Abstract
Rationale We wished to determine the influence of sex on the management and outcomes in acute respiratory distress syndrome (ARDS) patients in the Large Observational Study to Understand the Global Impact of Severe Acute Respiratory Failure (LUNG SAFE).
Methods We assessed the effect of sex on mortality, intensive care unit and hospital length of stay, and duration of invasive mechanical ventilation (IMV) in patients with ARDS who underwent IMV, adjusting for plausible clinical and geographic confounders.
Findings Of 2377 patients with ARDS, 905 (38%) were female and 1472 (62%) were male. There were no sex differences in clinician recognition of ARDS or critical illness severity profile. Females received higher tidal volumes (8.2±2.1 versus 7.2±1.6 mL·kg−1; p<0.0001) and higher plateau and driving pressures compared with males. Lower tidal volume ventilation was received by 50% of females compared with 74% of males (p<0.0001). In shorter patients (height ≤1.69 m), females were significantly less likely to receive lower tidal volumes. Surviving females had a shorter duration of IMV and reduced length of stay compared with males. Overall hospital mortality was similar in females (40.2%) versus males (40.2%). However, female sex was associated with higher mortality in patients with severe confirmed ARDS (OR for sex (male versus female) 0.35, 95% CI 0.14–0.83).
Conclusions Shorter females with ARDS are less likely to receive lower tidal volume ventilation, while females with severe confirmed ARDS have a higher mortality risk. These data highlight the need for better ventilatory management in females to improve their outcomes from ARDS.
Abstract
Shorter females with ARDS were less likely to receive lower tidal volume ventilation than shorter males, while mortality rates were higher in females with confirmed severe ARDS. Better ventilatory management may improve outcomes in females with ARDS. http://bit.ly/2JIsUBz
Introduction
Differences in the clinical management and outcomes of females versus males are well described [1]. The reasons underlying these differences are complex, with biological, organisational, case-mix, ethnicity, socioeconomic and local therapeutic traditions having an influence [2]. Differences related to age profile and disease severity and/or complexity may also play a role [3–5]. For example, females hospitalised with coronary artery disease are less likely to undergo invasive diagnostic and therapeutic intervention despite similar rates of presentation with acute myocardial infarction [1], but these differences may be related to older age at presentation. In females presenting with haemorrhagic stroke, the lower intervention rates is partially explained by more complex disease and older age at presentation [5].
In the critically ill, the impact of sex is less well understood. Sex may affect access to critical care, with females less likely to be admitted to the intensive care unit (ICU) [6]. However, once in the ICU there have been few studies indicating sex bias in the provision of care [3, 6]. In a prospective study of ICU admissions in Austria, while females had greater illness severity, males were independently more likely to receive invasive procedures, although outcomes were not different by sex [6]. Sex-based biological differences may influence the development and/or management of acute respiratory distress syndrome (ARDS) [7–9]. Han et al. [10] demonstrated that shorter patients (i.e. predominantly females) with severe sepsis-related acute lung injury received lung protective ventilation less frequently, a finding confirmed in an analysis of ARDS network trials [11]. Height is frequently inaccurately estimated, particularly in shorter patients [12], further increasing risk in females, given their shorter stature. Sex-specific hormonal differences can influence inflammation and immunological function [13, 14], which may impact on the risk of developing ARDS. The impact of sex on outcomes from ARDS is less clear [15–17].
We wished to address these issues in a secondary analysis of the Large Observational Study to Understand the Global Impact of Severe Acute Respiratory Failure (LUNG SAFE), a global multicentre cohort study [7]. Our primary objective was to determine the impact of sex on outcomes from ARDS. Secondary objectives were to assess differences in clinician recognition, patient management and progression of ARDS by sex.
Methods
Study design, patients and data collection
The detailed methods and protocol for LUNG SAFE have been published elsewhere [7]. In brief, LUNG SAFE was an international, multicentre, prospective cohort study, conducted during 4 consecutive weeks in the winter of 2014 in a convenience sample of 459 ICUs from 50 countries across six continents [7]. The study, funded by the European Society of Intensive Care Medicine (ESICM), was endorsed by multiple national societies/networks (supplementary appendix S1). National coordinators and site investigators were responsible for obtaining ethics committee approval, and for ensuring data integrity and validity (supplementary appendix S1). Further details are available in the supplementary material.
LUNG SAFE enrolled patients with acute hypoxaemic respiratory failure (AHRF) admitted to a study ICU who underwent invasive or noninvasive ventilation. Exclusion criteria were age <16 years or inability to obtain informed consent (where required). Patients were classified as having ARDS if they fulfilled all of the Berlin criteria [7]. We restricted subsequent analyses to patients that fulfilled ARDS criteria (93%) within 48 h of the onset of AHRF and who received invasive mechanical ventilation (IMV) (supplementary figure E1).
Data definitions
Our data definitions have been previously reported [7, 18, 19]. For the purposes of this analysis, sex assignment was made by the site investigators at the time of data entry. Lower tidal volume (LTV) ventilation was defined as a tidal volume ≤8 mL·kg−1 ideal body weight (IBW). In patients in whom plateau pressure was measured, lung protective ventilation (LPV) was defined as tidal volume ≤8 mL·kg−1 IBW and plateau pressure ≤30 cmH2O. From the variables originally collected we also derived dynamic compliance and body mass index (BMI). We used the threshold of 1.69 m (median height value) to classify shorter versus taller patients. Gross domestic product (GDP) per person was obtained through the World Bank database that gathers time series data for all countries on a variety of socioeconomic topics. GDP was used to define three major geoeconomic groupings: high-income countries in Europe, high-income countries in the rest of the world and middle-income countries (http://databank.worldbank.org/data/home.aspx). Duration of IMV was calculated as the number of days between the date of intubation and the date of extubation in the ICU (or death, if the patient died under IMV). Similarly, invasive ventilator-free days were calculated as the number of days from weaning from IMV to day 28, while patients who died before weaning were considered to have a ventilator-free day value of 0. Length of stay (LOS) in the ICU and hospital was evaluated as the number of days between the date of admission and the date of discharge from the ICU and hospital, respectively. Survival was evaluated at ICU and hospital discharge or at day 90, whichever occurred first. Because we previously observed a significant association between the presence of ARDS at day 2 and outcomes [20], ARDS severity was reclassified (“resolved” versus “confirmed” ARDS) on day 2 using the Berlin criteria.
Data management and statistical analyses
Descriptive statistics were reported for the study population stratified according to sex, and included proportions for categorical variables and mean with standard deviation or median (interquartile range) for continuous variables. No assumptions were made for missing data, which were rare. Comparisons between groups were performed using the Chi-squared test (or Fisher's exact test) for discrete variables and the t-test (or Wilcoxon–Mann–Whitney test) for continuous variables. The Shapiro–Wilk test was used to assess normality in data distribution.
The Kaplan–Meier approach was applied to assess the probability of discontinuing IMV in the ICU, and the probability of hospital survival and of being discharged alive during hospital stay. When assessing the probability of discontinuing IMV in the ICU, patients that weaned from IMV after 28 days in the ICU were considered as censored at day 28. When assessing the probability of being discharged alive from hospital, patients that died before day 90 were considered as censored at date of death, while patients discharged after day 90 were considered as censored at day 90. The log-rank test was used to compare curves between the female and male populations.
To evaluate the existence of a possible effect of sex on mortality, LTV ventilation during the first day of ARDS, LOS and duration of IMV adjusting for all plausible confounders, we applied generalised linear mixed models with random intercept, taking into account the correlation among patients within the same ICU of enrolment. Patients who died before ICU discharge were removed from the analysis of LOS and duration of IMV. In detail, the logistic-link function and binomial distribution of outcome were used to analyse mortality and LTV ventilation, while the log-link function and Poisson distribution were used for LOS and duration of IMV. In the first case, results were reported as odds ratio with 95% confidence interval, while in the second as incidence rate ratio with 95% confidence interval. As some ICUs had few observations to support the normal assumption, the bootstrap method was used (1000 samples randomly extracted) to estimate the model parameters.
Predictors used in the multivariable models were detected through the stepwise regression approach that combines forward and backward selection methods in an iterative procedure (significance level of 0.05 both for entry and retention). Potential independent predictors were: patient characteristics at baseline (age, sex, BMI, geoeconomic area), chronic disease (chronic obstructive pulmonary disease (COPD), diabetes mellitus, immuno-incompetence, cardiac failure, renal failure, liver failure), presence of ARDS risk factors, ICU characteristics (number of beds, proportion of ICU beds in hospital, number of beds per physician and per nurse, academic ICU), clinical parameters measured on day 2 from ARDS onset (total respiratory rate, tidal volume, presence of controlled ventilation, positive end-expiratory pressure (PEEP), standardised minute ventilation, ARDS severity, dynamic compliance, partial pressure of carbon dioxide (PaCO2), pH, nonpulmonary Sequential Organ Failure Assessment (SOFA) score, presence of adjunctive measures performed during 2 day from ARDS onset). Moreover, when sex was identified as a statistical significant predictor, we also evaluated its interaction with other selected predictors.
All p-values were two-sided, with p<0.05 considered as statistically significant. Statistical analyses were performed with R version 3.5.2 (www.R-project.org) and SAS version 9.4 (SAS Institute, Cary, NC, USA).
Results
Of the 12 906 patients enrolled in LUNG SAFE, 4499 developed criteria for AHRF: 1716 (38.1%) females and 2783 (61.9%) males (supplementary figure E1). Identical proportions of females and males were observed in the study population of 2377 patients who developed ARDS within 2 days from AHRF onset and were managed with IMV (supplementary figure E1). The proportions of males and females aged ≥50 years were also unchanged in the study population (table 1).
Characteristics of the 2377 invasively ventilated female and male patients with acute respiratory distress syndrome (ARDS)
In the study population, female patients (n=905) were shorter and had a higher BMI, and a greater frequency of immuno-incompetence (i.e. steroid use, cancer or haematological malignancies), while male patients (n=1472) had a higher frequency of COPD and chronic renal failure (table 1). There were differences in risk factor profile, with gastric aspiration, major trauma, pulmonary contusion and inhalation injury each less frequent in females (table 1). There was no difference in rates of clinician recognition of ARDS by sex.
There were no differences by sex with regard to the severity profile of ARDS or overall severity of illness as defined by SOFA score on day 1 or 2 of ARDS (table 2 and supplementary table E1). No statistically significant differences were observed between the sexes in the progression of ARDS severity, whether in mild (p=0.2191), moderate (p=0.3575) or severe (p=0.1613) patients on day 2 (supplementary figure E2). There were no sex differences in inspired oxygen use, arterial oxygen partial pressure/inspiratory oxygen fraction ratio or oxygen saturation on day 1 or 2.
Ventilatory management and illness severity of the 2377 invasively ventilated female and male patients with acute respiratory distress syndrome (ARDS)
The management of IMV differed by sex (figure 1, table 2 and supplementary table E1). Tidal volumes were higher in females, both in those that received controlled and assisted mechanical ventilation. Peak, plateau and driving pressures were higher, and standardised minute ventilation was higher, in females on days 1 and 2 compared with males (figure 1a and b, table 2, and supplementary table E2). In patients in whom plateau pressure was measured, more male patients received LPV (75% versus 51%; p<0.0001) (figure 1c and d). In contrast, there were no differences in PEEP levels used. PaCO2 was lower and minute volume higher in females on day 1 of ARDS (table 2), and a greater proportion of females were hypocapnic, while more males were hypercapnic (table 2).
Impact of sex on ventilatory parameters at day 1 of acute respiratory distress syndrome (ARDS) in invasively ventilated patients. IBW: ideal body weight; ECMO: extracorporeal membrane oxygenation; HFOV: high-frequency oscillatory ventilation. a, b) Cumulative frequency distribution of a) tidal volume and b) plateau pressure in females (n=278) and males (n=445). c, d) Distribution of tidal volume and plateau pressure in c) females (n=278) and d) males (n=445) by ARDS severity at onset. Data in (a)–(c) are limited to patients with available plateau pressure, and in whom either an assist control mode was used or, where a mode permitting spontaneous ventilation was used, the set and total respiratory rates were equal. Patients receiving HFOV or ECMO were also excluded. p-values in (d) refer to comparison with female patients. d) Box plot for tidal volume in the female and male populations stratified by quintiles of height in the study population. *: p<0.05 for the comparison between females and males. Pearson correlation coefficient between tidal volume and height in females and males is −0.3439 (p<0.0001) and −0.2855 (p<0.0001), respectively. The dashed lines in (a)–(d) indicate cut-offs of tidal volume 8 mL·kg−1 IBW and plateau pressure 30 cmH2O. e) Odds ratios (95% CI) for males versus females of receiving lower tidal volume ventilation (tidal volume ≤8 mL·kg−1 IBW).
Tidal volume sizes increased at lower height quintiles in both sexes (figure 1e). At the lowest quintile (height ≤1.60 m) where >80% were female, tidal volume was significantly higher in females (figure 1e). Tidal volumes were higher in females at each quintile of actual body weight (supplementary figure E3a). LOESS (locally weighted smoothing) regression demonstrated a clear relationship between lower height and higher tidal volume in females and males (supplementary figure E3b). At similar dynamic compliance, females received higher tidal volumes compared with males (supplementary figure E3c). Females continued to receive larger tidal volumes over each day of follow-up to day 14 (supplementary figure E3c and supplementary table E3). More females received adjunctive measures (p=0.0322), with extracorporeal membrane oxygenation (ECMO) and inhaled vasodilators used more frequently in females compared with males (table 2).
Multivariable models were constructed to examine factors associated with the use of LTV ventilation (supplementary table E4 and figure 1f). Factors including higher PaCO2, controlled ventilation, higher pH and geoeconomic area were associated with LTV use (supplementary table E6). In shorter patients (i.e. height less than or equal to the median value of 1.69 m), male sex was independently associated with use of LTV ventilation (figure 1f).
There were differences in the recovery profile from ARDS between the sexes. Female patients had a shorter duration of IMV and reduced length of ICU and hospital stay compared with males (table 3). In surviving patients, female patients had more invasive ventilator-free days in the ICU, a shorter duration of IMV and reduced length of ICU and hospital stay compared with males (table 3). The overall probability of being discharged alive by day 90 was higher for female patients (figure 2a).
Outcomes observed during follow-up in the 2377 invasively ventilated female and male patients with acute respiratory distress syndrome
Kaplan–Meier curves for main outcomes in acute respiratory distress syndrome (ARDS) patients invasively ventilated, stratified by sex. IMV: invasive mechanical ventilation; ICU: intensive care unit. The number of patients at risk refers to the end of the corresponding day. a) Probability of being discharged alive from hospital. Patients discharged alive/dead after day 90 or dead patients were considered as censored at day 90 or at date of death, respectively. b) Probability of discontinuing IMV in the ICU. Patients with weaning from IMV after 28 days in the ICU were considered as censored at day 28. c) Probability of hospital survival. Patients discharged alive before day 90 were considered alive at day 90. All p-values are from the log-rank test.
Overall ICU and hospital mortality rates were identical for both females and males (table 3). Kaplan–Meier analyses demonstrated no differences in the probability of discontinuing IMV or of hospital mortality during follow-up between the sexes (figure 2b and c). There were no differences in limitation of life-sustaining measures such as withdrawal or withholding of support between females and males (table 3).
When patients were stratified by severity of ARDS on day 2, ICU mortality (64% versus 46%; p=0.0066) and hospital mortality (68% versus 50%; p=0.0068) were significantly higher for females who had severe “confirmed” ARDS compared with males (supplementary figure E4 and supplementary table E2). Kaplan–Meier analyses of hospital survival in female and male patients stratified by ARDS severity (at day 2) demonstrated a significantly lower probability of survival in females with severe confirmed ARDS (figure 3).
Probability of hospital survival in female and male patients stratified by acute respiratory distress syndrome (ARDS) severity at day 2: a) resolved ARDS, b) mild ARDS, c) moderate ARDS and d) severe ARDS. The number of patients at risk refers to the end of the corresponding day. Patients discharged alive before day 90 were considered alive at day 90. All p-values are from the log-rank test.
After adjusting for all potential confounders and considering the correlation among patients within the same ICU of enrolment, female sex was independently associated with reduced length of IMV and shorter ICU and hospital stay in survivors (figure 4a and supplementary table E5). Female sex was associated with increased ICU and hospital mortality in patients with severe confirmed ARDS (figure 4b and supplementary table E5). Subsequent analysis of males and females with severe confirmed ARDS revealed no major differences in demographics or illness severity, but there were significant differences in ventilatory management between the sexes (supplementary table E6).
Effects of sex on main outcomes in patients with acute respiratory distress syndrome (ARDS). IMV: invasive mechanical ventilation; ICU: intensive care unit; IRR: incidence rate ratio. a) IRR for males versus females for length of ICU stay, length of hospital stay and duration of IMV. b) Odds ratios for males versus females for ICU and hospital mortality by ARDS severity at day 2 (resolved, mild, moderate and severe).
While there were geoeconomic area differences in the use of LTV ventilation (supplementary table E4) and in overall outcomes of ARDS (supplementary tables E5 and E7) as previously reported [21], geoeconomic area did not modify the relationships between sex and outcomes.
Discussion
Important sex-based differences exist in the demographics, management and outcomes of patients with ARDS. There are differences in the demographics and risk factors for ARDS between females and males. Female patients are more likely to receive nonprotective lung ventilation with higher tidal and minute volumes, and higher plateau and driving pressures, compared with males. While this was due in part to their lower height profile, shorter females were more likely to receive LTV ventilation than shorter males. While females had a shorter duration of IMV and a reduced length of ICU and hospital stay, ICU and hospital mortality rates were identical compared with males. Of particular concern is the fact that female sex was independently associated with higher mortality in patients with severe confirmed ARDS.
Demographic differences are consistent with previous studies which found that ARDS occurs more commonly in males than in females [10, 17]. We found no differences in age or clinician recognition of ARDS. Females with ARDS were shorter and had a higher BMI compared with males. Risk factors such as trauma, COPD and chronic renal failure were less common and immunosuppression more common in females, likely reflecting differences in smoking patterns, chronic kidney disease risk and cancer among females and males [22]. Differences in sex hormones, such as oestrogen and testosterone, can influence inflammation and immunological function [13, 14], which may impact on the risk of developing ARDS. In pre-pubertal children, where hormone levels should be equally low between males and females, ARDS due to nonseptic causes occurs equally. However, ARDS secondary to sepsis occurs at a greater frequency in male children, suggesting a biological reason for the difference in risk for ARDS where hormonal levels between males and females are equal [14]. In this analysis, in patients aged >50 years there was no increase in the proportion of females to males, suggesting that the frequency of ARDS does not increase in females post-menopause.
We next assessed if ventilatory management of ARDS differs between female and male patients with ARDS. A higher proportion of females received nonprotective lung ventilation and a greater proportion were hypocapnic. These differences in ventilation in females persisted out to day 14, suggesting this was sustained over time. If actual body weight rather than IBW is used to calculate the delivered tidal volume, the tidal volume may be inappropriately high [10]. Interestingly, at each quintile of actual body weight, females received significantly higher tidal volumes, while weight was not associated with LTV ventilation use. We found that in shorter patients, both female and male, clinicians were less likely to apply LTV ventilation, confirming and expanding prior findings [10]. While the delivery of higher tidal volumes to females patients has been previously described [11, 23], this has been attributed solely to their (shorter) height [10, 11, 23]. Our findings show that female sex is a factor, with shorter females significantly less likely to receive LTV ventilation than similarly sized males.
Mortality in females may have been lower if LPV had been appropriately used. Understanding the barriers to LPV implementation is important [10, 24, 25]. LPV was associated with an 8.8% reduction in mortality in one study, with another finding that each 1 mL·kg−1 increase in initial tidal volume above predicted body weight was associated with a 23% increase in mortality [26]. Overall, females are shorter than males and height measurement errors are magnified in shorter individuals, particularly where height is estimated [12]. In individuals undergoing abdominal surgery >4 h duration, female sex, short stature and obesity were associated with use of tidal volumes >10 mL·kg−1 [27]. The greater risk of shorter females receiving inappropriately high tidal volumes highlights the need for particular care and attention in calculating and applying LTVs to this cohort.
There were important differences in the recovery profile and outcomes between the sexes. Females with ARDS required shorter ICU stays and had more ventilator-free days. This did not appear to be a “survivor bias” as the finding persisted in surviving patients. Despite this, overall ICU and hospital mortality were identical in females and males with ARDS. Prior studies have reported similar findings, with female patients with critical illnesses including acute lung injury having shorter ICU stays and requiring less resource use [6, 15, 17, 28, 29], but with similar mortality rates to their male counterparts. There are pre-clinical and clinical data to suggest that females and males respond differently to inflammation [30] and to mechanical ventilation [17]. In one study, mechanically ventilated females had a faster alveolar fluid clearance rate compared with males [31]. These findings raise the possibility that females may recover from ARDS faster than males, but that this may be nullified by the lower rates of LPV in females. Additional studies are required to further examine these issues.
Our findings regarding outcomes from ARDS provide important additional insights compared with prior studies. Our finding of no overall differences in outcome from ARDS by sex confirms prior findings in a meta-analysis of outcomes from the ARDSnet studies [32]. Other studies have noted worse outcomes for females on mechanical ventilation, but with organ dysfunction explaining most of the differences [33]. We examined outcomes by day 2 severity, as we have previously demonstrated that reclassification at day 2 provides additional insights [20]. The finding that female patients with severe confirmed ARDS had increased mortality is a novel finding and is of concern. Differences in risk factor profile (e.g. greater pneumonia and nonpulmonary sepsis, lower trauma) may explain some of the outcome differences between the sexes, but overall these differences were relatively small. There were relatively little differences in illness severity profiles between males and females with severe confirmed ARDS. However, females with severe ARDS received higher tidal volumes and were exposed to higher airway pressures compared with males with severe ARDS, which is of concern given their worse outcome compared with their male counterparts. In this study, females were not more likely to receive orders relating to limitation of life-sustaining therapies following ICU admission, which is consistent with other studies [34, 35].
As previously reported, differences exist in the management of patients with ARDS based on geoeconomic region [21]. Differences in outcomes related to geoeconomic region were noted, with ICU and hospital mortality lowest in high-income countries. However, geoeconomic difference was not a factor in modifying the relationship between sex and mortality. The proportion of patients with severe ARDS was lower while use of LPV was higher in non-European high-income countries, which might explain the lowest proportion of female patients dying from ARDS in these countries compared with other geoeconomic areas [21]. It is reassuring that no sex-based differences in access to care for patients with ARDS were found in the study.
This study has several limitations related to study design, as have been previously described [7]. Limitations pertaining to this study include the possibility that all females with ARDS were not accounted for, i.e. due to nonadmission and palliation outside of the ICU. Associations of management with worse outcomes in females with severe ARDS were examined only for days 1 and 2 of ARDS. However, early management of ARDS is important to outcome and choosing a longer time-point to analyse would have decreased patient numbers due to ICU discharges or patient death.
In conclusion, higher proportions of shorter females with ARDS receive nonprotective lung ventilation compared with similarly sized males. There appear to be differences in recovery profiles between the sexes, with females requiring shorter ICU stays and more ventilator-free days. Of concern, females with severe confirmed ARDS have a higher mortality compared with their male counterparts. These findings suggest that females with ARDS require particular attention to their ventilatory management in order to optimise their outcomes.
Supplementary material
Supplementary Material
Please note: supplementary material is not edited by the Editorial Office, and is uploaded as it has been supplied by the author.
Supplementary material ERJ-00609-2019.Supplement
Shareable PDF
Supplementary Material
This one-page PDF can be shared freely online.
Shareable PDF ERJ-00609-2019.Shareable
Acknowledgement
A complete list of LUNG SAFE national coordinators, site investigators and national societies endorsing the study can be found in the supplementary material.
Footnotes
This article has supplementary material available from erj.ersjournals.com
Support statement: This work was funded and supported by the European Society of Intensive Care Medicine (ESICM; Brussels, Belgium), by St Michael's Hospital (Toronto, ON, Canada) and by the University of Milan-Bicocca (Monza, Italy). These funders had no role in the design and conduct of the study; management, analysis and interpretation of the data; preparation, review or approval of the manuscript; or decision to submit the manuscript for publication. Funding information for this article has been deposited with the Crossref Funder Registry.
Author contributions: Study concept and design: J.G. Laffey, G. Bellani, T. Pham and L. Brochard; acquisition of data: J.G. Laffey, G. Bellani, T. Pham and L. Brochard; analysis and interpretation of data: B.A. McNicholas, F. Madotto, E. Rezoagli, T. Pham, C.H. Masterson, S. Horie and J.G. Laffey; drafting of the manuscript: all authors; critical revision of the manuscript for important intellectual content: all authors; statistical analysis: F. Madotto, T. Pham and E. Rezoagli; obtained funding: G. Bellani and J.G. Laffey; study supervision: J.G. Laffey, G. Bellani and L. Brochard; approved the final draft of the manuscript: all authors.
Conflict of interest: B.A. McNicholas has nothing to disclose.
Conflict of interest: F. Madotto has nothing to disclose.
Conflict of interest: T. Pham has nothing to disclose.
Conflict of interest: E. Rezoagli has nothing to disclose.
Conflict of interest: C.H. Masterson has nothing to disclose.
Conflict of interest: S. Horie has nothing to disclose.
Conflict of interest: G. Bellani reports grants and personal fees from Draeger Medical, personal fees from Hamilton, Getinge and Dimar SRL, outside the submitted work.
Conflict of interest: L. Brochard reports grants from Covidien, grants and nonfinancial support from Fisher Paykel, Air Liquide and General Electric, nonfinancial support from Philips and Sentec, outside the submitted work.
Conflict of interest: J.G. Laffey has nothing to disclose.
- Received March 26, 2019.
- Accepted July 4, 2019.
- Copyright ©ERS 2019
References
