Article, Pulmonology

Initial mechanical ventilator settings and lung protective ventilation in the ED

a b s t r a c t

Objective: Mechanical ventilation with low tidal volumes has been shown to improve outcomes for patients both with and without acute respiratory distress syndrome. This study aims to characterize Mechanically ventilated patients in the emergency department (ED), describe the initial ED ventilator settings, and assess for associations between Lung protective ventilation strategies in the ED and outcomes.

Methods: This was a multicenter, prospective, observational study of mechanical ventilation at 3 academic EDs. We defined lung protective ventilation as a tidal volume of less than or equal to 8 mL/kg of predicted body weight and compared outcomes for patients ventilated with lung protective vs non-lung protective ventilation, including inhospital mortality, ventilator days, intensive care unit length of stay, and hospital length of stay.

Results: Data from 433 patients were analyzed. Altered mental status without respiratory pathology was the most common reason for intubation, followed by trauma and respiratory failure. Two hundred sixty-one patients (60.3%) received lung protective ventilation, but most patients were ventilated with a low positive end- expiratory pressure, high fraction of inspired oxygen strategy. Patients were ventilated in the ED for a mean of 5 hours and 7 minutes but had few ventilator adjustments. Outcomes were not significantly different between patients receiving lung protective vs non-lung protective ventilation.

Conclusions: Nearly 40% of ED patients were ventilated with non-lung protective ventilation as well as with low positive end-expiratory pressure and high fraction of inspired oxygen. Despite a mean ED ventilation time of more than 5 hours, few patients had adjustments made to their ventilators.

Introduction

Intubation and initiation of mechanical ventilation are frequently required in the emergency department (ED). Ventilator-induced lung in- jury is a well-described phenomena and is associated with inappropriate ventilator settings [1-2], and inappropriate ventilation has been demon- strated to cause pathophysiologic and histologic changes in animals in as little as 20 minutes [3-5]. Conversely, appropriate and tailored

? Sources of support: This study was supported by institutional funds. At Rhode Island Hospital, the work was funded in part by an institutional development grant from the University Emergency Medicine Foundation.

?? Prior presentations: This work has not been previously presented or published.

* Corresponding author.

E-mail addresses: [email protected] (S.R. Wilcox), [email protected] (J.B. Richards), [email protected] (D.F. Fisher), [email protected] (J. Sankoff), [email protected] (T.A. Seigel).

use of the ventilator has been shown to improve mortality in patients with Acute respiratory distress syndrome [6-7] and may have implications for other patients requiring mechanical ventilation. A growing body of literature has shown that ventilating patients without ARDS in Intensive care units and Operating rooms with low tidal volumes is associated with decreased incidence of ARDS [8-14]. However, this has not yet been demonstrated in the ED [15-16].

Although emergency physicians are skilled in definitive airway management, their management of mechanical ventilation is not well described [17]. One prior multicenter, observational study described ED ventilation practices in a cohort of 219 patients with acute respiratory failure and found that 55.7% of patients received lung protective ventila- tion [15]. In this study, there was a wide distribution of set tidal volumes, ranging from 4.3 to 12.2 mL/kg predicted body weight. These results demonstrate heterogeneity in practice patterns by emergency physicians and an opportunity to more broadly characterize the use of mechanical ventilation in the ED.

http://dx.doi.org/10.1016/j.ajem.2016.04.027

This study characterizes patients who are intubated and mechanically ventilated in the ED and describe the ventilation practices of emergency physicians, including the initial ventilator settings selected after intuba- tion. In addition, clinical outcomes associated with lung protective vs non-lung protective ventilation were assessed by univariate and multi- variate analyses. The primary outcomes of interest were inhospital mortality, ventilator days, ICU length of stay, and hospital length of stay.

Methods

This was a multicenter, prospective, observational study of emergency physician practice regarding mechanical ventilation, conducted at 3 academic EDs in the United States, from July 2011 to March 2013. All 3 EDs are level 1 trauma centers with greater than 100 000 ED visits a year and are staffed by board-certified emergency physicians and emer- gency medicine (EM) residents. Ventilator management was determined by clinicians caring for the patients. No institution had a prespecified ventilation protocol at the time of this study, and no emergency physician underwent additional training beyond usual clinical practice.

All consecutive adult patients (N 18 years old) receiving invasive mechanical ventilation via an endotracheal tube or tracheostomy tube were eligible for enrollment. Patients exclusively receiving noninvasive ventilation were not enrolled. Exclusion criteria included death upon arrival or during the patient’s ED course and direct transfer to the operating room from the ED. This study was funded in part by a univer- sity development grant, and the study duration and sample size were determined by convenience sampling during the grant funding period. Patients were screened and enrolled upon presentation to the ED while receiving invasive mechanical ventilation or after intubation in the ED. After enrollment, all data were collected by research assistants who were blinded to study endpoints. Research assistants rounded hourly in the ED, and data were collected via clinician interviews and review of patients’ medical records, using a standardized data collection form. Demographic information, patients’ Clinical courses, the indica- tion for intubation, and primary diagnoses were recorded. Patients were classified by indication for intubation according to prespecified categories of altered mental status with no overt respiratory pathology, trauma, cardiac arrest, respiratory failure, and “other” indications. Subgroups were defined by the most common indications for intuba-

tion, including altered mental status, trauma, and respiratory failure.

Initial ventilator settings in the ED were recorded, including mode of ventilation, tidal volume, peak inspiratory pressure, Plateau pressure (if available), set inspiratory pressure for patients on pressure control ventilation, set respiratory rate, actual respiratory rate, positive end- expiratory pressure (PEEP), and fraction of inspired oxygen (FIO2). Pa- tients on volume control and pressure-release volume control were considered to be on volume-targeted ventilation, and patients on pressure control were considered pressure-targeted ventilation. Any ventilator changes made in the ED were also recorded.

Patients’ heights and actual body weights were measured and recorded. Predicted body weight was calculated as 50 kg+ 2.3 kg for each inch in height more than 5 f. for male patients and 45.5 kg+

2.3 kg for each inch more than 5 f. for female patients [6]. Tidal volume in milliliters per kilogram was calculated by dividing the tidal volume for both volume- and pressure-controlled settings by predicted body weight. Lung protective ventilation was defined as a tidal volume of 8 mL/kg or less. For patients who had a plateau pressure checked in the ED, if the value was more than 30 cm H₂O, even if the tidal volume was less than or equal to 8 mL/kg, the patient was classified as having non-lung protective ventilation.

In addition to initial ventilator settings, patient outcomes, including time of extubation, date of hospital discharge, date of ICU transfer, and inhospital mortality, were recorded.

Institutional review boards of each of the 3 institutions approved the protocols with waiver for informed consent.

Outcomes for patients initially treated with lung protective ventila- tion in the ED were compared to those who did not receive lung protec- tive ventilation. The presence or absence of lung protective ventilation was considered a categorical variable, whereas the clinical outcomes were either continuous variable (duration of ventilation and ICU or hospital length of stay) or categorical variables (mortality.) Missing or incomplete data were excluded on a case-by-case basis for all analyses. The results were visually inspected for normalcy of distribution, which was confirmed with Kolmogorov-Smirnov tests for the entire data set and subgroups. Means and SDs were determined, and 2-tailed t tests with unequal variances were used to compare continuous variables, and ?² analyses were used to assess for differences between lung protective ventilation strategies and mortality. Univariate and multi- variable regression analyses were performed to predict mortality- based clinical parameters, including Patient sex, indication for intu- bation, lung protective ventilation, and vasopressor use in the ED. Analysis of variance was used to determine Univariate associations between clinical parameters and mortality, and multivariable analyses were performed by multiple linear regression. All analyses were performed using JMP Pro 11.0 (Cary, NC).

Results

Five hundred thirty-five patients were enrolled. Of these, 102 did not have data recorded to allow calculation of tidal volume in milliliters per kilogram, as 93 patients had no height recorded, and 9 did not have a tidal volume recorded, leaving 433 patients to evaluate for lung protec- tive or non-lung protective ventilation. Altered mental status without respiratory pathology was the most common reason for intubation (40.4%), followed by trauma (22.9%) and respiratory failure (18.2%).

Of the 433 patients, 261 (60.3%) received lung protective ventilation, and 172 (39.7%) received non-lung protective (Fig. 1). Two patients met criteria for lung protective ventilation by milliliters per kilogram, with 7.47 and 7.53 mL/kg, but had plateau pressures of 33 and 32 cm H₂O, respectively, and were therefore considered to have received non-lung protective ventilation.

Patients’ demographic data, indications for intubation, and disposi- tion are reported in Table 1. Patients receiving lung protective ventilation were significantly more likely to be male. Patients who were intubated due to trauma and primary neurologic disorders were significantly more likely to receive lung protective ventilation, but there were no differences in the frequency of lung protective ventilation for all other indications for intubation. Data regarding duration of intubation were available for 157 patients receiving lung protective ventilation and 116 patients receiving non-lung protective ventilation. Of these patients, those intubated for less than 24 hours were significantly more likely to receive non-lung protective ventilation (Table 1).

Most patients, 80.4%, were ventilated with volume-targeted ventila- tion, with most receiving assist control and volume control ventilation. The mean tidal volume was 552.2 mL (SD, +-120.0 mL), which resulted in a mean of 7.97 mL/kg (SD, +-2.09). The histogram of tidal volume distribution is shown in Fig. 2. The mean peak inspiratory pressure for all patients was 25.76 (SD, +-7.93) cm H₂O but was higher in patients with volume-targeted ventilation at 26.31 (SD, +- 8.09) cm H₂O and lower in patients on pressure-targeted at 23.49 (SD, +-6.81).

Plateau pressures were available for 253 (72.7%) of the 348 patients on volume-targeted ventilation. The mean plateau pressure was

17.9 cm H₂O. The median PEEP was 5 cm H₂O, with a mean of 5.4 cm H₂O. The median FIO2 was 1.0, with a mean of 0.87. Two hundred four- teen patients (49.4) were ventilated for the duration of their ED time with a PEEP of 5 or less and FIO2 of 100%. Fifty patients (11.5%) had the identical ventilator settings of volume control with a tidal volume of 500 mL/kg, PEEP of 5 cm H₂O, and FIO2 of 1.0.

Sixty-three percent of patients were intubated in the ED, and 37% were intubated in the field before arriving to the ED. The mean duration of time intubated in the ED was 5 hours and 7 minutes, with an SD of

Fig. 1. Patient distribution.

6 hours and 3 minutes. Of the 433 patients in this cohort, 337 (77.8%) had no ventilator changes made in the ED. Only 12 patients were initially on non-lung-protective ventilation and changed to lung protective in the ED, whereas 2 were on lung protective and changed to non-lung protec- tive. The 2 most common changes in ventilator settings were changes in the respiratory rate for 56 patients and changes in FIO2 for 46 patients. Positive end-expiratory pressure was only adjusted for 5 patients, increased in 4 cases, and decreased to 0 cm H₂O in one case.

Patients initially treated with a lung protective strategy received significantly lower tidal volume in milliliters per kilogram than patients not placed on lung protective ventilation for all patients (6.96 +- 0.74 vs 9.60 +- 2.49 cm H₂O; Pb .0001.) Tidal volumes were also significantly lower for patients receiving lung protective vs non-lung protective ventilation for all of the subgroups (Table 2). Plateau pressures were lower in the lung protective group, although both lung protective and non-lung protective plateau pressures were well less than 30 cm H₂O. Patients who were intubated because of respiratory failure had no sig- nificant differences in any ventilatory parameters whether they were in the lung protective or non-lung protective groups, beyond the

defining characteristics of tidal volume and milliliters per kilogram. Patients intubated for altered mental status who were in the lung protective ventilation group had significantly lower peak inspiratory pressures, plateau pressures, and minute ventilation as compared to pa- tients intubated for altered mental status in the non-lung protective group (Table 2).

Clinical outcomes were not significantly different between patients receiving lung protective vs non-lung protective ventilation, including all subgroups defined by indication for intubation (Tables 3 and 4).

By univariate analysis, there was an association between mortality and female sex, receiving lung protective ventilation, respiratory failure as the indication for intubation, and receiving vasopressors in the ED (Table 5). Multiple linear regression, however, demonstrated only vaso- pressor use in the ED as a significant predictor of mortality (F(4428) = 7.07; Pb .0001); female sex, respiratory failure as the indication for intubation, and the use of lung protective ventilation were not indepen- dently predictive of mortality in multivariable analyses.

Discussion

Table 1

Demographics

Lung protective, n (%)

Non-lung P

protective, n (%)

The objectives of this study were to characterize initial ventilator settings in the ED and to determine if initial ED ventilator settings were associated with the clinically relevant patient outcomes of inhospital mortality, ventilator days, ICU length of stay, and hospital length of stay. Numerous studies have shown that ventilation with

Male patients 209 (80.1) 83 (48.3) b.0001

Age in years 56.3 (SD, 19.1) 53.5 (SD, 18.5) .132

Location of intubation .207

ED	102 (39.1)	57 (33.1)
Before ED	159 (60.9)	115 (66.9)
Indication for intubation Altered mental status, no respiratory	104 (39.8)	71 (41.3)	.767
pathology
Respiratory failure	46 (17.6)	33 (19.2)	.683
Trauma	61 (23.4)	38 (22.1)	b.0001
Intracranial hemorrhage, stroke,	18 (6.9)	14 (8.1)	.003
seizures
Cardiac arrest	15 (5.7)	11 (6.4)	.784
Airway protection for edema or	7 (2.7)	1 (0.6)	.071
obstruction
Other	10 (3.8)	4 (2.3)	.363
Vasopressors in the ED 38 (14.6)		16 (9.3)	.738
Disposition Medical ICU	128 (49.0)	100 (58.1)	.063
Surgical/trauma ICU	68 (26.1)	36 (20.9)	.234
Neuro-ICU	44 (16.9)	22 (12.8)	.852
Cardiovascular ICU	17 (6.5)	12 (7.0)	.216
Not reported Duration of intubation b24 h	4 (1.5) 63 (40.1)	2 (1.2) 62 (53.4)	.741 .029
b48 h	91 (57.9)	78 (67.2)	.117

large tidal volumes and higher pressures in patients without ARDS contribute to the development of ARDS and other pulmonary complica- tions [8-14]. However, although strongly suspected to be clinically applicable to ventilator management and clinical outcomes in the ED, this phenomenon has not yet been demonstrated in ED patients [15-16]. One previous retrospective review of patients ventilated in the ED who later developed ARDS found that many of those patients were ventilated with larger than recommended tidal volumes [18], but the study was not designed to show an association between tidal volumes and outcomes. In our cohort of unselected patients receiving mechanical ventilation in the ED, lung protective ventilation was not associated with improvement in clinical outcomes.

In our study, lung protective ventilation was not defined solely base

on tidal volume. For patients with a set tidal volume of 8 mL/kg or less, if the plateau pressure was checked and found to be more than 30 cm H₂O, they were classified as having been placed on non-lung protective ventilation. This distinction is consistent with descriptions of lung protective ventilation in other studies and emphasizes the importance of maintaining both low distending pressures and low tidal volumes in mechanically ventilated patients [6-12].

Given the observational design of this study, confounding factors may have affected our findings, as the 2 groups were not evenly

Fig. 2. Distribution of tidal volumes in cc/kg, with a range from 4 to greater than 14.

matched. Patients who received non-lung protective ventilation were more likely to be extubated within the first 24 hours of admission, indi- cating lower baseline acuity in this group. Most patients in this study were intubated for altered mental status without respiratory failure, and this group with minimal pulmonary disease demonstrated no significant association between ventilator settings and clinical out- comes. However, univariate analyses demonstrate that receiving lung protective ventilation was associated with increased mortality. This supports the premise that higher acuity patients with respiratory failure may have received increased attention to their ventilator settings in the ED with lung protective ventilation. This possibility is further supported by the lack of difference between peak inspiratory pressures and plateau pressures for the lung protective and non-lung protective groups in patients intubated for respiratory failure and trauma. Although the tidal volumes were significantly lower in the lung protec- tive group, the set inspiratory pressures and plateau pressures were not, suggesting that the 2 groups were not evenly matched with regard to underlying Respiratory physiology. Therefore, as the findings of this observational study are counter to those of randomized controlled trials, we suspect that the outcomes of this study are more reflective of the underlying patient conditions rather than the effects of mechanical ventilation in the ED.

In addition to looking for differences between patients ventilated with lung protective or non-lung protective ventilation, we categorized ventilator settings and management in the ED. Based on initial ED ventilator settings, patients in our cohort overall had good respiratory system compliance, with a mean plateau pressure of only 18 cm H₂O, well below the goal plateau pressure of 30 cm H₂O for lung protective ventilation [6]. Although lung injury has been demonstrated to begin

within minutes to hours [3-5], with the low pressures seen in these pa- tients, non-lung protective ventilation may not have caused significant injury. Notably, even patients ventilated with non-lung protective ventilation in our study did not have tidal volumes as large as the control group in the original ARDSNet study, with only 9.6 mL/kg as compared to 12 mL/kg [6].

Many of the patients in our study had very similar ventilator settings, despite a wide range of heights and differing indications for intubation. Notably, in this cohort, most patients ventilated with lung protective ventilation were male. Previous work has demonstrated that female sex is associated with receiving larger than recommended tidal volumes [8,19-20], and our results are consistent with these findings. The observation that 80% of patients receiving lung protective ventilation were male may indicate a “one size fits all” approach to selecting tidal volumes without appreciating the need to reduce volumes for female patients, even if they are the same height as male patients. In the univariate analysis, female sex was strongly associated with an increased mortality. However, when controlled for other factors, including lung protective ventilation, this association was not present in the multivariable analysis.

Although we found significant differences for ventilator parameters of tidal volume and minute ventilation between the 2 groups, there were no significant differences in PEEP or FIO2. Patients in the ED were ventilated with relatively low PEEP, including those patients intubated primarily for respiratory failure. Although this may reflect the relative ease of oxygenating patients in his cohort, titrating PEEP to promote recruitment, reduce atelectrauma, and decrease FIO2 is a preferred ventilation strategy [8,15,21]. Emergency department providers have demonstrated knowledge of the importance of low tidal volumes in

Table 2

Comparison of ventilator parameters for lung protective and non-lung protective ventilation

	Lung protective,	Non-lung protective,	P
	(n +- SD)	(n +- SD)
mL/kg
All patients	7.94 +- 0.72	9.54 +- 2.47	b.0001
Altered mental status	7.05 +- 0.69	9.56 +- 2.1	b.0001
Respiratory failure	6.58 +- 0.78	9.07 +- 1.59	b.0001
Trauma	7.06 +- 0.70	9.28 +- 1.17	b.0001
Tidal volume (mL)
All patients	482 +- 71	552 +- 120	b.0001
Altered mental status	486 +- 72	547 +- 82	b.0001
Respiratory failure	447 +- 68	509 +- 87	.001
Trauma	510 +- 64	566 +- 71	.0001
Peak inspiratory pressure (cm H₂O)
All patients	25.1 +- 8.0	26.7 +- 7.8	.054
Altered mental status	23.5 +- 6.4	26.8 +- 8.1	.004
Respiratory failure	29.3 +- 10.8	28.9 +- 8.3	.881
Trauma	25.0 +- 7.4	24.8 +- 6.5	.874

Plateau pressure (cm H₂O)

All patients	17.5 +- 4.9	18.6 +- 5.0	.063
Altered mental status	16.2 +- 4.1	17.9 +- 4.5	.048
Respiratory failure	20.7 +- 5.7	20.4 +- 6.2	.878
Trauma	15.8 +- 3.5	17.4 +- 3.7	.112
Minute ventilation (L/min) All patients	9.1 +- 2.6	10.8 +- 3.9	b.0001
Altered mental status	8.9 +- 2.7	10.1 +- 2.8	.005
Respiratory failure	9.4 +- 3.0	9.8 +- 3.1	.610
Trauma	9.3 +- 2.1	11.0 +- 3.3	.007
PEEP (cm H₂O) All patients	5.5 +- 2.1	5.1 +- 1.6	.033
Altered mental status	5.2 +- 1.1	4.9 +- 1.5	.093
Respiratory failure	6.9 +- 3.2	6.0 +- 2.0	.121
Trauma	4.6 +- 1.8	4.9 +- 1.0	.217
FIO2 All patients	0.87 +- 0.22	0.86 +- 0.22	.513
Altered mental status	0.83 +- 0.25	0.84 +- 0.23	.731
Respiratory failure	0.91 +- 0.19	0.86 +- 0.21	.252
Trauma	0.91 +- 0.19	0.83 +- 0.25	.081

ARDS [16], but the results of our study raise the question of whether emergency physicians are as familiar with recommendations for management of PEEP and FIO2.

A key finding in this study is that ventilated patients were boarded in

the ED for more than 5 hours, a similar duration of ED length of stay reported in prior studies [22]. Ventilated patients are resource intensive and are at high risk for deterioration [23-24]. Despite prolonged ED ventilation durations, only 22.2% patients had any ventilator changes made in the ED, and most of those changes were to the respiratory rate and FIO2. Of the patients initially on non-lung protective ventilation, only 7% were changed to lung protective settings during their ED course. These data suggest that once ventilator settings are selected, adjust- ments to the ventilator are made only rarely. A 2013 study of 251 me- chanically ventilated ED patients with severe sepsis or septic shock had a similar finding, as 69.3% of patients did not have any ventilator changes made in ED [16]. Given that patients’ conditions often change within the first few hours of intubation and mechanically ventilated patients in the ED are at high risk for deterioration [25-26], these data

Table 3

Outcomes for Lung Protective vs. Non-Lung Protective Ventilation

Table 4

Outcomes by subgroup analysis

	Lung protective (n +- SD)	Non-lung protective (n +- SD)	P
Altered mental status Patients in each group	104	71
Mortality	26.2%	21.7%	.500
Mechanical ventilation duration (d)	4.66 +- 5.88	3.94 +- 4.06	.569
ICU length of stay (d)	6.93 +- 6.71	5.04 +- 4.47	.142
Hospital length of stay (d)	16.26 +- 29.57	10.72 +- 10.70	.124
Trauma Patients in each group	61	38
Mortality	15.5%	22.2%	.433
Mechanical ventilation duration (d)	4.10 +- 5.57	4.34 +- 3.72	.877
ICU length of stay (d)	5.69 +- 6.97	7.27 +- 6.03	.457
Hospital length of stay (d)	11.06 +- 11.28	12.16 +- 10.95	.661
respiratory failure patients in each group	46	33
Mortality	37.8%	35.5%	.841
Mechanical ventilation duration (d)	3.26 +- 2.90	3.84 +- 3.75	.605
ICU length of stay (d)	5.61 +- 4.69	10.36 +- 10.51	.159
Hospital length of stay (d)	13.35 +- 17.28	12.37 +- 15.46	.819

demonstrate an opportunity for emergency physicians to be more involved in the management of mechanical Ventilation parameters.

In a study published in 2015, Fuller et al [15] evaluated mechanical ventilation in the ED and found that 55.7% of patients in their cohort received lung protective ventilation, somewhat less than the 65% in our study. Although more patients in their study had ventilator changes made in the ED, at 68.5%, they also found a low frequency of titrating PEEP and high FIO2 administration. Our study supports the findings of this prior work, verifying that many patients continue to be ventilated with non-lung protective ventilation in the ED. In addition, emergency physicians in our study also made few ventilator changes in the ED, especially to PEEP, while favoring high FIO2.

This current study also found that most patients intubated in the ED

were intubated for reasons other than respiratory failure, consistent with other studies of mechanical ventilation in the ED [27]. If most patients in the ED are not intubated for primary respiratory failure and therefore are not difficult to oxygenate or ventilate, this may lead to less emphasis being placed on ventilation management in the ED. Patients’ acute presentations may be falsely reassuring with regard to their respiratory status and function, and careful adherence to low tidal volume strategies or making changes to ventilator settings may not be prioritized in patients intubated for altered mental status or airway protection. This observation identifies a potential missed oppor- tunity for optimized ventilator settings for all patients intubated in the ED and highlights an opportunity for increased education and emphasis on evidence-based ventilation strategies for all patients intubated in the ED, regardless of the indication for intubation.

The primary limitations arise from the observational nature of this study, as confounding factors may have significantly impacted the results, and the effect of confounders cannot be determined based on the available data. Furthermore, as an observational study, only correla- tive associations can be made and causal relationships cannot be deter- mined. In addition, despite being a prospective study, not all patients had heights recorded, thereby limiting the ability to calculate a predicted

Table 5

Lung

Not lung P

Univariate analyses

protective protective

	ventilation (n +- SD)	ventilation (n +- SD)		Parameter	df (between, within)	F	P
duration of mechanical ventilation (d)	4.20 +- 5.00	3.74 +- 3.67	.486	Female sex	1, 864	151.48	b.0001
ICU length of stay (d)	6.31 +- 6.71	6.29 +- 6.39	.983	Respiratory failure as indication for intubation	1, 864	724.53	b.0001
Hospital length of stay (d)	14.11 +- 22.95	12.50 +- 17.38	.467	Lung protective ventilation	1, 864	105.01	b.0001
Inhospital mortality	28.7%	29.7%	.825	Vasopressor use in the ED	1, 864	17.3	b.0001

body weight and calculate milliliters per kilogram of tidal volume. However, this lack of height documentation likely reflects current practices in EDs and demonstrates an opportunity for improvement. In addition, nearly 40% of patients were intubated before arrival in the ED. The ventilation provided by EMS clinicians was not recorded, and patients may have received hyperventilation before enrollment. Lastly, although 12 patients initially had non-lung protective ventilation and were changed to lung protective in the ED, we do not have the durations for each of these settings and, therefore, evaluated the initial ED settings with respect to outcomes. Given the small number of patients with changes to lung protection, a significant change in the outcomes is unlikely to be seen. Finally, the etiologies of respiratory failure were recorded in broad categories and the effects of more specific diagnoses on outcomes cannot be determined.

Conclusions

In this observational cohort study of a heterogeneous group of patients mechanically ventilated in the ED, a substantial number were ventilated with non-lung protective ventilation. The patients who received lung protective ventilation were less likely to be extubated within 24 hours of ED presentation, indicating significant heterogeneity between the 2 groups. Therefore, although lung protective ventilation was not associated with improved clinical outcomes, the effects of con- founding factors cannot be known. Few patients had changes made to their ventilators while in the ED, despite a mean ED ventilation time of more than 5 hours. Emergency medicine providers favored using high FIO2 with low PEEP for mechanically ventilated patients, and PEEP was rarely titrated in the ED. Future studies of mechanical ventilation in the ED should include large randomized controlled trials of lung protective ventilation in the emergency setting for patients at risk for progression to ARDS.

References

AJEM