a b s t r a c t

Background: prehospital airway management increasingly involves supraglottic airway insertion and a paucity of data evaluates outcomes in trauma populations. We aim to describe definitive airway management in traumat- ically injured patients who necessitated prehospital supraglottic airway insertion.

Methods: We performed a single institution retrospective review of multisystem injured patients (>=15 years) that received prehospital supraglottic airway insertion during 2009 to 2016. Baseline demographics, number and type of: supraglottic airway insertion attempts, definitive airway and complications were recorded. Primary outcome was need for tracheostomy. Univariate and multivariable statistics were performed.

Results: 56 patients met inclusion criteria and were reviewed, 78% were male. Median age [IQR] was 36 [24-56] years. Injuries comprised blunt (94%), penetrating (4%) and burns (2%). Median ISS was 26 [22-41]. Median number of Prehospital endotracheal intubation (PETI) attempts was 2 [1-3]. Definitive airway management in- cluded: (n = 20, 36%, tracheostomy), (n = 10, 18%, direct laryngoscopy), (n = 6, 11%, bougie), (n = 9, 15%, Glidescope), (n = 11, 20%, bronchoscopic assistance). 24-hour mortality was 41%. Increasing number of PETI was associated with increasing facial injury. On regression, increasing cervical and facial injury patterns as well as number of PETI were associated with definitive airway control via surgical tracheostomy.

Conclusions: After supraglottic airway insertion, operative or non-operative approaches can be utilized to obtain a definitive airway. Patients with increased craniofacial injuries have an increased risk for airway complications and need for tracheostomy. We used these factors to generate an evidence based algorithm that requires pro- spective validation.

Level of evidence: Level IV - Retrospective study.

Study type: Retrospective single institution study.

(C) 2017

Introduction

A functional and patent airway during prehospital resuscitation is a critical consideration of trauma resuscitation [1]. Several risk factors confound prehospital airway control such as obesity and craniofacial trauma [2]. Supraglottic devices may be utilized for the difficult airway

? Author contribution: Study design was developed by Matthew Hernandez, Johnathon Aho, David Morris, and Martin Zielinski. Data generation was performed by Matthew Hernandez. Data analysis and interpretation was performed by Matthew Hernandez, Johnathon Aho, Martin Zielinski, Scott Zietlow, Brian Kim, and David Morris. Manuscript writing was performed by Matthew Hernandez, Johnathon Aho, Scott Zietlow, Brian Kim, David Morris and Martin Zielinski.

* Corresponding author at: Division of Trauma Critical Care and General Surgery, Department of Surgery, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA.

E-mail addresses: [email protected] (M.C. Hernandez), [email protected] (J.M. Aho), [email protected] (M.D. Zielinski), [email protected] (B.D. Kim), [email protected] (D.S. Morris).

[1,3-5]. After supraglottic device insertion, methods to secure a defini- tive airway include direct laryngoscopy, blind tube exchange or fiber- optic guidance [6]. In cardiac arrest, supraglottic airway exchange may not be urgent as the primary focus is restoration of spontaneous circula- tion [7]. Conversely, the trauma resuscitation focuses on constant air- way assessment to gauge patency and adequate ventilation. This is the unique difference; maintenance of airway control and prevention of dysoxia while systematically triaging injury care by severity whereas medical resuscitations aim to restore and maintain cardiac flow.

In the prehospital setting, supraglottic devices provide initial airway control with ease of insertion [8-10]. These advancements come at the expense of potential complications. Morbidity such as gastric disten- sion, tube malposition and oropharyngeal edema resulting abrupt air- way occlusion can occur [5,11-15]. Management algorithms exist to secure a definitive airway after supraglottic device insertion; however, these recommendations are from variable populations [16-19]. For trau- ma patients necessitating prehospital Supraglottic airways, there is a

http://dx.doi.org/10.1016/j.ajem.2017.09.028

0735-6757/(C) 2017

hospital outcomes and definitive”>lack of evidence to adequately address definitive airway management. Therefore, we aimed to determine which definitive airway techniques were utilized after prehospital supraglottic device insertion in multisys- tem injured patients hypothesizing that patients with increased prehospital airway complications and craniofacial injury patterns would require advanced airway control, including surgical tracheostomy.

Methods

Patient cohort

From 2009 to 2016, we performed a single center retrospective study examining patients N 15 years old with multisystem trauma de- fined as an Injury Severity Score of >= 9 that necessitated prehospital in- sertion of a supraglottic airway (patients receiving only a King Airway Device, King LT-D, Noblesville, IN). Patients were identified from the Mayo Clinic Trauma Center registry. Institutional review board approval was obtained prior to data review. Patients that refused research con- sent, received prehospital endotracheal tube intubation (PETI), were pregnant, or without multisystem trauma were excluded.

Prehospital airway institutional protocol

Patients were transported by a critical care trained rotor wing team or ground transport. Injured patients that require advanced prehospital airway management meet criteria for our highest level trauma activa- tion, which includes Emergency Medicine, Surgery, and Anesthesia providers to be present at patient arrival. Each prehospital airway inter- vention at our facility is reviewed in detail by the directors of medical transportation, Emergency Medicine, Trauma Surgery, and Anesthesia divisions. A prehospital advanced airway control algorithm (Fig. 1) has been defined and implemented by this group to standardize difficult air- way management in the prehospital setting. This algorithm is designed for use after clearly defined “failure” of standard prehospital endotra- cheal intubation (PETI) attempts and after non-invasive ventilation is determined to be inadequate.

Outcomes and predictors

The primary outcome for this study was need for tracheostomy. If a tracheostomy was not performed and instead an endotracheal tube ex- change (ETT) was performed, the method of ETT was recorded (direct laryngoscopy, bougie, Glidescope, bronchoscopic assistance). Patient demographics, transportation method and duration, traumatic mecha- nism, trauma severity (ISS and Abbreviated Injury Scores (AIS)), admis- sion vital signs (heart rate, respiratory rate, systolic and diastolic blood pressure and oxygen saturation), Glasgow Coma Score , 24 h and overall mortality, frequency and type of prehospital airway compli- cations, and number of PETI, and durations of intensive care, mechanical ventilation or overall hospital stay were abstracted from the electronic record. Mortality was defined according to definitions reported previ- ously [20].

Statistical analyses

Summary statistical and univariate analyses were performed. Con- tinuous variables were described using means with standard deviations (SD) if normally distributed and medians with inter-quartile ranges [IQR] for non-normally distributed data and two tailed t-tests were per- formed between definitive airway techniques, endotracheal tube ex- change (ETT) versus surgical tracheostomy. Categorical variables were summarized as proportions, and differences were evaluated using chi- square analysis. All p-values were considered significant at p b 0.05. Clinical and statistically significant variables were evaluated to assess for risk factors for 24-h mortality using nominal logistic regression

with 95% confidence intervals (CI). Data was analyzed with JMP (SAS In- stitute, Inc. Cary NC). We utilized GraphPad Prism (GraphPad Software, Inc. La Jolla CA) for all visual graphics.

Results

Baseline demographics

The study cohort consisted of 56 patients with multisystem trauma and supraglottic airway insertion. The median [IQR] ISS was 26 [22- 41] and 78% of patients were male. The mean (+-SD) age was 39.6 (+-21.2) years. Most patients (72%) were transported via rotor wing and median [IQR] transport time was 20 [13 - 33] minutes. Mean (+-SD) body mass index (BMI) was 29.2 (+-6.6). The median [IQR] head, neck and facial abbreviated injury scores (AIS) were: head 4 [3- 5], neck 2 [0-3], face 1 [0-2] respectively, Table 1. Mechanisms of injury included blunt (n = 53, 94%), penetrating (n = 2, 4%), and burn (n = 1, 2%).

Prehospital airway characteristics

In the prehospital setting, supraglottic device indications included failed PETI (n = 56, 100%). The median [IQR] attempts at PETI were 2 [2-3]. The number of failed PETI attempts increased in patients with in- creased craniofacial injury patterns Fig. 2. At arrival, all patients had a patent and functional airway provided by the supraglottic device. Dur- ing prehospital resuscitation, there were 35 (63%) complications includ- ing significant laryngeal or oropharyngeal edema preventing PETI (n = 22, 63%) and supraglottic airway dislodgement (n = 13, 37%).

In hospital outcomes and definitive airway management

Techniques for in hospital definitive airway included ETT (n = 36; 64%) or surgical tracheostomy (n = 20; 36%). For patients managed with ETT, 50% (n = 18) were performed in the emergency room and 50% (n = 18) were performed in the operating room. Table 2 compares outcomes and secondary predictors by definitive airway and this dem- onstrates the association of increased craniofacial injury patterns with a need for surgical tracheostomy. In patients who required definitive airway with a surgical tracheostomy, compared to ETT, there was an in- creased median facial AIS (4 [3-4] versus 1 [0-2], p b 0.0001). There was no statistically significant difference in median head AIS (4 [2-5] versus 4 [2-5]) injury severity but there was an approach to statistical signifi- cance and likely clinical significance in patients with increased median cervical AIS (2 [0-3] versus 1 [0-2], p = 0.08). Multivariable analysis demonstrated that the following factors were independently associated with need for surgical tracheostomy compared to ETT in patients with a prehospital supraglottic rescue airway: Facial AIS >= 3, cervical AIS >= 3, and number of PETI attempts, Table 3.

During definitive airway management (open tracheostomy or endo- tracheal tube exchange (ETT)), the median [IQR] oxygen saturation nadir was significantly lower in patients that received ETT compared to open tracheostomy, (84% [75-89] versus 92% [88-94], p = 0.007). This difference disappeared within 10 min of definitive airway manage- ment completion (99 [96-99] versus 99 [96-100], p = 0.8). There were no long-term complications from surgical tracheostomy or ETT during follow up, median 13 [1-37] months.

There were 23 patients that expired. Causes for mortality included myocardial infarction (n = 3), pulmonary contusion (n = 4), and ten- sion pneumothorax (n = 3), traumatic brain injury (n = 5), and hem- orrhagic shock (n = 8). No deaths were related to inpatient airway complications. There was less overall mortality in those receiving tra- cheostomy compared to those undergoing ETT (n = 8, (24%) versus n

= 27, (77%), p = 0.01). With respect to 24-hour mortality, a more pro- nounced difference existed between patients undergoing tracheostomy compared to ETT (n = 3, (13%) versus n = 20, (87%), p = 0.004).

Fig. 1. Institutional clinic prehospital transport airway management algorithm (after first failed attempt).

Discussion

This analysis underscores several challenges, prehospital airway complications, and important clinical outcomes in patients receiving

supraglottic airway insertion. We demonstrate that after prehospital supraglottic airway insertion in polytrauma, definitive airway manage- ment via surgical tracheostomy was associated with severe craniofacial injury patterns and multiple PETI attempts. Moreover, we

Table 1

Patient characteristics comparing definitive airway technique

Characteristic	Overall (n = 56) No. (%)	Endotracheal tube exchange (ETT) (n = 36) No. (%)	Tracheostomy (n = 20) No. (%)	p value
Baseline demographics Age, median [IQR]	36 [24-56]	33 [21-55]	41 [26-59]	0.6
Male	44 (79)	16 (80)	28 (78)	0.7
BMI, median [IQR]	29 [24-33]	31 [2-34]	28 [24-33]	0.6
Prehospital metrics prehospital transport time, median [IQR]	25 [15-33]	25 [18-39]	21 [12 -32]	0.2
Patient physiology Heart rate, median [IQR]	110 [96-116]	112 [101-117]	110 [81-116]	0.3
Respiratory rate, median [IQR]	22 [17-26]	24 [20-27]	21 [16-37]	0.1
Oxygen saturation, median [IQR]	96 [87-99]	87 [66-94]	84 [68-96]	0.8
Systolic blood pressure, median [IQR]	104 [80-123]	106 [72-128]	103 [84-122]	0.7
Diastolic blood pressure, median [IQR]	60 [50-80]	64 [50-81]	61 [50-80]	0.8
Predictors ISS, median [IQR]	26 [22-41]	35 21-43]	26 [22-38]	0.3

BMI: body mass index, ISS: injury severity score, IQR: interquartile range, ETT: endotracheal tube exchange.

demonstrated that those who received surgical tracheostomy as a defin- itive airway had minimal oxygen saturation fluctuation compared to those with endotracheal tube exchange (ETT). This analysis provides preliminary data for the role of adjunct airways, granular complication data, and initial implications for definitive airway management in adult trauma patients.

Airway assessment and control is a primary objective during trauma resuscitation. For supraglottic airways in trauma, minimal evidence ex- ists to guide definitive airway management (16). Subramanian et al. re- ported that for patients presenting with supraglottic airways, few had associated trauma, and in those patients the definitive airway was man- aged using a surgical tracheostomy in the majority (5). The present study is comprised entirely of multisystem trauma patients. Our analy- sis provides a trauma only cohort that analyzes subsequent definitive airway management. Patients with multisystem trauma pose a signifi- cant challenge for airway control. In the prehospital setting, two priori- ties are of equal value 1) minimization of airway Iatrogenic injury and 2) rapid transport to definitive care. Supraglottic devices, such as the King LT(TM), are one method to provide a temporary airway. Since these de- vices are increasingly utilized, a comprehensive analysis of the indica- tions, complications, and subsequent definitive airway management has not been adequately provided. We demonstrated that for patients with increasing craniofacial trauma, definitive airway management after supraglottic airway insertion is associated with tracheostomy compared to ETT. This finding echoes other work which has advocated for the use of tracheostomy in the management of definitive airways

craniofacial injury patterns with surgical tracheostomy justifies our pro- posed definitive airway algorithm, as shown in Fig. 3. Patients who meet criteria would have, at the minimum, definitive airway evaluation in the operative room as opposed to trauma resuscitation bay. Application of these criteria may prevent premature attempts to evaluate patients with potentially compromised airways in less controlled environments such as a trauma resuscitation bay. This algorithm needs to be prospec- tively validated, ideally in a multi-center trial, given the overall small number of patients any one center would see in a given period of time. We found that the decision to perform surgical tracheostomy com- pared to endotracheal tube exchange (ETT) is difficult in trauma pa- tients. Craniofacial injury patterns made prehospital airway control and subsequently definitive airway management more difficult. Routine tube exchange may not be warranted or safe in trauma patients. This may be due to multiple PETI attempts or supraglottic airway insertion, which may inflict oropharyngeal trauma and can complicate definitive airway management. Our multivariable analysis demonstrated that the number of PETI attempts and increased craniofacial trauma patterns were independently associated with definitive airway managed with surgical tracheostomy compared to ETT. These data highlight that care- ful selection of patients necessitating surgical tracheostomy may be

Table 2

Patient and injury characteristics, by type of definitive airway, values otherwise reported as medians with interquartile range

after supraglottic airway device insertion [16].

Definitive airway management requires rapid assessment and an ac- tionable plan amidst the multiple priorities that occur in trauma resus- citation. The association of increasing number of PETI attempts and

Outcome Tracheostomy N= 20

Endotracheal tube exchange (ETT)

N= 36

Overall mortality rate	40%	75%	0.02
24 hour mortality rate	15%	56%	0.004
Post definitive airway	11.1%	17%	1
complication rate
Facial AIS	4 [3-4]	0 [0-2]	0.001
Head AIS	2 [1-3]	0 [0-1]	0.003
Cervical AIS	4 [2-5]	4 [3-5]	0.6
Blunt trauma rate	90%	97%	0.4
Supraglottic airway dwell	104 [87-138]	61 [46-80]	0.001
time (minutes)
# of PETI failed attempts	2 [2-3]	1 [0-2]	0.001
ISS	35 [21-43]	26 [23-38]	0.4
Prehospital airway	95%	44%	0.0001
complication rate
Duration of stay (days)	2 [1-7]	14 [4-25]	0.002
Duration of mechanical	2 [1-4]	6 [1 -13]	0.01
ventilation (days)

p-Value

Fig. 2. Increased facial trauma, (AIS grade), associated with increased prehospital endotracheal intubation. * denotes p b 0.05.

PETI: prehospital endotracheal intubation, AIS: abbreviated injury score, ISS: injury sever- ity score, ETT: endotracheal tube exchange.

Table 3

multivariable regression demonstrating risk factors independently associated with need for surgical tracheostomy compared to endotracheal intubation

	OR	95% CI	p-Value	McFadden’s R2
Number of prehospital failed endotracheal intubation (PETI) attempts	1.4	1.1, 1.7	0.01	0.71
Cervical AIS >= 3	1.8	1.2, 4.1	0.03
Facial AIS >= 3	1.3	1.1, 2.2	0.02

Abbreviations: CI, confidence interval; AIS, abbreviated injury score; OR, odds ratio; PETI: prehospital endotracheal intubation.

possible and that injury patterns or prehospital airway events contrib- ute dramatically.

Several limitations exist in this work. The study focuses on a small, retrospective sample with variability in prehospital airway manage- ment. While small it a systematic investigation into the complication profile of supraglottic airways in the trauma population. Furthermore, our data demonstrate a clear selection bias based on clinical judgment - patients with risk factors for Difficult airways were managed more fre- quently with Surgical airways. A prospective, randomized comparison of surgical airway and endotracheal tube exchange (ETT) is unlikely to ever be performed, and indeed, is likely unethical. Our data support a cautious approach in regard to ETT in the emergency department that emphasizes recognition of risk factors for airway distortion and a low threshold for surgical airway in high-risk patients.

Conclusion

After supraglottic rescue airway insertion, trauma resuscitation should focus on early airway management based on patient factors, such as distorted or injured anatomy, as well as EMS factors such as number of previous intubation attempts. We demonstrate that these

factors influence the route of definitive airway management, with tra- cheostomy favored for patients with increased craniofacial injury pat- terns. Future research is necessary to confirm our findings.

References

1. Hubble MW, Brown L, Wilfong DA, Hertelendy A, Benner RW, Richards ME. A meta- analysis of prehospital airway control techniques part I: orotracheal and nasotracheal intubation success rates. Prehosp Emerg Care 2010;14(3):377-401.
2. Combes X, Jabre P, Jbeili C, Leroux B, Bastuji-Garin S, Margenet A, et al. Prehospital standardization of medical airway management: incidence and risk factors of diffi- cult airway. Acad Emerg Med 2006;13(8):828-34.
3. Hilton MT, Wayne M, Martin-Gill C. Impact of system-wide king LT airway imple- mentation on orotracheal intubation. Prehosp Emerg Care 2016;3127(July):1-8.
4. van Tulder R, Schriefl C, Roth D, Stratil P, Thalhammer M, Wieczorek H, et al. Laryn- geal tube practice in a metropolitan ambulance service: a five-year retrospective ob- servational study (2009-2013). Prehosp Emerg Care 2016;3127(July):1-7.
5. Subramanian A, Garcia-Marcinkiewicz AG, Brown DR, Brown MJ, Diedrich DA. In reply: definitive airway management in emergency department patients with a King laryngeal tube(TM) in place: a simple and safe approach. Can J Anesth 2016; 63(5):640.
6. Koumpan Y, Murdoch J, Beyea JA, Kahn M, Colbeck J. A case report. A A Case Rep 2017;8(6):139-41.
7. Jensen JL, Cheung KW, Tallon JM, Travers AH. Comparison of tracheal intubation and alternative airway techniques performed in the prehospital setting by paramedics: a systematic review. Can J Emerg Med 2010;12(2):135-40.

   Fig. 3. Proposed algorithm describing pathway for management of a supraglottic rescue airway device in patients with trauma.

   Wiese CHR, Semmel T, Muller JU, Bahr J, Ocker H, Graf BM. The use of the laryngeal tube disposable (LT-D) by paramedics during out-of-hospital resuscitation-an obser- vational study concerning ERC guidelines 2005. Resuscitation 2009;80(2):194-8.

8. Hagberg C, Bogomolny Y, Gilmore C, Gibson V, Kaitner M, Khurana S. An evaluation of the insertion and function of a new supraglottic airway device, the King LT(TM), dur- ing spontaneous ventilation. Anesth Analg 2006;102(2):621-5.
9. Frascone RJ, Wewerka SS, Griffith KR, Salzman JG. Use of the King LTS-D during med- ication-assisted airway management. Prehosp Emerg Care 2009;13(4):541-5.
10. Katz SH, Falk JL. Misplaced Endotracheal tubes by paramedics in an urban emergen- cy medical services system. Ann Emerg Med 2001;37(1):32-7.
11. Lutes M, Worman DJ. An unanticipated complication of a novel approach to airway management. J Emerg Med Elsevier Inc 2010;38(2):222-4.
12. Gaither JB, Matheson J, Eberhardt A, Colwell CB. Tongue engorgement associated with prolonged use of the King-LT laryngeal tube device. Ann Emerg Med Elsevier Inc 2010;55(4):367-9.
13. Matioc AA, Genzwuerker HV. Why go blind when you can see? J Emerg Med Elsevier Inc 2012;42(6):702-3.
14. Schalk R, Seeger FH, Mutlak H, Schweigkofler U, Zacharowski K, Peter N, et al. Com- plications associated with the Prehospital use of Laryngeal tubes-a systematic

    analysis of risk factors and strategies for prevention. Resuscitation. European Resus- citation Council, American Heart Association, Inc., and International Liaison Commit- tee on Resuscitation. Elsevier Ireland Ltd; 2014. p. 1629-32.

    Khaja SF, Chang KE. Airway algorithm for the management of patients with a king LT. Laryngoscope 2014;124(5):1123-7.

15. Van Zundert AAJ, Gatt SP, Kumar CM, Van Zundert TCRV, Pandit JJ. “Failed supraglottic airway”: an algorithm for suboptimally placed supraglottic airway de- vices based on videolaryngoscopy. BJA Br J Anaesth 2017;118(5):645-9.
16. Mayglothling J, Duane TM, Gibbs M, McCunn M, Legome E, Eastman AL, et al. Emer- gency tracheal intubation immediately following traumatic injury: an Eastern Asso- ciation for the Surgery of Trauma practice management guideline. J Trauma Acute Care Surg 2012;73(5):S333-40.
17. Warner MA, Smith HM, Zielinski MD. Impaired Ventilation and oxygenation after emergency Cricothyrotomy. A A Case Rep 2016;7(10):212-4.
18. Baraniuk S, Tilley BC, Junco DJ, Fox EE, Van Belle G, Wade CE, et al. Pragmatic ran- domized optimal platelet and plasma ratios (PROPPR) trial. Injury 2015;45(9): 1287-95.