Effects of Timing of Invasive Mechanical Ventilation in Patients with Shock. An Analysis of the Multicenter Prospective Observational VOLUME

Abstract

Objectives: Describe the variation in practice and identify predictors of invasive mechanical ventilation (IMV) use in shock. Explore the association between the timing of IMV initiation (“Early” vs. “Delayed”) on shock duration. Design: Multicenter, prospective, observational cohort study between September 2017 and February 2018 Setting: 34 hospitals in the United States and Jordan. Patients: Consecutive, adult, critically ill patients with shock, defined as a systolic blood pressure less than or equal to 90mm Hg, mean arterial pressure less than or equal to 65mm Hg, or need for a vasopressor medication. Interventions: None. Measurements and Main Results: “Early” IMV was defined as starting IMV 0–6 hours of shock onset and “Delayed” IMV was defined as starting IMV between 6 and 48 hours of shock onset. The primary outcome was shock–free days, defined as the number of days without shock after the first 48 hours of shock onset. Variation and predictors of IMV use were examined within the whole cohort as well as the subgroup of those intubated within 0–48 hours of shock onset. Mixed effects modeling with hospital site as a random effect showed that there was 7% variation by site in the use and timing of IMV in this shock cohort. In a propensity–matched model for the timing of IMV, “Early” IMV after shock onset was associated with more shock–free days when compared to “Delayed” IMV in those intubated within 0–48 hours of shock onset (Beta coefficient 0.65 days, 95% CI 0.14-1.16 days). Conclusions: Timing of IMV initiation for patients in shock has potentially important implications for patient outcomes and merits further study.

Keywords

shock invasive mechanical ventilation timing variation in practice

Introduction

Invasive mechanical ventilation (IMV) is a frequently utilized treatment for critically ill patients with respiratory distress and shock.¹ However, the decision to initiate IMV for these conditions varies considerably, affected not only by patient–related factors such as illness severity and trajectory of acute illness, but also by system and clinician–related factors.² Variation in ventilation strategies exists even with primary pulmonary disorders such as acute respiratory distress syndrome (ARDS), where delays in starting IMV have been associated with a worse prognosis despite a lower severity of illness at presentation.³ Within a cohort of patients with septic shock, later IMV initiation (greater than 12 hours after study entry) was associated with prolonged need for organ support.⁴ These studies examining timing of IMV initiation have similarly found that delays (ranging from 12 to 48 hours) are associated with worse outcomes.^1,3,4

For patients in shock, IMV may provide benefit. Regardless of etiology, shock is typically characterized by hypoperfusion or reduction in oxygen delivery to vital organs, leading to organ dysfunction. Organ dysfunction in shock usually occurs due to an imbalance between the oxygen needed by the organs and the oxygen delivered from the blood. Targeting the underlying cause of shock is the mainstay of the treatment. Using IMV to unload the muscles of breathing, thereby lowering their oxygen consumption, may also provide additional benefit.^5,6 In a small observational study of the effect of different ventilation strategies, support with assist-control reduced oxygen consumption by more than 20% compared with continuous positive airway pressure.⁷ The early IMV strategy may have the beneficial effect of directing the limited oxygen delivery to other vital organs during the shock state.

Though starting early IMV may reduce oxygen consumption by respiratory muscles and improve perfusion of other vital organs, it is associated with additional risks such as increased need for sedation or analgesia, delirium, ventilator–associated pneumonia, and chronic respiratory failure. The subset of patients in whom the benefits of early ventilatory support outweigh the risks is unknown but important to guide clinical practice. Using data from a multicenter observational cohort of patients with shock, we describe variations in practice and identify predictors of IMV use in shock. Then, we explore the association between the timing of IMV initiation on shock duration. We hypothesize that early initiation of IMV will have more shock free days.

Materials and Methods

Study Setting & Population

This is a secondary analysis of the Observation of variation in fluids administered and characterization of vasopressor requirements in shock (VOLUME–CHASERS) study,⁸ an observational, multicenter, prospective cohort of all consecutive adult patients (≥ 18 years) considered for ICU admission who had shock, defined as a systolic blood pressure < 90 mm Hg, mean arterial blood pressure < 65 mm Hg, or vasopressor initiation. The study sites included academic and community hospitals across the United States and one site in Amman, Jordan. This study was conducted over a 2– to 4–week duration between September 2017 and February 2018 through the Discovery Network, the Society of Critical Care Medicine's research network (ClinicalTrials.gov NCT03190408). Based on a previous observational study of 776 patients with septic shock, of whom 633 were intubated within 12 hours and 30 (4%) of whom had delayed intubation (greater than 12 hours),⁴ we estimated our target accrual in the VOLUMES-CHASER cohort, with a more conservative estimate of 50% IMV rate, will give us a cohort of approximately 480 patients to adequately power our study. This study was approved by the Institutional Review Board at all 34 participating hospitals and informed consent was waived by all sites.

For this secondary analysis, we included all patients with shock in the cohort except those patients identified to have shock onset in the operating room or post anesthesia care units, whose shock state resolved prior to IMV initiation, or those without any oxygenation or ventilation data. These patients were excluded as the timing of IMV was unclear, the outcome (shock) resolved prior to the exposure (IMV), or they had no ventilation and oxygenation data recorded.

Study Design & Measurements

The VOLUME–CHASERS data were collected for four distinct timeframes: a) the baseline or pre–shock period (data from 12 hours before the onset of shock); b) the peri–shock period (data from 12 hours before and after the onset of shock); c) the resuscitation period (data from the first 24 hours after the onset of shock, collected in intervals of 0-3, 3-6, 6-12, and 12-24 hours after shock onset); and d) post shock period (data collected daily for 7 days after the resuscitation period). See Supplemental Figure A2. All patients were followed until death in the hospital or discharge. ICU and hospital lengths of stay, need for new renal replacement therapy, and use of IMV were also collected.

Baseline patient characteristics included demographics, comorbidities, and medication history. Severity of illness was calculated using the Acute Physiology and Chronic Health Evaluation (APACHE) III⁹ and Sequential Organ Failure Assessment (SOFA)¹⁰ scores from 12 hours before to 12 hours after shock onset. Additional clinical data included shock etiology, location of shock onset, and ICU type. Treatment data included total fluid (crystalloid, colloid, and blood products) and vasopressor (norepinephrine, epinephrine, vasopressin, dopamine, and phenylephrine) administration, as well as the use of mechanical ventilation in all shock periods.

We defined a priori “early” IMV as IMV initiation within 0–6 hours of shock onset and “delayed” IMV as IMV initiation within 6–48 hours of shock onset. The primary outcome was shock–free days, defined as the number of days without shock after the first 48 hours of shock onset. Variation and predictors of IMV use were examined within the whole cohort as well as the subgroup of those intubated within 0–48 hours of shock onset.

Statistical Analysis

We conducted bivariate analyses for those with and without IMV use as well as those with “early” and “delayed” IMV initiation. We compared continuous variables between groups using Wilcoxon Mann Whitney U test and independent t–tests as appropriate and categorical variables between groups using Chi–square analysis.

Variation and predictors of IMV and “early” IMV Use

We used a multivariable, logistic, mixed effect regression model, with hospital site as a random effect, to determine predictors of IMV use and “early” IMV initiation in the subgroup of those on IMV within 0–48 hours of shock onset. Significant bivariate associations, (meeting a threshold of p <0.1), informed covariate selection. We also tested a candidate list of covariates, selected a priori based on clinical significance, including age, sex, race, APACHE III and SOFA score, shock etiology and location, and total amount of baseline fluid. Intraclass correlation coefficients estimated variation between sites.

Association between IMV timing and shock–free days

To examine the association between timing of IMV on shock–free days, we examined only those patients in the cohort initiated on IMV between 0–48 hours of shock onset. We measured shock–free days from 48 hours until 7 days after shock onset. Using the predictors of “early” versus “delayed” IMV from the previous analysis, we generated propensity scores for “early” IMV, and then matched patients 1:1 using a propensity score caliper of 0.1¹⁰ (details of propensity matching in Figure A3 and Table A1 in Appendix). These variables were most predictive for the need for “early” IMV and were therefore used to generate the propensity scores. For the propensity matched cohort, we used a multivariable, mixed–effect, linear regression model, with hospital site as the random effect to determine the association between IMV timing and shock–free days. We also conducted an a priori sensitivity analysis, of the association between IMV initiated between 0–3, 3–6, 6–12, 12–24, and 24–48 hours after shock onset and shock free days.

We selected parsimonious models based on Akaike information criteria (AIC) and evaluated the effects of the covariates on the outcomes using a two–tailed statistical test with significance of p–value less than 0.05. The data analysis for this paper was generated using SAS software, Version 9.4 of the SAS System for Windows (SAS Institute Inc., Cary, NC).

Results

Baseline Characteristics

Of the 1639 patients from the VOLUME–CHASERS cohort, 1603 (97.8%) were included in this secondary analysis (See Supplemental Figure A1). IMV was used in 967 (60.3%) of these patients, with almost half initiated within 0–48 hours of shock onset (n = 468, 48.4%). Of these intubated patients, 338 (72.2%) were initiated on IMV between 0–6 hours after shock onset (“early” group) and 130 (27.8%) between 6–48 hours after shock onset (“delayed” group).

Compared to patients who received IMV, patients who were never received IMV were older, had a lower severity of illness (APACHE–III), and more often developed shock outside the ICU (Appendix Table A1.) The primary etiology of shock was similar between both groups. In the IMV subgroup, “early” IMV patients had a high severity of illness in the peri–shock period and more often had liver disease or cirrhosis (Table 1.). The total number of days on the ventilator was similar between both “early” versus “delayed” groups (median [IQR], 3 [2-7] vs. 3 [1-8] days, p = 0.34). In this cohort, noninvasive positive pressure ventilation use varied from 0.5%–23.3% by site, and high flow nasal oxygen use varied 1.3%–14.5% by site. Of those patients in the “early” IMV group 68 (20.1%) were on noninvasive positive pressure ventilation (NIV) versus 36 (27.7%) in the late “IMV” group, and 12 (3.6%) were on high flow nasal oxygen (HFNO) in the early “IMV” versus 5 (3.8%) in the late “IMV group.

Table 1.

Baseline Cohort Characteristics of the Subgroup on IMV Between 0–48 Hours After Shock Onset by the “Early” (Within 6 Hours of Shock Onset) Versus “Late” (Between 6–48 Hours) IMV use After Shock Onset.

Variables	Full subgroup n = 468	“Early” IMV n = 338	“Late” IMV n = 130	p-value^a
Age, years, median (IQR)	64 (55–75)	65 (56–75)	63 (51–76)	0.42
Body Mass Index, median (IQR)	27.1 (22.7–32.0)	27.0 (22.7–31.0)	27.7 (23.0–35.0)	0.053
Race, n (%)
White	246 (52.6)	169 (50.0)	77 (59.2)	0.21
Black	89 (19.0)	64 (18.9)	25 (19.2)
Hispanic	68 (14.5)	54 (16.0)	14 (10.8)
Other	65 (13.9)	51 (15.1)	14 (10.8)
Sex, n (%)
Male	275 (58.8)	208 (61.5)	67 (51.5)	0.049
Female	193 (41.2)	130 (38.5)	63 (48.5)	0.049
Past Medical History
AIDS	8 (1.7)	4 (1.2)	4 (3.1)	0.23
Metastatic cancer	61 (13.0)	44 (13.0)	17 (13.1)	0.99
Lymphoma	16 (3.4)	8 (2.4)	8 (6.2)	0.08
Leukemia	35 (7.5)	21 (6.2)	14 (10.8)	0.09
Obstructive lung disease	61 (13.0)	45 (13.3)	16 (12.3)	0.77
Hepatic Failure	34 (7.3)	32 (9.5)	2 (1.5)	0.003
Cirrhosis without Failure	48 (10.3)	46 (13.6)	2 (1.5)	0.0001
Chronic kidney disease, not on dialysis	56 (12.0)	39 (11.5)	17 (13.1)	0.65
End Stage Renal Disease on dialysis	27 (5.8)	19 (5.6)	8 (6.2)	0.82
Congestive Heart Failure	58 (12.4)	36 (10.7)	22 (16.9)	0.07
Diabetes Mellitus	141 (30.1)	104 (30.8)	37 (28.4)	0.63
Hypertension	257 (54.9)	187 (55.3)	70 (53.9)	0.77
Primary etiology of shock, n (%)
Septic shock	323 (69.0)	227 (67.2)	96 (73.6)	0.43
Hypovolemic shock	71 (15.2)	552 (15.4)	19 (14.6)
Cardiogenic shock	47 (10.0)	37 (11.0)	10 (7.7
Other shock	27 [5.8]	22 [6.5]	5 [3.9]
ICU type, n [%]
Medical ICU	210 [50.1]	169 [50.2]	41 [50.0]	0.77
Medical/Surgical ICU	88 [21.0]	71 [21.1]	17 [20.7]
Surgical ICU	49 [11.7]	37 [11.0]	12 [14.6]
Other ICU [Cardiac/Cardiac Surgery/Neuro]	72 [17.2]	60 [17.8]	12 [14.6]
Shock onset locations, n [%]
ICU	171 [36.5]	118 [34.9]	53 [40.8]	0.25
Emergency department	203 [43.4]	152 [45.0]	51 [39.2]
Wards	82 [17.5]	57 [16.9]	25 [19.2]
Other	12 [2.6]	11 [3.3]	1 [0.8]
Severity of illness in the peri-shock period [12 hours before and 12 hour after shock onset]
Acute Physiology and Chronic Health Evaluation III, median [IQR]	98 [78–120]	100.5 [80–124]	90.5 [75–115]	0.01
Sequential Organ Failure Assessment, median [IQR]	10 [7–12]	10 [7–13]	7 [5–11]	<0.0001
Baseline fluid/hour prior to shock onset, mL/hour, median [IQR]^b	135.3 [53.8–340.0]	132.5 [58.3–334.4]	140.8 [50.5–403.3]	0.85

Abbreviations: IMV = invasive mechanical ventilation; IQR = Intraquartile range; AIDS = Acquired Immune Deficiency Syndrome; ICU = Intensive care unit; ml = milliliter

Variables was compared between the No IMV and IMV Use. Mann-Whitney rank sum test for continuous variables and Chi-square test for categorical variables were used.

Baseline fluid/hr was only avaiable in 1036 patients

Variation and predictors of IMV and “early” IMV Use

The variation in IMV use by hospital site was 6.9% (ICC, 0.069; 95% CI 0.04-0.18, p = 0.01). The adjusted prediction model for IMV use showed that age decreased the probability of IMV use (beta coefficient, −0.003; 95% CI −0.004 – −0.001), while male sex (beta coefficient, 0.08; 95% CI 0.03-0.14), APACHE III score (beta coefficient, 0.004; 95% CI 0.003 to 0.005), and total number of vasopressors used in the resuscitation period (beta coefficient, 0.11; 95% CI 0.07 to 0.14) increased the probability of IMV use. Compared to the Emergency Department, shock onset in the ICU and the wards increased the probability of IMV use (beta coefficient, 0.26; 95% CI 0.17 to 0.34, and beta coefficient, 0.12; 95% CI 0.02 to 0.22, respectively). (See Appendix Table A2 for complete model.)

The variation in the timing of IMV use (“early” vs. “delayed” IMV) by site was 7.2% (ICC 0.07, 95% CI 0.03-0.28, p = 0.042) (See Figure 1). Figure 1 shows the significant site to site variation in the percent of patients who were intubated after more than 6 hours of shock onset against the percent of patients with in-hospital mortality, weighted by the number of patients from each site. Higher SOFA scores from peri–shock period increased the odds of “early” IMV (OR 1.19, 95% CI 1.12-1.27), while comorbidities such as leukemia, congestive heart failure, and cirrhosis reduced the odds of “early” IMV (OR 0.35, 95% CI 0.16-0.78; OR 0.51, 95% CI 0.27-0.95; and OR 0.18, 95% CI 0.04-0.76, respectively) (Table 2).

Table 2.

Predictors of “Early” (Within 6 Hours of Shock Onset) Versus “Late” (Between 6–48 Hours) IMV use in the Subgroup of Patients Initiated on IMV Between 0–48 Hours After Shock Onset.

Covariate	OR^a	95% CI		p-value
Age	1.01	1.00	1.03	0.06
Body Mass Index	0.98	0.96	1.01	0.12
Female versus Male	0.71	0.46	1.11	0.13
SOFA Score from peri-shock period	1.19	1.12	1.27	<0.0001
Leukemia	0.35	0.16	0.78	0.01
Congestive Heart Failure	0.51	0.27	0.95	0.04
Cirrhosis	0.18	0.04	0.76	0.02

Abbreviations: IMV = invasive mechanical ventilation; OR = Odds ratio; 95% CI = 95% Confidence Interval; SOFA = Sequential Organ Failure Assessment

Multivariate logistic regression model adjusting for age, body mass index, sex, Sequential Organ Failure Assessment score in the peri-shock period, and comorbidities.

Association between timing of invasive mechanical ventilation on patient morbidity and mortality

In the propensity–matched model, the use of “early” IMV after shock onset was associated with more shock–free days when compared to “delayed” IMV in those intubated within 0–48 hours of shock onset (Beta coefficient 0.65 days, 95% CI 0.14-1.16 days) (Table 3). In an a priori sensitivity analysis, patients who started IMV 0–3, 3–6, and 12–24 hours all had more shock–free days than patients who started IMV 24–48 hours after shock onset (Appendix Table A3). When comparing between groups, those who started IMV 0–3 hours after shock onset had more shock–free days than those who started IMV 24–48 hours (Beta coefficient 1.1 days, 95% CI 0.012-2.10 days, p–value = 0.046, adjusted using Tukey–Kramer adjustment). In a multivariable mixed effects model for in-hospital mortality, with hospital site as a random effect, there was no association between “early” IMV and in-hospital mortality. (Beta estimate 0.29, 95% CI −0.45 – 1.02).

Table 3.

Association Between “Early” (Within 6 Hours of Shock Onset) IMV use and Shock-Free Days in the Subgroup of Patients Initiated on IMV Between 0–48 Hours After Shock Onset.

Covariate	Estimate^a	95% CI		p-value
“Early” versus “Late initiation of IMV	0.65	0.14	1.16	0.01
Race
Caucasian	Reference
Black	−0.28	−1.00	0.44	0.43
Hispanic	−0.70	−1.51	0.12	0.09
Other	0.16	−0.60	0.92	0.67
Shock Etiology
Septic	Reference
Cardiogenic	−0.64	−1.49	0.21	0.13
Hypovolemic	0.58	−0.16	1.31	0.12
Other	0.31	−0.77	1.39	0.57
Total number of vasopressors used in resuscitation period	−0.02	−0.25	0.21	0.87
Total amount of crystalloid (ml) used in resuscitation period	0.0014	−0.004	0.002	0.53

Abbreviations: IMV = invasive mechanical ventilation; 95% CI = 95% Confidence Interval; SOFA = Sequential Organ Failure Assessment

^aMultivariate linear mixed effects model adjusting for race, etiology of shock, and total number of vasopressors and total amount of crystalloid used in the resuscitation period. This analysis used a 1:1 propensity matched sample for likelihood for “Early” IMV use. Hospital site was used as a random effect.

Discussion

The decision to initiate IMV is complex and is often influenced by factors external to the individual patient; understanding all the potential benefits of early IMV initiation is an important consideration when making the decision.² Bedside providers have to weigh the potential benefits of IMV—decreased oxygen demand and controlled lung protective ventilation— against the potential harms. In this multicenter observational study, we found that early initiation of IMV was associated with increased number of shock–free days compared to delayed IMV initiation. While early IMV has been shown to reduce oxygen demand and shock in small physiologic studies,⁷ our study shows one potential benefit of early IMV initiation in a larger, more diverse patient cohort.

We did not find that “early” IMV reduced in–hospital mortality despite employing a propensity-matched sample to reduce confounding associated with the decision to initiate IMV. While our data suggest that “early” IMV use was associated with more shock–free days, this did not translate into a reduction in mortality risk. Our study cohort lacked sufficient detail to permit examining factors that influence decisions to start invasive ventilation, such as institution specific resource availability, utilization, and culture. Furthermore, the role and impact of IMV initiation for varying degrees of shock (shock treated with fluid of 1 vasopressor vs. refractory shock) is also unclear. Because many factors affect the risk for in-hospital mortality, we did not expect to find that 'early IMV' was associated with a reduced risk for dying in the hospital. This is also similar to other therapeutic studies, such as adjuvant steroids for septic shock, which have shown improvement in shock days without showing an improvement in mortality.^11–13 The association between the timing of intubation and long-term outcomes such as mortality and persistent organ dysfunction merits further study.

As with previous studies,² we also found significant variation in IMV use and timing in our cohort. There are likely many factors contributing to this, such as the use of noninvasive support with NIV and HFNO, and underlying comorbidities. We found a wide range of the use of noninvasive support in this shock cohort. Since noninvasive support is often been used to reduce the need for IMV, this wide variation in noninvasive use likely contributed to some of the variation in the timing of IMV.^14,15 Though noninvasive ventilation benefits patients with respiratory failure in the setting of COPD or cardiogenic pulmonary edema, its impact on oxygen demand and shock severity in patients with shock are unknown. Underlying comorbidities of the patients also contributed widely to the variation in IMV use as well as its timing. While more patients with underlying obstructive lung disease were on IMV in this cohort, fewer patients with congestive heart failure required IMV. In the subgroup of those initiated on IMV within 0–48 hours of shock onset, those with underlying leukemia, congestive heart failure, and cirrhosis were less likely to be on “early” IMV. Though we lack detailed data about noninvasive ventilation use and duration for this cohort, this association may reflect attempts to use of non–invasive ventilation, which has been shown to be beneficial in the immunocompromised oncological and congestive heart failure populations.

A strength of our analysis is our use of propensity matching. As with most observational studies, causal inferences made with this analysis is susceptible to confounding. We adjusted for several factors, including underlying demographics, comorbidities, severity of illness, and site to account for some of the confounding associated with the need for IMV. Our propensity–matched analysis further mitigated potential confounding. As IMV initiation is known to cause hypotension in some patients due to the loss of sympathetic vasoconstriction, or the use of sedative medications, we further limited the analysis of the association between IMV and shock–free days only to those on IMV after the onset of shock. To reduce the immortal time bias and determine the association between “early” versus “delayed” IMV, only patients intubated within 0–48 hours after shock onset were included in the primary analysis, and the primary outcome of shock–free days, not shock days, was measured only after 48 hours after shock onset.

A limitation of our analysis is that our data set was incomplete. Because this study was initially designed to assess the association between physiological assessment and vasopressor use on outcomes, some of the detail about IMV is lacking. However, we used the data in the intervals available and defined “Early” IMV as 0–6 hours after shock onset based on historical early goal directed therapy guidelines for sepsis resuscitation.¹⁶ Given the available data structure, we were only able to explore the data using this pre–specified cutoff for “early” IMV and were unable to explore timing of IMV as a continuous variable to identify what might be the best time to start IMV. Though our sensitivity analysis showed that those initiated on IMV 0–3 hours after shock onset had more shock–free days compared to those who started IMV 24–48 hours after shock onset, these data lacked the sample size to assess a specific time at which IMV initiation affected short–term (shock–free days) and long–term (mortality) outcomes. Although this study suggests some benefits associated with starting IMV early, further study is required to understand the impact of invasive mechanical ventilation in patients with shock more fully, including long–term patient outcomes and cost implications.

Conclusions

Our data show “early” IMV could be beneficial in increasing shock–free days. As initiation of IMV is a nuanced decision often happening quickly at the bedside of a critically ill patient, our analysis shows that shock may also be an important factor that needs to be considered when making this decision and needs further study.

Supplemental Material

sj-docx-1-jic-10.1177_08850666221081102 - Supplemental material for Effects of Timing of Invasive Mechanical Ventilation in Patients with Shock. An Analysis of the Multicenter Prospective Observational VOLUME–CHASERS Cohort

Supplemental material, sj-docx-1-jic-10.1177_08850666221081102 for Effects of Timing of Invasive Mechanical Ventilation in Patients with Shock. An Analysis of the Multicenter Prospective Observational VOLUME–CHASERS Cohort by Neha N. Goel, Jen–Ting Chen, Russel Roberts, Jonathan Sevransky, Michelle N. Gong, Kusum S. Mathews and in Journal of Intensive Care Medicine

Footnotes

Acknowledgment

[NNG]: reports no additional conflict of interest, other than the above funding sources; [JTC]: reports no conflict of interest; [RR]: reports no conflict of interest; [JS]: reports no conflict of interest; [MNG]: reports no conflict of interest; [KSM]: reports no additional conflict of interest, other than the above funding sources.

Funding

NNG has received study support from the NIH National Heart, Lung, and Blood Institute (Award: DHHS – 1T32 HL129974–PI: Richardson).

KSM has received study support from the NIH National Heart, Lung, and Blood Institute (Awards: 1K23HL130648–PI: Mathews).

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.

Ethical Approval

Not applicable, because this article does not contain any studies with human or animal subjects.

ORCID iD

Neha N. Goel

Supplemental Material

Supplemental material for this article is available online.

References

Bauer

Gajic

Nanchal

, et al. Association between timing of intubation and outcome in critically ill patients: a secondary analysis of the ICON audit. J Crit Care. Dec 2017;42:1-5. Epub 2017 Jun 16.

Bauer

Kumbamu

Wilson

, et al. Timing of intubation in acute respiratory failure associated with sepsis: a mixed methods study. Mayo Clin Proc. Oct 2017;92(10):1502-1510.

Kangelaris

Ware

Wang

, et al. Timing of intubation and clinical outcomes in adults with acute respiratory distress syndrome. Crit Care Med. Jan 2016;44(1):120-129.

Delbove

Darreau

Hamel

Asfar

Lerolle

. Impact of endotracheal intubation on septic shock outcome: a post hoc analysis of the SEPSISPAM trial. J Crit Care. Dec 2015;30(6):1174-1178.

Viires

Sillye

Aubier

Rassidakis

Roussos

. Regional blood flow distribution in dog during induced hypotension and low cardiac output. Spontaneous breathing versus artificial ventilation. J Clin Invest. Sep 1983;72(3):935-947.

Aubier

Trippenbach

Roussos

. Respiratory muscle fatigue during cardiogenic shock. J Appl Physiol Respir Environ Exerc Physiol. Aug 1981;51(2):499-508.

Manthous

Hall

Kushner

Schmidt

Russo

Wood

. The effect of mechanical ventilation on oxygen consumption in critically ill patients. Am J Respir Crit Care Med. Jan 1995;151(1):210-214.

Chen

Roberts

Fazzari

, et al. Variation in fluid and vasopressor use in shock with and without physiologic assessment: a multicenter observational study. Crit Care Med. Oct 2020;48(10):1436-1444.

Zimmerman

Wagner

Draper

Wright

Alzola

Knaus

. Evaluation of acute physiology and chronic health evaluation III predictions of hospital mortality in an independent database. Crit Care Med. Aug 1998;26(8):1317-1326.

10.

Austin

. An Introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. May 2011;46(3):399-424.

11.

Annane

Bellissant

Bollaert

, et al. Corticosteroids for treating sepsis in children and adults. Cochrane Database Syst Rev. Dec 6 2019;12 (12):CD002243.

12.

Venkatesh

Finfer

Cohen

, et al. Adjunctive glucocorticoid therapy in patients with septic shock. N Engl J Med. . Mar 1 2018;378(9):797-808.

13.

Sprung

Annane

Keh

, et al. Hydrocortisone therapy for patients with septic shock. N Engl J Med. . Jan 10 2008;358(2):111-124.

14.

National Hospital Ambulatory Medical Care Survey: 2011 Emergency Department Summary Tables. Ambulatory Health Care Data Centers for Disease Control and Prevention (CDC): The Ambulatory and Hospital Care Statistics Branch; 2011.

15.

Demoule

Girou

Richard

Taille

Brochard

. Increased use of noninvasive ventilation in French intensive care units. Intensive Care Med. Nov 2006;32(11):1747-1755.

16.

Rivers

Nguyen

Havstad

, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med. Nov 8 2001;345(19):1368-1377.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.22 MB