinical practice and effect of carbon dioxide on outcomes in mechanically ventilated acute brain-injured patients: a secondary analysis of the ENIO study.

Intensive Care Med, 50 (2024), pp. 234-246

http://dx.doi.org/10.1007/s00134-023-07305-3 | Medline

[23]

C. Robba, D. Poole, M. McNett, et al.

Mechanical ventilation in patients with acute brain injury: recommendations of the European Society of Intensive Care Medicine consensus.

Intensive Care Med, 46 (2020), pp. 2397-2410

http://dx.doi.org/10.1007/s00134-020-06283-0 | Medline

[24]

P.M. Kochanek, H. Bayır.

Titrating the dose of oxygen after severe traumatic brain injury in the era of precision medicine.

J Neurotrauma, 34 (2017), pp. 3067-3069

http://dx.doi.org/10.1089/neu.2017.5159 | Medline

[25]

C. Robba, D. Battaglini, R. Cinotti, et al.

Individualized thresholds of hypoxemia and hyperoxemia and their effect on outcome in acute brain injured patients: a secondary analysis of the ENIO study.

Neurocrit Care, 40 (2024), pp. 515-528

http://dx.doi.org/10.1007/s12028-023-01761-x | Medline

[26]

A.S. Neto, C.S.V. Barbas, F.D. Simonis, et al.

Epidemiological characteristics, practice of ventilation, and clinical outcome in patients at risk of acute respiratory distress syndrome in intensive care units from 16 countries (PRoVENT): an international, multicentre, prospective study.

Lancet Respir Med, 4 (2016), pp. 882-893

http://dx.doi.org/10.1016/S2213-2600(16)30305-8 | Medline

[27]

L. Pisani, A.G. Algera, A. Serpa Neto, A. Ahsan, A. Beane, K. Chittawatanarat, et al.

Epidemiological characteristics, ventilator management, and clinical outcome in patients receiving invasive ventilation in intensive care units from 10 Asian Middle-income Countries (PRoVENT-iMiC): an international, multicenter, prospective study.

Am J Trop Med Hyg, 104 (2021), pp. 1022-1033

http://dx.doi.org/10.4269/ajtmh.20-1177 | Medline

[28]

R. Cinotti, J.C. Mijangos, P. Pelosi, et al.

Extubation in neurocritical care patients: the ENIO international prospective study.

Intensive Care Med, 48 (2022), pp. 1539-1550

http://dx.doi.org/10.1007/s00134-022-06825-8 | Medline

[29]

PROVE Network [Internet]. Available from: https://www.provenetwork.org/home.

[30]

S. Cuschieri.

The STROBE guidelines.