Predictors of Respiratory Support Use in Emergency Department Patients With COVID-19-Related Respiratory Failure
================================================================================================================

* Neha N Goel
* Erin Eschbach
* Daniel McConnell
* Bryan Beattie
* Sean Hickey
* John Rozehnal
* Evan Leibner
* Gary Oldenburg
* Kusum S Mathews

## Abstract

**BACKGROUND:** Given the known downstream implications of choice of respiratory support on patient outcomes, all factors influencing these decisions, even those not limited to the patient, warrant close consideration. We examined the effect of emergency department (ED)–specific system factors, such as work load and census, on the use of noninvasive versus invasive respiratory support.

**METHODS:** We conducted a multi-center retrospective cohort study of all adult subjects with severe COVID-19 requiring an ICU admission from 5 EDs within a single urban health care system. Subject demographics, severity of illness, and the type of respiratory support used were obtained. Using continuous measures of ED census, boarding, and active management, we estimated ED work load for each subjects’ ED stay. The subjects were categorized by type(s) of respiratory support used: low-flow oxygen, noninvasive respiratory support (eg, noninvasive ventilation [NIV] and/or high-flow nasal cannula [HFNC]), invasive mechanical ventilation, or invasive mechanical ventilation after trial of NIV/HFNC. We used multivariable logistic regression to examine system factors associated with the type of respiratory support used in the ED.

**RESULTS:** A total of 634 subjects were included. Of these, 431 (70.0%) were managed on low-flow oxygen alone, 108 (17.0%) on NIV/HFNC, 54 (8.5%) on invasive mechanical ventilation directly, and 41 (6.5%) on NIV/HFNC prior to invasive mechanical ventilation in the ED. Higher severity of illness and underlying lung disease increased the odds of requiring invasive mechanical ventilation compared to low-flow oxygen (odds ratio 1.05 [95% CI 1.03–1.07] and odds ratio 3.47 [95% CI 1.37–8.78], respectively). Older age decreased odds of being on invasive mechanical ventilation compared to low-flow oxygen (odds ratio 0.96 [95% CI 0.94–0.99]). As ED work load increased, the odds for subjects to be managed initially with NIV/HFNC prior to invasive mechanical ventilation increased 6–8-fold.

**CONCLUSIONS:** High ED work load was associated with higher odds on HFNC/NIV prior to invasive mechanical ventilation.

*   respiratory support
*   noninvasive ventilation
*   system factors
*   ED workload

## Introduction

Over the past 2 decades, use of noninvasive respiratory support as first-line treatment for acute respiratory failure increased by approximately 50% in all acute care settings, especially within emergency departments (EDs).1-4 These therapies include noninvasive ventilation (NIV) and high-flow nasal cannula (HFNC). Whereas the evidence supporting the use of NIV for acute hypoxemic respiratory failure (AHRF) is mixed, HFNC has been shown to be beneficial for this condition,5,6 including specific use for COVID-19-related AHRF.7-11 More recently, other trials have shown that CPAP may reduce the risk of tracheal intubation and mortality over conventional oxygen, but HFNC did not show the same benefit.12 With general recommendations for use during surge or in limited-resource settings, these studies cite the benefits of noninvasive respiratory support in mitigating the need for high-intensity resources such as invasive mechanical ventilation and ICU admissions.7 Whereas these studies have explored how subject-related factors such as degree of hypoxemia, breathing frequency, and comorbidities can affect the choice of respiratory support,7-10 none has explored the influence of resource limitations and work load on the choice of respiratory support.

Many variables, unrelated to the patient, influence selection of respiratory support, including system variables (eg, site, census, resources) and provider variables (eg, training discipline, experience).13 As failure of noninvasive respiratory support requiring invasive mechanical ventilation initiation portends increased morbidity and mortality,14-21 it is important to identify possible reasons for specific respiratory support device selection to guide appropriate subject management. For this investigation, we aimed to examine how ED system factors affected the choice of respiratory support in the ED during the initial COVID-19 surge in New York City after controlling for subject-specific factors. In this multi-center retrospective cohort study of all ED subjects with severe COVID-19 infection, we first describe ED utilization of respiratory support during the initial surge of COVID-19 in New York City and then examine subject- and system-level predictors of respiratory support modality choice.

### QUICK LOOK

#### Current Knowledge

The use of noninvasive respiratory support such as noninvasive ventilation and high-flow nasal cannula has seen a dramatic increase in the past decades. This increase was accentuated by the recent COVID-19 pandemic and resource limitations. Whereas there is robust evidence supporting the use of noninvasive respiratory support for acute respiratory failure, the choice of these therapies might not always be influenced by patient factors alone.

#### What This Paper Contributes to Our Knowledge

In addition to subject factors, system factors such as ED work load influenced the choice of respiratory support in the ED during the COVID-19 pandemic. Given the known downstream implications of respiratory support on patient morbidity and mortality, all factors influencing these patient care decisions, even those not limited to the subject, warrant closer consideration when choosing respiratory support.

## Methods

### Study Setting and Population

The cohort included subjects from 5 EDs within the Mount Sinai Health System across New York City boroughs of Manhattan, Brooklyn, and Queens. Participating hospitals included both tertiary academic referral centers and community hospitals catering to the local population. The 5 EDs included in this study have a combined 2,646 hospital beds and see an annual ED volume of more than 400,000 subjects; this volume was greatly increased during the time of the COVID-19 surge between March–July 2020.

The subject cohort included all ED patients, age 18 y and older, who presented to a participating ED between March 1–July 10, 2020, with confirmed COVID-19 infection via polymerase chain reaction testing for SARS-CoV-2 within 48 h of arrival and required an ICU stay during their hospital course. Those patients with out-of-hospital cardiac arrest were excluded.

### Study Design, Measurements, and Outcomes

Clinical data from electronic health records, operational metrics, respiratory billing data, and hourly ED census data were extracted from the health system’s enterprise data warehouse. All electronic data capture was validated, further expanded with standardized manual abstraction, and entered into a secure online database by trained reviewers. All subjects were followed until death in the hospital or discharge.

Subject characteristics included demographics (eg, age, race, ethnicity, sex), baseline comorbidities, severity of illness calculation using the Mortality Probability Model III (MPM-III) scores on ED arrival,22 and any care limitations on arrival to the ED (such as do not intubate). Clinical course variables included the modality, duration, and settings of respiratory support; and adjuvant interventions such as steroids, anticoagulation, sedatives, diuresis, or vasopressors administered in the ED. The modalities of respiratory support included low-flow oxygen (nasal cannula or non–rebreather mask), HFNC, NIV (bi-level positive airway pressure or CPAP), and invasive mechanical ventilation. Additional clinical data such as ED, ICU, and hospital lengths of stay, as well as time and location of intubation, were collected.

The primary exposures being studied were system factors for ED work load and capacity strain.23,24 Measurements of total ED census and work load were obtained in 5-min intervals. Measurements to estimate ED work load included (1) number of ED subjects with high Emergency Severity Index (ESI)25 of 1 (resuscitation) or 2 (emergent), (2) number of active ED subjects undergoing initial workup and resuscitation, and (3) number of high-acuity ED subjects that required in-patient admission. These measurements were then expressed as a ratio of the total ED census in 5-min intervals. Then, these ratios were averaged across the duration of each subject’s ED stay to approximate ED-specific work load and resource utilization during the subjects stay in the ED.24 We then divided this averaged ED work load variables (ratio of ED work load/total ED census) into quartiles to test the role of increasing work load and resource utilization on the choice of respiratory support.

The primary outcome of interest was the type of respiratory support used in the ED and was categorized into 4 groups: (1) those on low-flow oxygen (< 10 L/min) only, (2) those receiving noninvasive respiratory support (NIV and/or HFNC [those receiving >30 L/min]), (3) those on invasive mechanical ventilation directly, or (4) those on invasive mechanical ventilation after a trial of noninvasive respiratory support.

This project was approved by the health system’s institutional review board under expedited review procedure, with a waiver of informed consent.

### Statistical Analysis

We conducted univariate and multivariate analyses to both describe the types of respiratory support used in the ED and the subjects that received them. Continuous variables were compared using Wilcoxon-Mann-Whitney U test or independent *t* tests as appropriate, and categorical variables were compared using chi-square analysis or Fisher exact test as appropriate. Intraclass correlation coefficients (ICCs) estimated variation between sites.

#### Predictors of the Type of Respiratory Support.

We used a multivariable, multivariate logistic regression model to determine predictors of the type of respiratory support used. A candidate list of covariates, including specific subject and system characteristics as well as those with significant multivariate associations (meeting a threshold of *P* < .1), informed covariate selection.

We selected parsimonious models based on Akaike information criteria and evaluated the effects of the covariates on the outcomes using a 2-tailed statistical test with significance of *P* value < .05. Model diagnostics, collinearity testing, and interactions between covariates of interest (age, site, severity of illness, and ED system factors) were also conducted. The data analysis for this paper was generated using SAS software, Version 9.4, SAS system for Windows (SAS Institute, Cary, North Carolina).

## Results

### Types of Respiratory Support Used in the ED

A total of 634 subjects presented with severe COVID-19 infections requiring an ICU stay during their hospitalization. A majority of these subjects (68.0%) only required low-flow oxygen while in the ED, whereas 17.0% was treated with noninvasive respiratory support, 8.5% was on invasive mechanical ventilation directly, and 6.5% was trialed on noninvasive respiratory support prior to being on invasive mechanical ventilation in the ED. There was no significant variation in the choice of respiratory support strategy among the 5 EDs or by which month they presented during the pandemic (ICC 0.02 [95% CI 0.01–0.67], *P* = .19). HFNC was used for 8.0% of the subjects for a median duration of 4 (interquartile range [IQR] 1–9) h. NIV was used for 17.8% for a median duration of 3 (IQR 1–7) h. Baseline characteristics (age, race, sex, body mass index) and comorbidities did not vary between various modalities of respiratory support (Table 1). Those who required invasive mechanical ventilation had a higher severity of illness upon ED arrival.

View this table:
[Table 1.](http://rc.rcjournal.com/content/67/9/1091/T1)

Table 1. 
Baseline Subject Characteristics and Outcomes of Different Respiratory Support Strategies in the Emergency Department

During the study period, the average median ED census for the 5 sites was 48.4 (IQR 39.5–57.7) subjects. As a percentage of the total ED census, the other ED work load measures were as follows: ED subjects with a high ESI of 1 or 2 were median 41.6% (IQR 33.2–51.2); active ED subjects undergoing initial workup and resuscitation were median 37.7% (IQR 20.7–48.7), and high-acuity ED subjects that required in-patient admission were median 62.3% (IQR 51.3–79.3). Table 2 details these ED system factors by modality of respiratory support used.

View this table:
[Table 2.](http://rc.rcjournal.com/content/67/9/1091/T2)

Table 2. 
Baseline Emergency Department System-Level Characteristics for Study Cohort

#### Predictors of the Respiratory Support Strategy.

Higher severity of illness on ED arrival (MPM-III) increased the odds of requiring invasive mechanical ventilation directly compared to low-flow oxygen (odds ratio 1.05 [95% CI 1.03–1.07]) (Table 3A). Having underlying lung disease such as COPD or asthma also increased the odds of requiring invasive mechanical ventilation directly compared to low-flow oxygen (odds ratio 3.47 [95% CI 1.37–8.78]). Higher severity of illness also increased the odds of invasive mechanical ventilation even after trial of noninvasive support compared to low-flow oxygen (odds ratio 1.03 [95% CI 1.02–1.05]). Older age decreased odds of being on invasive mechanical ventilation compared to low-flow oxygen (odds ratio 0.96 [95% CI 0.94–0.99]). There was no interaction between age and severity of illness, age and ED work load or site, or severity of illness and ED work load or site. The time (month or week) during which the subject presented after the pandemic onset was not significantly associated with a specific modality of respiratory support and was not included in the final model selection.

View this table:
[Table 3.](http://rc.rcjournal.com/content/67/9/1091/T3)

Table 3. 
Patient-Level Predictors of the Choice of Respiratory Support Strategies. Results From a Multivariable Logistic Regression Model Controlling for Patient and System-Level Predictors

The 4 variables measuring ED work load demonstrated collinearity, so the percentage of high-acuity ED subjects remained in the model. The percentage of high-acuity ED subjects remained in the model as the strongest individual predictor, and this metric likely encompassed a subject population contributing the highest burden on the ED staff and resources. In the adjusted model, compared to times of lower ED work load (quartile 1), during times of higher ED workloads (quartiles 3 and 4), the odds for subjects to be managed initially with noninvasive respiratory support prior to invasive mechanical ventilation increased 6–9-fold, respectively, (quartile 3 odds ratio 6.77 [95% CI 1.40–32.66], and quartile 4 odds ratio 8.94 [95% CI 1.72–46.54]) (Table 3B). Two of the 5 EDs were significantly more likely to use noninvasive support compared to low-flow oxygen (Appendix Table 1, see related supplementary materials at [http://www.rcjournal.com](http://www.rcjournal.com)). There was no interaction between site and ED work load.

## Discussion

We found that in times of increased ED work load subjects were significantly more likely to be trialed on noninvasive ventilation strategies prior to invasive mechanical ventilation. Building upon previous studies that have explored subject-level predictors of HFNC or NIV failure,8,10 we specifically explored ED-specific system factors as patients are first resuscitated here and are, therefore, more vulnerable to increased morbidity and mortality if care is delayed.23,24,26,27 At times of increased demand, few EDs have staffing and space flexibility to accommodate surges in critical care demand28; as such, our results likely reflect providers’ choice to trial a noninvasive strategy, which often requires less resources than invasive mechanical ventilation at times of high work load. During the initial surge of COVID-19 in New York City, unknown aerosolization risks and evolving practice guidelines may have excluded noninvasive treatment modalities in favor of early intubation for certain patients.29,30 Further, initially, conflicting recommendations regarding the use and benefit of NIV and HFNC for COVID-19 might have added to the uncertainty about which respiratory support modality was optimal and when to escalate to invasive modes.31-33 Whereas our data did not show that the month or week after the pandemic onset was significantly associated with respiratory support selection, they did show how system factors and resource limitations influenced choice of respiratory support for AHRF during the COVID-19 pandemic. Surge capacity of individual EDs, the ability to adequately monitor subjects on noninvasive respiratory support, and ventilator availability might have also played a role in selecting the mode of ventilation.

Our data also showed that even within the same health system the ED site where the subjects presented influenced the type of respiratory support they received, despite controlling for subject-related factors. However, we did not find any interaction between the ED site and the ED work load, suggesting that specific EDs are not more or less influenced by strain than others. The recent COVID-19 pandemic has highlighted a problem that previously existed but was significantly exacerbated by the rapid need for increased critical care capacity and the high-quality critical care outside the ICU.34,35 At times of surge, EDs and wards are unlikely to be adequately resourced for ongoing management of acute respiratory failure, which in turn affects resource utilization and subject outcomes.36,37 Given the known risks of increased morbidity and mortality for patients who fail of noninvasive respiratory support,14-20 and our data showing that factors beyond the subject influence support selection and management, it is important to consider and address these system factors that influence choice of respiratory support when developing clinical management guidelines or protocols. Institutional culture, such as ease and training with noninvasive support, as well as policy, such as limitations placed on the use of noninvasive support outside negative-pressure rooms, also likely played a role in device selection but was not accounted for in our analysis and warrants closer consideration. Further studies examining organizational factors, staff reallocation, and leveraging technology to guide care are needed to optimize respiratory care delivery.38

We also found that age and severity of illness predicted the types of respiratory support used in the ED. Whereas it is unsurprising that increasing severity of illness increased the likelihood of respiratory strategies that involved invasive mechanical ventilation in the ED, older age decreased the likelihood of invasive mechanical ventilation, despite older subjects generally having a higher severity of illness. Given that ventilators were a scarce resource during the height of the COVID-19 pandemic, this may reflect an effort to allocate invasive mechanical ventilation to those who are more likely to benefit. Further, given clinicians might be more inclined to have goals-of-care conversations with older individuals, they might also keep older patients on noninvasive respiratory support longer. During times of resource limitations, systems often adopt a crisis standard of care in lieu of usual care practices.39 Whereas several strategies were used to guide increasing surge capacities during the pandemic,40-44 rationing scare resources became a necessary practice.45 Fortunately, we did not find an interaction between age and ED work load when selecting a choice of respiratory support. Despite this and existing guidance on how care should be allocated,46,47 the impact of rationing during the pandemic on choice of respiratory support strategies, equity, and disparities needs further study.48

A major strength of our study was our capture of respiratory care delivery in the ED, specifically details of the duration, settings, and changes between modalities. In contrast to many studies examining crowding in the ED,49,50 we had continuous ED operations and census data streams allowing us to more accurately adjust for and evaluate the impact of ED strain and work load on respiratory support strategies. The inclusion of several sites, including community and tertiary centers, increases the generalizability of our findings. Our approach and inclusion of system factors in our modeling are informative to a more comprehensive understanding of respiratory care delivery.

Limitations of our study include its retrospective nature, which does not account for real-time variations in practice guidelines that were common during the pandemic. We tested our models for variation among respiratory support strategies by both site and by the month during which the subject presented. Whereas these models did not find significant variations in practice, we were not powered to assess this variation as some sites had a limited sample size of subjects on different support strategies. We tested all available subject and ED-specific operational factors in our model for predictors of respiratory support strategy; however, this quantitative modeling cannot completely capture the complex nature of real-time bedside decisions made by the provider and needs further study using mixed methods.13 Further, whereas we combined the use of noninvasive respiratory support, both HFNC and NIV, into one group for pragmatic reasons (patients were often switched between the modalities), it is important to note that the resources used and the burden on the health care staff are not the same for both HFNC and NIV as is a limitation of our study.

## Conclusions

In summary, both patient and ED system factors affected the choice of respiratory support strategy used for subjects with severe COVID-19 infections. Increasing ED work load significantly increased the odds of trialing noninvasive respiratory support prior to invasive mechanical ventilation in critically ill ED subjects. The impact of these decisions on subject outcomes needs further study.

## Acknowledgments

This work was supported in part through the computational and data resources and staff expertise provided by scientific computing at the Icahn School of Medicine at Mount Sinai.

## Footnotes

*   Correspondence: Neha N Goel MD MSCR, Icahn School of Medicine at Mount Sinai, One Gustave L. Levy Place, Box 1232, New York, NY 10029. E-mail: neha.goel{at}mountsinai.org
*   Dr Mathews has received study support from the National Institutes of Health National Heart, Lung, and Blood Institute (Awards: 1K23HL130648-PI: Mathews). The other authors have disclosed no conflicts of interest.

*   The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Heart, Lung, and Blood Institute or the National Institutes of Health.

*   Supplementary material related to this paper is available at [http://www.rcjournal.com](http://www.rcjournal.com).

*   Copyright © 2022 by Daedalus Enterprises

## References

1.  1.Stefan MS, Shieh MS, Pekow PS, Rothberg MB, Steingrub JS, Lagu T, et al. Epidemiology and outcomes of acute respiratory failure in the United States, 2001 to 2009: a national survey. J Hosp Med 2013;8(2):76-82.
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=23335231&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

2.  2.Schnell D, Timsit JF, Darmon M, Vesin A, Goldgran-Toledano D, Dumenil AS, et al. Noninvasive mechanical ventilation in acute respiratory failure: trends in use and outcomes. Intensive Care Med 2014;40(4):582-591.
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=24504643&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

3.  3.Demoule A, Girou E, Richard JC, Taille S, Brochard L. Increased use of noninvasive ventilation in French intensive care units. Intensive Care Med 2006;32(11):1747-1755.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1007/s00134-006-0229-z&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=16799775&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 
    
    [Web of Science](http://rc.rcjournal.com/lookup/external-ref?access_num=000241453400014&link_type=ISI) 

4.  4.Ozsancak Ugurlu A, Sidhom SS, Khodabandeh A, Ieong M, Mohr C, Lin DY, et al. Where is noninvasive ventilation actually delivered for acute respiratory failure? Lung 2015;193(5):779-788.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1007/s00408-015-9766-y&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=26210474&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

5.  5.Frat JP, Thille AW, Mercat A, Girault C, Ragot S, Perbet S, et al; REVA Network. High-flow oxygen through nasal cannula in acute hypoxemic respiratory failure. N Engl J Med 2015;372(23):2185-2196.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1056/NEJMoa1503326&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=25981908&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

6.  6.Rochwerg B, Granton D, Wang DX, Helviz Y, Einav S, Frat JP, et al. High-flow nasal cannula compared with conventional oxygen therapy for acute hypoxemic respiratory failure: a systematic review and meta-analysis. Intensive Care Med 2019;45(5):563-572.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1007/s00134-019-05658-2&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

7.  7.Calligaro GL, Lalla U, Audley G, Gina P, Miller MG, Mendelson M, et al. The utility of high-flow nasal oxygen for severe COVID-19 pneumonia in a resource-constrained setting: a multi-center prospective observational study. EClinicalMedicine 2020;28:100570.
    
    

8.  8.Zucman N, Mullaert J, Roux D, Roca O, Ricard JD; Contributors. Prediction of outcome of nasal high flow use during COVID-19-related acute hypoxemic respiratory failure. Intensive Care Med 2020;46(10):1924-1926.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1007/s00134-020-06177-1&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=32671470&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

9.  9.Demoule A, Vieillard Baron A, Darmon M, Beurton A, Geri G, Voiriot G, et al. High-flow nasal cannula in critically ill patients with severe COVID-19. Am J Respir Crit Care Med 2020;202(7):1039-1042.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1164/rccm.202005-2007LE&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

10. 10.Mellado-Artigas R, Mujica LE, Ruiz ML, Ferreyro BL, Angriman F, Arruti E, et al; COVID-19 Spanish ICU Network. Predictors of failure with high-flow nasal oxygen therapy in COVID-19 patients with acute respiratory failure: a multi-center observational study. J Intensive Care 2021;9(1):23.
    
    

11. 11.Ferreyro BL, Angriman F, Munshi L, Del Sorbo L, Ferguson ND, Rochwerg B, et al. Association of noninvasive oxygenation strategies with all-cause mortality in adults with acute hypoxemic respiratory failure: a systematic review and meta-analysis. JAMA 2020;324(1):57-67.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1001/jama.2020.9524&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

12. 12.Perkins GD, Ji C, Connolly BA, Couper K, Lall R, Baillie JK, et al; RECOVERY-RS Collaborators. Effect of noninvasive respiratory strategies on intubation or mortality among patients with acute hypoxemic respiratory failure and COVID-19: the RECOVERY-RS randomized clinical trial. JAMA 2022;327(6):546-558.
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

13. 13.Bauer PR, Kumbamu A, Wilson ME, Pannu JK, Egginton JS, Kashyap R, et al. Timing of intubation in acute respiratory failure associated with sepsis: a mixed-methods study. Mayo Clin Proc 2017;92(10):1502-1510.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1016/j.mayocp.2017.07.001&link_type=DOI) 

14. 14.Penuelas O, Esteban A. Noninvasive ventilation for acute respiratory failure: the next step is to know when to stop. Eur Respir J 2018;52(2):1801185.
    
    [Abstract/FREE Full Text](http://rc.rcjournal.com/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MzoiZXJqIjtzOjU6InJlc2lkIjtzOjEyOiI1Mi8yLzE4MDExODUiO3M6NDoiYXRvbSI7czoyNDoiL3Jlc3BjYXJlLzY3LzkvMTA5MS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

15. 15.Ozyilmaz E, Ugurlu AO, Nava S. Timing of noninvasive ventilation failure: causes, risk factors, and potential remedies. BMC Pulm Med 2014;14(1):19.
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=24520952&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

16. 16.Bellani G, Laffey JG, Pham T, Madotto F, Fan E, Brochard L, et al; ESICM Trials Group. Noninvasive ventilation of patients with acute respiratory distress syndrome. Insights from the LUNG SAFE study. Am J Respir Crit Care Med 2017;195(1):67-77.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1164/rccm.201606-1306OC&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=27753501&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

17. 17.Chandra D, Stamm JA, Taylor B, Ramos RM, Satterwhite L, Krishnan JA, et al. Outcomes of noninvasive ventilation for acute exacerbations of chronic obstructive pulmonary disease in the United States, 1998–2008. Am J Respir Crit Care Med 2012;185(2):152-159.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1164/rccm.201106-1094OC&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=22016446&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 
    
    [Web of Science](http://rc.rcjournal.com/lookup/external-ref?access_num=000299311500007&link_type=ISI) 

18. 18.Stefan MS, Nathanson BH, Higgins TL, Steingrub JS, Lagu T, Rothberg MB, et al. Comparative effectiveness of noninvasive and invasive ventilation in critically ill patients with acute exacerbation of chronic obstructive pulmonary disease. Crit Care Med 2015;43(7):1386-1394.
    
    

19. 19.Kangelaris KN, Ware LB, Wang CY, Janz DR, Zhuo H, Matthay MA, et al. Timing of intubation and clinical outcomes in adults with acute respiratory distress syndrome. Crit Care Med 2016;44(1):120-129.
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

20. 20.Kang BJ, Koh Y, Lim CM, Huh JW, Baek S, Han M, et al. Failure of high-flow nasal cannula therapy may delay intubation and increase mortality. Intensive Care Med 2015;41(4):623-632.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1007/s00134-015-3693-5&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=25691263&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

21. 21.Menga LS, Cese LD, Bongiovanni F, Lombardi G, Michi T, Luciani F, et al. High failure rate of noninvasive oxygenation strategies in critically ill subjects with acute hypoxemic respiratory failure due to COVID-19. Respir Care 2021;66(5):705-714.
    
    [Abstract/FREE Full Text](http://rc.rcjournal.com/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6ODoicmVzcGNhcmUiO3M6NToicmVzaWQiO3M6ODoiNjYvNS83MDUiO3M6NDoiYXRvbSI7czoyNDoiL3Jlc3BjYXJlLzY3LzkvMTA5MS5hdG9tIjt9czo4OiJmcmFnbWVudCI7czowOiIiO30=) 

22. 22.Higgins TL, Teres D, Copes WS, Nathanson BH, Stark M, Kramer AA. Assessing contemporary intensive care unit outcome: an updated Mortality Probability Admission Model (MPM0-III). Crit Care Med 2007;35(3):827-835.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1097/01.CCM.0000257337.63529.9F&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=17255863&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 
    
    [Web of Science](http://rc.rcjournal.com/lookup/external-ref?access_num=000244470800020&link_type=ISI) 

23. 23.Owyang CG, Kim JL, Loo G, Ranginwala S, Mathews KS. The effect of emergency department crowding on lung-protective ventilation utilization for critically ill patients. J Crit Care 2019;52:40-47.
    
    

24. 24.Goel NN, Durst MS, Vargas-Torres C, Richardson LD, Mathews KS. Predictors of delayed recognition of critical illness in emergency department patients and its effect on morbidity and mortality. J Intensive Care Med 2022;37(1):52-59.
    
    

25. 25.Tanabe P, Gilboy N, Travers DA. Emergency Severity Index version 4: clarifying common questions. J Emerg Nurs 2007;33(2):182-185.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1016/j.jen.2006.11.009&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=17379042&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

26. 26.Zhang Z, Bokhari F, Guo Y, Goyal H. Prolonged length of stay in the emergency department and increased risk of hospital mortality in patients with sepsis requiring ICU admission. Emerg Med J 2019;36(2):82-87.
    
    [Abstract/FREE Full Text](http://rc.rcjournal.com/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NzoiZW1lcm1lZCI7czo1OiJyZXNpZCI7czo3OiIzNi8yLzgyIjtzOjQ6ImF0b20iO3M6MjQ6Ii9yZXNwY2FyZS82Ny85LzEwOTEuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

27. 27.Jo S, Jeong T, Jin YH, Lee JB, Yoon J, Park B. ED crowding is associated with inpatient mortality among critically ill patients admitted via the ED: post hoc analysis from a retrospective study. Am J Emerg Med 2015;33(12):1725-1731.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1016/j.ajem.2015.08.004&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

28. 28.Carr BG, Addyson DK, Kahn JM. Variation in critical care beds per capita in the United States: implications for pandemic and disaster planning. JAMA 2010;303(14):1371-1372.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1001/jama.2010.394&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=20388892&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 
    
    [Web of Science](http://rc.rcjournal.com/lookup/external-ref?access_num=000276597300021&link_type=ISI) 

29. 29.Fowler RA, Guest CB, Lapinsky SE, Sibbald WJ, Louie M, Tang P, et al. Transmission of severe acute respiratory syndrome during intubation and mechanical ventilation. Am J Respir Crit Care Med 2004;169(11):1198-1202.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1164/rccm.200305-715OC&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=14990393&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 
    
    [Web of Science](http://rc.rcjournal.com/lookup/external-ref?access_num=000221677400008&link_type=ISI) 

30. 30.Cheung JC, Ho LT, Cheng JV, Cham EYK, Lam KN. Staff safety during emergency airway management for COVID-19 in Hong Kong. Lancet Respir Med 2020;8(4):e19.
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

31. 31.Cheung TM, Yam LY, So LK, Lau AC, Poon E, Kong BM, et al. Effectiveness of noninvasive positive pressure ventilation in the treatment of acute respiratory failure in severe acute respiratory syndrome. Chest 2004;126(3):845-850.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1378/chest.126.3.845&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=15364765&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 
    
    [Web of Science](http://rc.rcjournal.com/lookup/external-ref?access_num=000224148900032&link_type=ISI) 

32. 32.Alhazzani W, Moller MH, Arabi YM, Loeb M, Gong MN, Fan E, et al. Surviving sepsis campaign: guidelines on the management of critically ill adults with coronavirus disease 2019 (COVID-19). Crit Care Med 2020.
    
    

33. 33.Sa N-S. Respiratory support for patients with COVID-19 infection. Lancet Respir Med 2020;8(4)e18.
    
    

34. 34.Mitchell R, Banks C; authoring working party. Emergency departments and the COVID-19 pandemic: making the most of limited resources. Emerg Med J 2020;37(5):258-259.
    
    [FREE Full Text](http://rc.rcjournal.com/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiRlVMTCI7czoxMToiam91cm5hbENvZGUiO3M6NzoiZW1lcm1lZCI7czo1OiJyZXNpZCI7czo4OiIzNy81LzI1OCI7czo0OiJhdG9tIjtzOjI0OiIvcmVzcGNhcmUvNjcvOS8xMDkxLmF0b20iO31zOjg6ImZyYWdtZW50IjtzOjA6IiI7fQ==) 

35. 35.Mitchell R, Nou G. A “new normal”: harnessing the experience of COVID-19 for sustained improvements in global emergency care. Lancet Reg Health West Pac 2020;1:100012.
    
    

36. 36.Eriksson CO, Stoner RC, Eden KB, Newgard CD, Guise JM. The association between hospital capacity strain and in-patient outcomes in highly developed countries: a systematic review. J Gen Intern Med 2017;32(6):686-696.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1007/s11606-016-3936-3&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

37. 37.McQuillan P, Pilkington S, Allan A, Taylor B, Short A, Morgan G, et al. Confidential inquiry into quality of care before admission to intensive care. BMJ 1998;316(7148):1853-1858.
    
    [Abstract/FREE Full Text](http://rc.rcjournal.com/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6MzoiYm1qIjtzOjU6InJlc2lkIjtzOjEzOiIzMTYvNzE0OC8xODUzIjtzOjQ6ImF0b20iO3M6MjQ6Ii9yZXNwY2FyZS82Ny85LzEwOTEuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

38. 38.Zebrowski AM, Branas C, Sims S, Doorley R, Buckler D, Carr BG. 2 Right-Sizing Response: The Optimization of Critical Care Resources during COVID-19. Ann Emerg Med 2020;76(4).
    
    

39. 39.Committee on Guidance for Establishing Crisis Standards of Care for Use in Disaster Situations; Institute of Medicine. Crisis standards of care: a systems framework for catastrophic disaster response. Washington, District of Columbia: National Academies Press; 2012.
    
    

40. 40.Abir M, Nelson C, Chan EW, Al-Ibrahim H, Cutter C, Patel K, et al. Critical care surge response strategies for the 2020 COVID-19 outbreak in the United States: RAND Corporation; 2020.
    
    

41. 41.Halpern NA, Tan KS. United States resource availability for COVID-19. Society of Critical Care Medicine. Available at: [https://sccm.org/Blog/March-2020/United-States-Resource-Availability-for-COVID-19](https://sccm.org/Blog/March-2020/United-States-Resource-Availability-for-COVID-19). *Accessed October 23, 2020*.
    
    

42. 42.Quah LJJ, Tan BKK, Fua T-P, Wee CPJ, Lim CS, Nadarajan G, et al. Reorganizing the emergency department to manage the COVID-19 outbreak. Int J Emerg Med 2020;13(1):32.
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

43. 43.Mathews KS, Seitz KP, Vranas KC, Duggal A, Valley TS, Zhao B, et al. Variation in initial US hospital responses to the coronavirus disease 2019 pandemic. Crit Care Med 2021.
    
    

44. 44.Hickey S, Mathews KS, Siller J, Sueker J, Thakore M, Ravikumar D, et al. Rapid deployment of an emergency department-intensive care unit for the COVID-19 pandemic. Clin Exp Emerg Med 2020;7(4):319-325.
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

45. 45.Butler CR, Wong SPY, Wightman AG, O'Hare AM. US clinicians' experiences and perspectives on resource limitation and patient care during the COVID-19 pandemic. JAMA Netw Open 2020;3(11):e2027315.
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

46. 46.White DB, Lo B. A framework for rationing ventilators and critical care beds during the COVID-19 pandemic. JAMA 2020;323(18):1773-1774.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1001/jama.2020.5046&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

47. 47.Emanuel EJ, Persad G, Upshur R, Thome B, Parker M, Glickman A, et al. Fair allocation of scarce medical resources in the time of COVID-19. N Engl J Med 2020;382(21):2049-2055.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1056/NEJMsb2005114&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=32202722&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

48. 48.Lewis EG, Breckons M, Lee RP, Dotchin C, Walker R. Rationing care by frailty during the COVID-19 pandemic. Age Ageing 2021;50(1):7-10.
    
    

49. 49.Stang AS, Crotts J, Johnson DW, Hartling L, Guttmann A. Crowding measures associated with the quality of emergency department care: a systematic review. Acad Emerg Med 2015;22(6):643-656.
    
    [CrossRef](http://rc.rcjournal.com/lookup/external-ref?access_num=10.1111/acem.12682&link_type=DOI) 
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=25996053&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom) 

50. 50.Gaieski DF, Agarwal AK, Mikkelsen ME, Drumheller B, Cham Sante S, Shofer FS, et al. The impact of ED crowding on early interventions and mortality in patients with severe sepsis. Am J Emerg Med 2017;35(7):953-960.
    
    [PubMed](http://rc.rcjournal.com/lookup/external-ref?access_num=http://www.n&link_type=MED&atom=%2Frespcare%2F67%2F9%2F1091.atom)