A combination of oxygenation and driving pressure can provide valuable information in predicting the risk of mortality in ARDS patients

Background Acute respiratory distress syndrome (ARDS) is a common life-threatening condition in critically ill patients. Itis also an important public health issue because it can cause substantial mortality and health care burden worldwide. The objective of this study was to investigate therisk factors that impact ARDS mortality in a medical center in Taiwan. Methods This was a single center, observational study thatretrospectively analyzed data from adults in 6 intensive care units (ICUs) at Taichung Veterans General Hospital in Taiwan from 1st October, 2018to30th September, 2019. Patients needing invasive mechanical ventilation and meeting the Berlin definition criteria were included for analysis. Results A total of 1,778 subjects were screened in 6 adult ICUs and 370 patients fulfilled the criteria of ARDS in the first 24 hours of the ICU admission. Among these patients, the prevalenceof ARDS was 20.8% and the overall hospital mortality rate was 42.2%. The mortality rates of mild, moderate and severe ARDS were 35.9%, 43.9% and 46.5%, respectively. In a multivariate logistic regression model, combination of driving pressure (DP) > 14cmH2O and oxygenation (P/F ratio)≤150 was an independent predictor of mortality (OR2.497, 95% CI 1.201–5.191, p = 0.014). Patients with worse oxygenation and a higher driving pressure had the highest hospital mortality rate(p<0.0001). Conclusions ARDS is common in ICUs and the mortality rate remains high. Combining oxygenation and respiratory mechanics may better predict the outcomes of these ARDS patients.

This statement is required for submission and will appear in the published article if the submission is accepted.Please make sure it is accurate.

Unfunded studies
Enter: The author(s) received no specific funding for this work.Yes -all data are fully available without restriction Tick here if the URLs/accession numbers/DOIs will be available only after acceptance of the manuscript for publication so that we can ensure their inclusion before publication.

Introduction
Acute respiratory distress syndrome (ARDS) is a life-threatening condition which may result from a variety of pulmonary (e.g., pneumonia and aspiration) and extra-pulmonary causes (e.g., non-pulmonary sepsis, trauma and pancreatitis).ARDS is characterized by acute, diffuse alveolar inflammation and flooding, leading to increased capillary permeability, which results in lung tissue edema and loss of aeration.Clinically, patients with ARDS often present with hypoxemia, pulmonary congestion, and decreased respiratory compliance.Patients with ARDS often need invasive mechanical ventilation to rescue their hypoxemia.Although there have been recent advances in the management of ARDS, including protective mechanical ventilation [1], prone positioning [2] and extracorporeal life support [3], the mortality rate remains high [4].In addition, patients who survive ARDS may also suffer from significant sustained disabilities, both physically and psychologically, leading to a decreased quality of life and even increased mortality in later years [5,6].
ARDS is common in critically ill patients who are admitted to intensive care units (ICUs), but often goes unrecognized and undertreated [4].In the United States, ARDS affects approximately 200,000 patients per year and results in about 75,000 deaths annually [7].ARDS is also an important public health issue as it is a common cause of death for severe respiratory system infections, especially influenza [8] and coronavirus disease 2019 (COVID-19) [9,10].
In Taiwan, the epidemiological data, patterns of care and outcomes of patients with ARDS are not clear.Given that ARDS has as an important impact on public health burden, we conducted a prospective observational study to investigate its incidence and factors associated with the outcomes of critically ill patients with ARDS.

Patient and study design
This was an observational study of six adult ICUs in a single tertiary referral institute in central Taiwan.The ICUs were functionally divided into medical, respiratory, neurological, surgical, cardiovascular and trauma.We retrospectively analyzed the data of critically ill patients with ARDS who were admitted to the adult ICUs of Taichung Veterans General Hospital (TCVGH) from the 1 st October 2018 to the 30 th September 2019.Adult patients (age >20 years old) who were admitted to the ICUs needing invasive mechanical ventilation and who met the Berlin definition criteria for ARDS [11] were included for analysis.Patients with ARDS were identified by a respiratory therapist and verified by two critical care physicians with a respiratory specialty.During the study period, we prospectively collected data for a quality improvement program via a clinical audit of daily practice in patients with acute respiratory failure who needed mechanical ventilation in the ICUs.This study was approved by the Institutional Review Board of Taichung Veterans General Hospital Taiwan (IRB number: CE20049B).Data were collected and accessed after IRB approval on February 21, 2020.Given that the research was limited to the secondary use of data previously collected during daily practice and the fact that the patients' identities were completely anonymized, the requirement for informed consent from the study subjects was waived by the IRB of Taichung Veterans General Hospital Taiwan due to no more than minimal risk to the subjects.All methods were carried out in accordance with the IRB's guidelines and regulations.

Data collection
All data was collected from medical records, including the patients' demographic data, comorbidities, and basic laboratory tests.Disease severity scores, including Acute Physiology and Chronic Health Evaluation II (APACHE II) [12], and Sequential Organ Failure Assessment (SOFA) scores [13] were determined on the first, third and seventh days of ICU admission.Mechanical ventilation parameters were also collected, including fraction of inspired oxygen (FiO2), positive end-expiratory pressure (PEEP, cmH2O), tidal volume (VT) adjusted by predicted body weight (PBW), plateau pressure (Pplat) and peak airway pressure (Ppeak) on the first three days of ICU admission.Pplat was measured at end-inspiration by a 0.5 to 1 second inspiration hold on a mechanical ventilator within 24 hours of the initiation of mechanical ventilation.Driving pressure was calculated as the difference between Pplat and PEEP [14].
Information on major comorbidities, including cardiovascular disease, cerebrovascular disease, dementia, chronic pulmonary disease, rheumatic disease and malignancy was collected.The Charlson comorbidity index (CCI) [15] was calculated using the International Classification of Disease 10 th Revision [16] diagnosis codes from the patients' medical records.Diabetes was defined by medical history and a laboratory exam with HbA1c >6.5%.The severity of ARDS was classified by thePaO2/FiO2 (P/F )ratio according to the Berlin definition [11].Mechanical ventilator parameters, including mode, FiO2, VT, PEEP, respiratory rate (RR), peak inspiratory pressure (PIP), and Pplat were collected in the first 24 hours of ICU admission.ARDS severity was determined using the worst partial pressure of oxygen to fraction of inspired oxygen ratio within the first 24 hours following ARDS diagnosis.

Statistical analysis
Data is presented as frequency (percentages) for categorical variables and as mean ±

Patient enrollment and clinical features
Of 1,778 patients with acute respiratory failure who needed invasive mechanical ventilation, 370 (20.8%) fulfilled the criteria of ARDS in the first 24 hours of ICU admission (Figure 1).1).

ARDS severity
In this cohort of 370 patients, 103 (27.8%) had mild ARDS, 196 (53%) had moderate and 71 had severe (19.2%).The increasing severity of ARDS was paralleled by a worsening of the APCHE II and SOFA scores at admission and on the following ICU days.There was no significant difference in age, sex, admission source, ICU type, CCI and etiology among the three severity groups.Patients with higher severity ARDS had a higher body mass index (BMI).Among the co-morbidities, dementia was more common in milder ARDS patients.The clinical outcomes, including ICU stay, hospital stay, ventilator-days and hospital mortality, were not significantly different among the three groups (Table S1).

Mechanical ventilation and adjunctive treatments in ARDS
Mechanical ventilation management on the first day of ICU admission varies with ARDS severity.Volume-targeted ventilation (82.4%) was preferred in this cohort of patients.For the ventilator settings, FiO2, PEEP and respiratory rate increased with ARDS severity, whereas VT decreased.The measured pressures, including PIP, Pplat and driving pressure changed in parallel with ARDS severity.The average VT was 7.5 ± 1.6 mL/kg PBW and 65.5% of these patients had ≤8mL/kg PBW.Pplat was measured in 84.3% of patients and the average Pplat was 21.5 ± 5.2 cm H2O; 96.2% of these patients had ≤30 cm H2O.A total of 64.4% of these patients received protective mechanical ventilation as defined by a VT of ≤8 mL/kg PBW and a Pplat ≤30 cm H2O.
The average driving pressure was 12.9 ± 4.1 cm H2O and 66.4% of these patients had ≤14 cm H2O.Positive end-expiratory pressure was relatively low with the average PEEP 8.8 ± 4.1 cm H2O in all patients and 10.2 ± 3.0 cm H2O in severe patients.
The use of adjunctive treatments in patients with ARDS was relatively low but increased with ARDS severity.In patients with severe ARDS, the percentages of patients using extracorporeal membrane oxygenation, prone position, and recruitment maneuver were 5.6% , 15.5%, and 26.8%, respectively.Neuromuscular blockade was used in 43.0% of all patients and in 64.8% of severe ARDS patients (Table S2).

Characteristics of ARDS patients according to driving pressure
We further analyzed the characteristics of patients by driving pressure.Driving pressure at admission was available in 318 (85.9%) of the patients.Those with a driving pressure >14 cmH2O had a significantly lower survival rate (33.6% vs 52.3%, p=0.001) (Table S3).There was no significant difference in age, gender, BMI, comorbidities, admission source, type of ICU, etiology, length of ICU stay, length of hospital stay or ventilator days between groups of driving pressure (DP) ≤14cmH2O and DP >14 cmH2O.In ARDS patients with different DPs, the ventilator parameters including PIP and Pplat, were significantly lower in the DP ≤14cmH2O group compared with the DP >14 cmH2O group (p<0.0001)(Table S4).

Mortality of ARDS
The hospital mortality in this ARDS cohort was 42.2% and the hospital mortality rate was paralleled by ARDS severity, which was 35.9% in mild, 43.9% in moderate and 46.5% in severe ARDS.There was no significant difference in age, sex, BMI, admission source, and type of ICU between the survivors and non-survivors.
Non-survivors had higher APACHE II and SOFA scores on the days following ICU admission.Non-survivors also had higher CCI scores than survivors.Among the comorbidities, the incidence of rheumatic disease, hepatic disease and malignancy were higher in the non-survivors, but diabetes was more common in survivors (Table 1).Among mechanical ventilator parameters, VT and PEEP were not significantly different between the survivors and non-survivors.However, PIP, Pplat and driving pressure were higher in non-survivors compared with survivors (  3).
As oxygenation (P/F ratio) and respiratory mechanics (driving pressure) were both associated with mortality in ARDS, we constructed an analysis which combined these two factors.Patients with a driving pressure >14 cm H2O (Fig. 2A, p=0.025) and a P/F ratio ≤150 (Fig. 2B, p=0.002) had a higher hospital mortality rate.By combining these factors, patients with worse oxygenation and higher driving pressure had the highest hospital mortality rate (Fig. 2C, p<0.0001).

Discussion
The present study was carried out in 6 ICUs of a referral medical center in central Taiwan.ARDS appears to represent an important health problem, which is both common and has a high mortality, in critically ill patients receiving invasive mechanical ventilation.We found that CCI, driving pressure and APACHE II scores were independent risk factors of mortality, and that a combination of oxygenation status and respiratory mechanics gives better outcome prediction in ARDS patients.
Adhesion to contemporary ventilation strategies and adjuncts in ARDS patients remains unsatisfactory.These findings also indicate the potential for improvement in the management of patients with ARDS.
The burden of critical illness is increasing worldwide because of aging populations, natural disasters, conflicts, and higher-risk medical therapies [17].The number of critically ill patients admitting to ICUs is increasing and most of these patients need mechanical ventilation [18].In the LUNG SAFE study, approximately a quarter of patients admitted to the ICU and who received mechanical ventilation developed ARDS [4], and there is a geographical variation in the incidence of ARDS, which is probably due to the health care system, medical resource availability, and ethnics.Previous epidemiological studies of ARDS in Taiwan are limited due to the small-size of case series [19] and the imprecise case definition used [20].As there has been a substantial evolution in the health care system over the past two decades, there has been the development of a new ARDS definition [11], and epidemiological data on the incidence, pattern of care, and outcome of ARDS patients in Taiwan is urgently needed.Our study demonstrates that the incidence of ARDS remains high, and that the mortality rate of these patients is doubled compared with patients without ARDS.
This highlights the fact that critical care teams should pay more attention to the prevention and management of patients with, or at risk of ARDS.

The presence of comorbidities is common in critically ill patients admitted to
ICUs.There is a trend that the number of comorbidities increases over time in non-surgical patients needing invasive mechanical ventilation [18].The presence of chronic illness, such as active neoplasm, hematological neoplasm, and chronic liver failure, are independent risk factors of worse outcomes in ARDS patients [21].The concept of limiting life-sustaining therapies or measures in critically ill patients varies among Asian countries and regions [22].Comorbidities are not modifiable and can also serve as a key factor in decisions to forgo life-sustaining treatment [23].In the present study, we calculated CCI and found that it independently predicted mortality in ARDS patients.Since the majority of decisions to limit care in critically ill patients are made after the development of ARDS, the presence of comorbidities can provide important information when making this decision.
In real-life practice, recognition of ARDS is often delayed or missed, especially in those with mild disease.However, the mortality of ARDS patients remains high even in those with mild ARDS.In this cohort of ARDS patients, the mortality rate of mild disease was 35.9%, which is similar to the LUNG SAFE study, but still much higher than in those without ARDS.The mortality rate is even higher if ARDS worsens over the following days.Pham et al. emphasized the need to pay close attention to those with mild ARDS [24].Failure to recognize ARDS in a timely fashion can lead to failure to implement strategies that improve survival of ARDS.
Timely implementation of lung protection with low VT is fundamental to successful treatment in ARDS [25].In severe influenza ARDS patients, first VT, shortly after intubation, is associated with increased mortality [26].In this study, although ARDS was recognized in a timely manner at ICU admission, the outcomes of these patients were not satisfactory.This can be attributed to adherence to lung protection and the application of adjunct therapies.Another important issue for ARDS in real-life practice is that it is often under recognized.The LUNG SAFE study also indicated that a lower ratio of healthcare professionals to ICU patients, was associated with reduced recognition of ARDS.A global epidemiological investigation into the burden of critical illness also suggested that ICU organization has an important effect on the risk of death [27].Taken together, our results also highlight the importance of timely identification and proper management of ARDS, both of which may be improved by investing resources to improve the workforce in ICUs.
As hypoxemia is the major clinical presentation of ARDS, the severity of ARDS is traditionally classified by the P/F ratio [28].As variations in PEEP and FiO2 levels can impact the P/F ratio, it is now considered that PEEP or continuous positive airway pressure ≥5 cm H2O is required to define and classify ARDS severity.However, the difference in mortality between mild and severe ARDS was around 10% in LUNG SAFE and our study.Respiratory mechanics, such as Pplat [29] and driving pressure [14], are good predictors of outcomes in ARDS patients.In addition, improvement of respiratory mechanics, such as an increase in dynamic driving pressure [30] or a decrease in PaCO2 [31], are associated with improved survival in prone positioning of severe ARDS patients.The respiratory response to prone positioning was more relevant when PaCO2 rather than the P/F ratio was used [32].In this study, by combining respiratory mechanics and oxygenation status, we found that patients with a driving pressure >14 cm H2O and a P/F ≤150 have worst outcomes.ARDS is also an inflammatory disease, both locally and systematically, and hyper inflammation is associated with clinical outcomes.We therefore suggest the need to develop a new classification model which combines oxygenation, respiratory mechanics and inflammation parameters and which may better predict outcomes for ARDS patients.
A major strength of this study was that the data was collected prospectively by a quality improvement program for a year.In contrast to previous studies which were conducted over 4 weeks (LUNG SAFE) to 2 months [33], this epidemiological study avoided any seasonal variation in ARDS incidence.This study also had several limitations.First, this was a single center study and is inherently limited for external validation.As the Taiwan National Health Insurance is a single payer system, the hospitals share high homogeneity in their practice patterns and ICU organization.As this was an observational study without intervention, the results of this study are representative of real-life information about the diagnosis and management of ARDS in ICUs.Second, we only included patients with invasive mechanical ventilation within 24 hours of ICU admission.Patients with non-invasive ventilation were not included.These patients either had mild ARDS with a good prognosis or were those designated as 'do not intubate' due to their advanced age or end-stage disease.We also did not include ARDS which developed after the first day of ICU mission because most ARDS cases are early onset in ICU admission.However, all these limitations should be taken into account when interpretating the information collected.

Conclusions
ARDS is common among critically ill patients who are admitted to the ICU and who use invasive mechanical ventilation.Hospital mortality remains high and could potentially be improved by adherence to currently available evidence.Both CCI and driving pressure are independent predictors of outcomes in ARDS patients.However, combining oxygenation status and respiratory mechanics may better predict ARDS patients at risk of mortality.

List of abbreviations
the submission guidelines for detailed requirements.View published research articles from PLOS ONE for specific examples.
Funded studies Enter a statement with the following details: Initials of the authors who received each award • Grant numbers awarded to each author • The full name of each funder • URL of each funder website • Did the sponsors or funders play any role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript?• NO -Include this sentence at the end of your statement: The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.• YES -Specify the role(s) played.• * typeset Competing Interests Use the instructions below to enter a competing interest statement for this submission.On behalf of all authors, disclose any competing interests that could be perceived to bias this work-acknowledging all financial support and any other relevant financial or nonfinancial competing interests.This statement is required for submission and will appear in the published article if the submission is accepted.Please make sure it is accurate and that any funding sources listed in your Funding Information later in the submission form are also The authors have declared that no competing interests exist.Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation declared in your Financial Disclosure statement.View published research articles from PLOS ONE for specific examples.NO authors have competing interests Enter: The authors have declared that no competing interests exist.Authors with competing interests Enter competing interest details beginning with this statement: I have read the journal's policy and the authors of this manuscript have the following competing interests: [insert competing interests here] /A" if the submission does not require an ethics statement.General guidance is provided below.Consult the submission guidelines for detailed instructions.Make sure that all information entered here is included in the Methods section of the manuscript.This study was approved by the Institutional Review Board of Taichung Veterans General Hospital Taiwan (IRB number: CE20049B).Given that the research was limited to the secondary use of data previously collected during daily practice and the patients' identity were completely unidentifiable, the informed consent was hence waived Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation Format for specific study types Human Subject Research (involving human participants and/or tissue) Give the name of the institutional review board or ethics committee that approved the study • Include the approval number and/or a statement indicating approval of this research • Indicate the form of consent obtained (written/oral) or the reason that consent was not obtained (e.g. the data were analyzed anonymously) • Animal Research (involving vertebrate animals, embryos or tissues) Provide the name of the Institutional Animal Care and Use Committee (IACUC) or other relevant ethics board that reviewed the study protocol, and indicate whether they approved this research or granted a formal waiver of ethical approval • Include an approval number if one was obtained • If the study involved non-human primates, add additional details about animal welfare and steps taken to ameliorate suffering • If anesthesia, euthanasia, or any kind of animal sacrifice is part of the study, include briefly which substances and/or methods were applied • Field Research Include the following details if this study involves the collection of plant, animal, or other materials from a natural setting: Field permit number • Name of the institution or relevant body that granted permission • Data Availability Authors are required to make all data underlying the findings described fully available, without restriction, and from the time of publication.PLOS allows rare exceptions to address legal and ethical concerns.See the PLOS Data Policy and FAQ for detailed information.

Figure 1 .
Figure 1.Flow chart of patient enrollment.

Figure 2 .
Figure 2. Prediction of hospital survival when driving pressure and P/F ratio are combined.DP, Driving pressure; P/F ratio, ratio of partial pressure of oxygen to fraction of inspired oxygen Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation standard deviation or median (interquartile range [IQR]) for continuous variables.Differences between alive and dead groups were determined using the Student's t-test and the Mann-Whitney U test for continuous variables and the Chi-squared test for categorical variables.One-way analysis of variance was used to compare variables among the three ARDS severity groups.Kaplan-Meier analysis was performed to test the mortality correlation of driving pressure above or below 14 cmH2O in different ARDS severities.The log-rank test was used to compare mortality between two groups.A multivariate logistic regression model was conducted to identify independent variables that predicted mortality.Statistical significance was set at a two-sided p value of <0.05.All data were analyzed using SPSS software version 22.0 (SPSS Inc., IBM Corp, Armonk, NY, USA).
The hospital mortality of patients with ARDS was 42.2%, while in those without ARDS it was just 21.0%.In this cohort of ARDS patients, the Most of these ARDS patients were admitted via the emergency room (81.9%) and there were more in the medical (79.7%)ICU compared with the surgical (20.3%)ICU.The overall disease severity in this cohort of ARDS patients was high in terms of APACHE II and SOFA scores at admission and on the following ICU days.

Table 1 .
Demographic data of ARDS patients categorized by hospital mortality ARDS, acute respiratory distress syndrome; BMI, Body Mass Index; ICU, intensive care unit; APCHE II, Acute Physiology And Chronic Health Evaluation II; SOFA, The sequential organ failure assessment score; CCI, P value represents comparisons between the survivors and non-survivors ARDS patients. a

Table 2 .
Mechanical ventilation and arterial blood gas on admission a P value represents comparisons between the survivors and non-survivors ARDS patients.

Table 3 .
Multivariate logistic regression model for hospital mortalityAbbreviations: ARDS, acute respiratory distress syndrome; MAP, Mean Airway Pressure; BMI, Body Mass Index; APCHE II, Acute Physiology And Chronic Health Evaluation II; CCI, Charlson comorbidity index; CI., confidence interval.
a Univariate analysis.bAll variables included in the multivariable analysis are reported in this Table