Abstract
Patients who require mechanical ventilation in the prehospital and emergency department environments experience high mortality and are at high risk of ventilator-associated ventilator-induced lung injury and ARDS. In addition, little attention has been given in the literature, trainee education, or clinical emphasis to ventilator management in these patients. ARDS and ventilator-induced lung injury are time-sensitive disease processes that develop early in mechanical ventilation and could potentially be prevented with early lung-protective ventilation. Prehospital and emergency department ventilation, in general, is characterized by potentially injurious tidal volume, high FIO2, and low PEEP. Recent literature highlights improved subjects outcomes in the setting of early lung-protective ventilation in both subjects with and those without ARDS. This review of the literature led us to recommend that lung-protective ventilation with avoidance of hyperoxia be the default goal ventilator strategy for all patients with prehospital and emergency department mechanical ventilation. This can be achieved by delivering low tidal volumes with stepwise, concurrent titration of FIO2 and PEEP to facilitate adequate oxygenation.
- mechanical ventilation
- prehospital
- emergency department
- ventilator-associated lung injury
- lung-protective ventilation
Introduction
Patients who require mechanical ventilation in the prehospital and emergency department environment experience high mortality and morbidity.1 Mechanical ventilation can lead to iatrogenic injury via ventilator-induced lung injury as well as hyperoxia. Phenotypically, this usually presents as worsening pulmonary mechanics, pneumonia, and/or ARDS, with the peak incidence occurring early in the course of mechanical ventilation (ie, day 1 or day 2).2,3 This suggests that appropriate management of mechanical ventilation immediately after endotracheal intubation is crucial, and emerging evidence has demonstrated a vital role for appropriate ventilator management in the prehospital and emergency department treatment of patients who are critically ill. Compared with the ICU or intraoperative environment, mechanical ventilation in the prehospital and emergency department setting has historically received very little attention in terms of research, trainee education, and clinical emphasis.4–6 As such, potentially injurious practice patterns are common.1,7,8
In this article, we discuss the landscape of mechanical ventilation and advances in scientific understanding in the care of patients with acute respiratory failure in the prehospital and emergency department settings. We provide recommendations for the provision of mechanical ventilation to patients in both of these environments. Although we recognize the importance of airway management and noninvasive positive-pressure ventilation, these topics are outside the scope of this review. In addition, many factors (ie, sedation, fluid administration, transfusions) play a role in the ultimate outcome of patients on mechanical ventilation in the emergency department; this review focused on the delivery of invasive mechanical ventilation.9,10
The Concept of Ventilator-Induced Lung Injury as a Time-Sensitive Emergency
Excessive stretch, regional lung overdistention, and repetitive airway opening all play roles in ventilator-induced lung injury and ARDS.11 Biologic mediators and hyperoxia can contribute to progressive pulmonary dysfunction, multiple organ failure, and death.12–14 In volutrauma, overdistention of alveoli results in damage to the intercellular junctions and the cellular membranes due to increased strain on pneumocytes.15 Similarly, barotrauma occurs when pneumocytes are damaged due to an increase in transalveolar pressure or stress. Cyclic recruitment–de-recruitment as alveoli collapse between respirations results in atelectrauma, which increases stress at any given pressure due to reduced compliance.16,17 These forces are especially important in ARDS, when the amount of lung tissue available for gas exchange is reduced, often referred to as a “baby lung.”17–19 Hyperoxia and the resultant reactive oxygen species formation are thought to cause injury both in the lung parenchyma and in sites distal to the pulmonary system.12,13 Time- and dose-dependent increases in inflammatory markers have been observed in experimental animals that received mechanical ventilation.20–22
Before knowledge of the existence of ventilator-induced lung injury, normalization of oxygenation and ventilation was given priority. As such, mechanical ventilation strategies involved the delivery of high tidal volume (VT) (12–15 mL/kg predicted body weight [PBW]) and low levels of PEEP.23 Subsequently, seminal work in the field provided strong evidence that the mechanical ventilator can cause harm, as animal and ex vivo models clearly demonstrated the mitigation of lung injury by the application of lung-protective concepts (ie, lower VT, more-appropriate PEEP).24–26 Perhaps most important to this review, on prehospital and emergency mechanical ventilation, these effects were observed over the course of only a few hours.24–26
Clinical studies that involved subjects without ARDS support this premise.4,27,28 Data that compared various strategies of lung-protective ventilation versus conventional ventilation during surgery showed that early application of lung protection for comparatively short durations can mitigate pulmonary and systemic inflammation and is associated with a reduction in pulmonary and extrapulmonary complications.29–31 Similarly, higher driving and plateau pressures among patients in the emergency department were associated with increased progression to ARDS. There was a dose-dependent, stepwise relationship with an increasing incidence of ARDS with increasing driving pressure, plateau pressure, compliance, and mechanical power.32
In subjects without ARDS and in the ICU, observational studies,2,33,34 a small randomized trial,27 and 2 systematic reviews4,28 showed an association between higher VT and increased incidence of ARDS, with a typical peak incidence around ICU day 2. This suggests that initial ventilator dosing influences downstream complications. In addition, in patients with ARDS, delayed delivery of lung-protective ventilation is associated with increased mortality.35
Multiple studies have shown that initial ventilator settings, both in the emergency department and the prehospital setting, influence ventilator settings that subjects received in the ICU. These settings remained unchanged in up to 75% of subjects through the first 24 h.1,2,33,36–38 Similar therapeutic momentum has been documented in other areas of critical care, such as sedation and antibiotic dosing, and may result in prolonged iatrogenic risk.9,39 Therefore, the most immediate period of care is a potential therapeutic target to increase adherence to best practice. Because mortality in patients in the emergency department and on mechanical ventilation can exceed 30%1 and can be as high as 50% if ARDS develops,37 early use of these practices may have the potential to have a large impact on patient outcome. In sum, ARDS and ventilator-induced lung injury occur early in patients on mechanical ventilation and are time-sensitive processes that benefit from early recognition and treatment.
Prehospital Mechanical Ventilation
The Landscape of Prehospital Ventilation
Prehospital mechanical ventilation is delivered during ∼4% of emergency medical service activations annually in the United States.40 Much of the data about prehospital care of patients who are critically ill comes from the realm of interfacility transport; however, even these data are limited to a few primarily observational studies. The majority of patients (73–83%) receive volume control continuous mandatory ventilation during transport, with a minority receiving pressure control continuous mandatory ventilation or volume control intermittent mandatory ventilation. In a cohort study of subjects with hypoxemic respiratory failure, the mean ± SD FIO2 was high during transport (0.95 ± 0.12), and mean ± SD PEEP was relatively low (9.6 ± 4.7 cm H2O).41 The investigators note, however, that FIO2 is often increased preemptively in the prehospital setting to levels higher than would be in the ICU to prevent critical desaturation.41 The VT values are often high, with one cohort study that showed that low VT ventilation occurred in only 14% of subjects on prehospital ventilation.42
For the majority of patients who require prehospital ventilation, a transport ventilator is not available and bag-valve-mask is frequently used to provide ventilation, even among patients transported on aeromedical units.42 Although, to our knowledge, there are no studies that compared patient-centered outcomes and the use of transport ventilators versus bag-valve-masks, one small randomized controlled trial (28 subjects) found that paramedics randomized to use transport ventilators believed that they were better able to perform patient care tasks than paramedics randomized to the use of a bag-valve-mask.43 In addition, the standard adult bag-valve-mask delivers high-volume, low-PEEP ventilation contrary to current recommendations for lung-protective ventilation.44 Manual ventilation with bag-valve-masks in simulated resuscitation scenarios has been shown to often deliver with high peak pressures, in certain scenarios that exceeded 100 cm H2O, even among experienced respiratory therapists.45 Previous reviews recommend the use of adjustable, disposable PEEP valves when providing ventilation via bag-valve-masks for preoxygenation before intubation.46 The impact of PEEP valves on patient-centered outcomes in bag-valve-mask ventilation has not been studied, and it is unclear how often they are used in practice.
Emergency medical services providers are trained to ventilate in a way that achieves observable chest rise, yet bag-valve-masks do not provide feedback on the delivered VT. Pneumotachograph devices are not routinely used or carried by emergency medical services providers. One simulation study demonstrated that the use of a standard pediatric bag-valve-mask resulted in a significantly greater proportion of VT in the 6–8 mL/kg PBW range than the use of adult bag-valve-masks (17.7% versus 5.1%).44 Median (interquartile range) VT delivered via endotracheal tube were also significantly greater when an adult bag-valve-mask (981.5 [901–1085] mL) was used compared with a pediatric bag-valve-mask (663 [615-696] mL).44 A separate simulation study demonstrated that gripping the bag-valve-mask with fewer fingers, in conjunction with pediatric bag size resulted in an even greater proportion of volumes being in a lung-protective range when compared with an adult bag-valve-mask alone (46.4% versus 0.4%).47 Analysis of these data indicated that injurious ventilation could occur in patients who receive ventilation via adult bag-valve-masks, although no patient-centered outcomes exist.
Complications Associated With Prehospital Mechanical Ventilation
Critical events (ie, major resuscitative procedure, hemodynamic deterioration, or inadvertent extubation) occur in as many as 1 in 20 aeromedical transports of patients who are critically ill, and the need for mechanical ventilation is independently associated with a 2- to 3-fold increase in risk of critical events during transport.48,49 Hypoxemic episodes during transport have been reported with relatively high frequency across studies that measured this end point (1.3–28%).48,50 Despite these risks, transfers of patients to facilities with higher levels of care are generally considered safe and deaths during transport are relatively rare, having occurred in 0.0–0.1% of transports described in the literature.41,48,51
Hypocapnia secondary to hyperventilation also occurs frequently during prehospital ventilation. This has been most commonly documented among patients with traumatic brain injury and occurred in up to 79% of patients.52,53 Prehospital hyperventilation and the resulting hypocapnia are associated with poor outcomes, including worsened mortality in multiple analyses.54–56 The major mechanism of this injury is believed to be decreased cerebral blood flow and vasoconstriction that causes ischemia in cerebral tissue.54–57 Use of prehospital quantitative end-tidal capnometry to avoid unintentional hypocapnia has been associated with a decreased incidence of hyperventilation.58
Clinical Impact of Prehospital Ventilatory Care
Changing transport practices have impacted the clinical course of prehospital patients. In particular, dedicated critical-care transportation teams can provide a similar level of care as an ICU. Patients treated by these teams experience fewer critical events than those treated by advanced life support paramedics.48 Across multiple studies, critical-care transport teams with training in complex ventilator management are associated with improved PaO2 after transfer from outside facilities in patients with hypoxic respiratory failure.41,50,51 Wilcox et al41 describe a cohort in which high rates of neuromuscular blockade were observed; 58 subjects (43.3%) received initial neuromuscular blockade from the critical care transport team. The transporting team changed ventilator settings during transport in 89% of the subjects, most commonly decreasing VT (35.9% of subjects), increasing PEEP (29.1%), and increasing FIO2 (30.1%).41 These changes were associated with increases in PaO2 on arrival at the receiving facility.41 Increasing FIO2 and PEEP, and administration of neuromuscular blockade were most strongly associated with increased PaO2 after transport. In addition, ventilator changes were associated with reduced peak inspiratory pressure and trended toward reduced plateau pressures. Prehospital mechanical ventilation management not only influences oxygenation and critical events during transport but may also carry downstream effects as well. VT provided by prehospital aeromedical crews have been shown to influence initial hospital VT, both in the emergency department and ICU.42
Emergency Department Ventilation
Landscape of Mechanical Ventilation Provided in the Emergency Department
Conservative estimates show that, in the United States, 250,000 patients are on mechanical ventilation in the emergency department annually.59 This rate is increasing,60,61 along with overcrowding and emergency department boarding of patients who are critically ill.62,63 A survey of emergency department directors revealed that >90% of emergency departments report problems with crowding and that daily crowding occurred in 39% of emergency departments, which resulted in delayed care and diagnosis in almost 40% of the patients.64 Crowding and prolonged boarding is associated with worsened mortality and prolonged mechanical ventilation duration.61,65–67 Increased duration of mechanical ventilation in the emergency department has been independently associated with increased mortality.67
Until recently, mechanical ventilation in the emergency department has received little attention in the literature outside of initial airway management.4 Survey studies showed that emergency physicians and trainees are often uncomfortable with ventilator management,5,6 and multiple studies showed that potentially injurious ventilation is commonly delivered in the emergency department.1,8,38 Volume control continuous mandatory ventilation is the most common mode of mechanical ventilation used in the emergency department (65–90% of patients).1,8,68 Analysis of the observational data from the emergency department showed that subjects received mean levels of PEEP of ∼5 cm H2O and high FIO2.1,8 A single-center study demonstrated median (interquartile range) emergency department VT to be 8.8 (7.8–10.0) mL/kg PBW and that lung-protective ventilation was delivered in only 27.1% of the subjects.1
Similar findings were observed in a multi-center study. Although a greater proportion of subjects in this cohort received lung-protective ventilation (55.7%), 11.4% still received VT of >10 mL/kg PBW.8 In a cohort at a different network of centers described by Wilcox et al,68 approximately half of the subjects received ventilation with both FIO2 of 1.0 and PEEP of ≤5 cm H2O, and nearly 40% of the subjects received nonprotective ventilation. The median FIO2 in this cohort was 1.0, and the median PEEP was 5 cm H2O. Patients often receive prolonged exposure to both high VT values (median VT = 230 [0-320] min) and high FIO2 (median VT = 251 [148-373] min) while in the emergency department.7 Initial ventilator settings remain static in up to 78% of subjects in the emergency department for the duration of ventilation, which suggested that the historical practice of ventilator management in the emergency department did not involve active titration of settings.1,67,68
Clinical Impact of Mechanical Ventilation in the Emergency Department
As demonstrated by several cohort studies, the historical approach to mechanical ventilation in the emergency department involved the following: (1) relatively high VT; (2) PEEP of 5 cm H2O; (3) FIO2 of 1.0; and (4) the delivery of mechanical ventilation in the supine, flat position.1,8,68 The LOV-ED (Lung-Protective Ventilation Initiated in the Emergency Department) trial was designed to target these practice patterns through a quality-improvement initiative with protocolized dosing of VT, PEEP, FIO2, and head of bed elevation for subjects in the emergency department who are on mechanical ventilation.36,69 This protocol was largely driven by respiratory therapists who measured accurate heights in all subjects to effectively implement lung-protective VT based on PBW and titrated FIO2 and PEEP to maintain adequate oxygenation.36,69 The protocol that was used clinically is displayed in Figure 1. This protocol was effectively implemented, with a significant increase in lung-protective ventilation from 48.2% in the pre-intervention cohort to 96.2% in the intervention cohort. Among subjects without ARDS, the post-intervention cohort received lower median dosing of FIO2 (median (IQR), 0.4 [0.4–0.6] versus 0.80 [0.5–1.0]) and VT (8.1 [7.3–9.1] mL/kg PBW versus 6.3 [6.0–6.7] mL/kg PBW).36
The LOV-ED protocol was associated with a reduction in mortality from 34.1% to 19.6% and a reduction in pulmonary complications (composite outcome of ARDS and ventilator-associated conditions) from 14.5% to 7.4%. In addition, ventilator, ICU, and hospital-free days were greater among the subjects who received protocolized lung-protective ventilation while in the emergency department, with mean differences of 3.7, 95% CI 2.3–5.1 d; 2.4, 95% CI 1.0–3.7 d; and 2.4, 1.2–3.6 d, respectively.36 Similar results were seen in the subjects with ARDS. Receipt of the emergency department–based lung-protective intervention was the only predictor of subjects with ARDS ever receiving lung protection in the ICU and was associated with a mortality reduction from 54.8% to 39.5% and with an increase in ventilator-free days from 7.7 to 11.6.37 Results of both studies are detailed in Table 1.
Although analysis of these data indicated that emergency department lung-protective ventilation is associated with improved patient outcome, it is unclear whether the observed clinical benefit was secondary to mitigation of ventilator-induced lung injury, reduced hyperoxia, or some therapeutic combination. Analysis of the clinical data showed that hyperoxia is associated with worse outcomes across a wide range of patients, including acute coronary syndrome and after cardiac arrest. However, these data had largely been limited to hyperoxia observed in the ICU.70,71 An a priori planned substudy7 of the LOV-ED trial demonstrated increased mortality in the subjects with hyperoxia (PaO2 >120 mm Hg) in the emergency department (29.7% vs 19.4%). There was a dose-dependent relationship between increasing ranges of hyperoxia and observed mortality as well (Fig. 2). Although clinicians recognize the negative impacts of hyperoxia, this typically is not reflected in their oxygen administration patterns.72,73 As such, hyperoxia is common in patients on mechanical ventilation in the ICU and the emergency department, and a possible target for improved outcomes.7,72
Summary and Recommendations
As demonstrated in this review, mechanical ventilation in the prehospital and emergency department settings (1) influences how the ventilator is managed after ICU admission, and (2) impacts patient outcome. Providers in these arenas, therefore, should strive to achieve the most-appropriate and safe ventilator settings on an individual patient level. Interfacility transfer of patients who are hypoxemic to higher levels of care (ie, a facility with extracorporeal membrane oxygenation capability) is feasible and safe, despite risks of deterioration and desaturation to the patient. In the prehospital environment, strong consideration should be given to avoidance of adult bag-valve-masks in all patients to avoid the dangers of hyperventilation and hypocapnia as well as the delivery of unnecessarily large VT values. When possible, a transport ventilator should be used or the adult bag-valve-mask should be replaced with a pediatric-sized bag-valve-mask to minimize delivery of VT that exceed lung-protective ventilation targets.44 Hypocapnia can be avoided by titration of ventilation based on end-tidal capnometry to reduce the risk of hyperventilation. Despite a paucity of patient-centered outcome data, we agree with previous recommendations to use a disposable, adjustable PEEP valve when providing bag-mask ventilation, titrating PEEP dosing to patient oxygen saturation.
For the vast majority of patients who receive mechanical ventilation in the prehospital or emergency department setting, we recommend that lung-protective VT (6–8 mL/kg PBW) be the default approach. It should also be noted that data from large academic medical centers demonstrated that ∼8% of patients on mechanical ventilation in the emergency department have acute lung injury.3,74 Therefore, dosing VT closer to the 6 mL/kg PBW end of the range as an initial approach may serve to improve the outcome in this cohort. To avoid hyperoxia, as opposed to the traditional approach of administering FIO2 of 1.0 at the initiation of mechanical ventilation, we recommend starting at 0.3–0.4 and only titrating up when needed, and in combination with PEEP. To streamline care, we recommend bundled ventilator protocols to help achieve implementation of best practices, and recommend a team approach, with heavy involvement from respiratory therapy. The lung-protective ventilation protocol used successfully in the LOV-ED study is displayed in Figure 1.
Although we recommend the effective implementation of protocols to reduce the unnecessary practice variability that surrounds postintubation mechanical ventilation, this does not replace the clinical decision making at the bedside with respect to dynamic ventilator adjustments. The implementation of a lung-protective ventilation protocol has proven safe and feasible, and is associated with improved outcome in patients on mechanical ventilation. However, it is not appropriate for all patients with acute respiratory failure (ie, life-threatening acidemia, expiratory flow limitation, and intrinsic PEEP [asthma, COPD]).75,76 Although high minute ventilation is often used to reduce PaCO2 transiently in the setting of acute brain herniation, maintenance of normocapnia is recommended in patients with brain injury.77
Protocols in patients with and without ARDS allow for set frequency adjustments up to 30–35 breaths/min36,78; In most patients, this does not result in clinically important increases in intrinsic PEEP and can typically maintain normocapnia.79 This method has been shown to be safe in patients with brain injury and gives providers the flexibility to titrate ventilation to achieve appropriate PaCO2 levels while providing lung-protective VT.80 Further, lung-protective ventilation in the setting of brain injury is well tolerated physiologically and is associated with improved outcomes.81 Therefore, the presence of brain injury should NOT preclude clinicians from attempting to use lung-protective ventilation. Finally, among the most important findings of this review was that there is a relative paucity of literature in the realm of prehospital and emergency ventilation. Given the importance of this topic, we believe that this is an area that is ripe for further study.
Footnotes
- Correspondence: Robert J Stephens, Division of Emergency Medicine, Washington University School of Medicine in St. Louis, 660 S Euclid Avenue, St. Louis, MO, 63110. email: stephensr{at}wustl.edu.
The authors have disclosed no conflicts of interest.
- Copyright © 2019 by Daedalus Enterprises
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.
- 19.↵
- 20.↵
- 21.
- 22.↵
- 23.↵
- 24.↵
- 25.
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵