Introduction: Laryngeal injuries are rare but life-threatening airway emergencies. Increased understanding of the epidemiology of these injuries can inform treatment and improve outcomes. We aimed to characterize the demographics and management of adult laryngeal trauma. Methods: The National Trauma Data Bank (NTDB) was queried from 2007 to 2015 for patients ≥18 years old with laryngeal trauma. Patient demographics, injury characteristics, and treatment course were collected. Outcomes were assessed via multivariate logistic regression. Results: From 7.3 million patients, 6,890 (0.1%) patients with laryngeal trauma were identified. Eighty-five percent of patients were male, and the median age was 40. Of these patients, 343 (5.0%) were dead on arrival and of the remaining patients, 510 (7.8%) of patients were deceased at discharge. Common concomitant injuries included facial fractures (27%), intracranial injuries (21%), and rib and sternum fractures (19%). The most common cause of injury was motor vehicle accident (26%), followed by assault with firearms/explosives (12%) and assault with cutting instruments (8%). Forty-three percent of patients received mechanical ventilation and 15% received surgical repair. After correcting for gender, age, and injury severity, firearm injuries (odds ratio [OR] 3.46, 95% CI: [2.88–4.15]) and cutting/piercing injuries (OR 2.23, 95% CI: [1.89–2.64]) were positively associated with the need for mechanical ventilation. Motor vehicle trauma (OR 0.63, 95% CI: [0.46–0.84]) was negatively associated with surgical repair while striking injuries (OR 1.61, 95% CI: [1.25–2.06]) were positively associated. Lastly, shorter time to tracheostomy was significantly associated with shorter ICU stays (p < 0.0001). Conclusion: This study is the largest epidemiologic study of laryngeal trauma to date and identifies the risk of surgical intervention with firearm and cutting injuries as well as the importance of earlier time to tracheostomy for ICU management.

The larynx contains cartilaginous structures, ligaments, and muscles that maintain the airway and generate voice and sound. Trauma to this area can be an acute threat to life or cause long-term sequelae, including dysphonia, dysphagia, and airway stenosis [1‒4]. Previous literature have cited incidence rates of less than 1% and mortality rates of 2% [3, 5, 6]. Injuries are relatively uncommon due to the protective nature of the mandible and the sternum but can occur from blunt force, including car accident trauma (impact on the neck from dashboard or steering wheel), strangulation, or violent punches or impact. Penetrating injuries can occur from knives, sharp glass, or firearms. Severe injuries can compromise airway, major vessels in the neck, or impact nerve branches [3, 5‒8].

Management paradigms are largely based on studies from single institutions. In brief, laryngeal trauma should be managed with thorough history and imaging given appropriate clinical context. If impending airway obstruction is noted, emergent airway through tracheostomy, intubation, or cricothyrotomy is indicated. Imaging evaluation, including laryngoscopy and CT imaging, can then triage mucosal versus cartilaginous injures and guide further management for open reductions, internal fixations, and/or stenting [1, 5, 8‒11]. Characterization of laryngeal injuries is limited to these smaller scale studies with at most several hundred patients [1, 6, 12]. The nature of these studies is limited to cases with otolaryngologist evaluation and intervention. To the best of our knowledge, there is no consensus on incidence of these injuries nor a populational understanding of cause or management of these injuries.

We sought to better understand the epidemiology of these laryngeal injuries, including mechanisms of injury, co-occurring injuries, and initial hospital management. We also aimed to analyze associations of injury characteristics and interventions.

A waiver was obtained from the University of Pennsylvania’s Institutional Review Board, approval number 848781 because this study was conducted using a de-identified dataset. We queried the National Trauma Data Bank from 2007 to 2015 for laryngeal trauma injuries. The National Trauma Data Bank is a US trauma data registry maintained by the American College of Surgeons. The dataset contains over 7.3 million de-identified records from trauma centers across the USA. Records are incidence based and include information regarding presentation to the hospital and in-hospital injury data. Demographic information and clinical injury scores are included as well [13‒15].

Cohort Selection

We identified patients between ages 18 and 89. Laryngeal trauma was determined by ICD9 Diagnosis Codes 807.5, 807.6, 874, 874.01, 874.1, 874.11. These codes include open and closed fractures of the larynx as well as wounds to the larynx.

Covariates

Demographic information including patient sex, age, and race were analyzed. To characterize the injury on a standardized scale, Injury Severity Score and Glasgow Coma Scale were recorded. Injury Severity Score (ISS) is a numerical score that is the sum of the squares of the three highest Abbreviated Injury Scores (AIS) [16]. The AIS grades six different anatomical regions (head/neck, face, chest, abdomen, extremity, external/other) on a scale of 0–6 (0 = no injury, 1 = minor, 2 = moderate, 3 = serious, 4 = severe, 5 = critical, 6 = maximum) [17]. We further categorized the ISS score into mild (ISS <9), moderate (9 ≤ ISS ≤15), and severe (ISS >15). The Glasgow Coma Scale is a 15 point scale that quantifies mental status using eye, verbal, and motor responses [18]. Brain injuries were classified as severe (GCS ≤8), moderate (GCS 9–12), or mild (GCS ≥13). Due to data availability, not every patient has these data recorded.

Concomitant Injuries

For each record, multiple ICD diagnostic codes can be recorded. We analyzed the other diagnostic ICD codes in additional to the larynx injury diagnostic code.

Mechanism of Injury

The mechanism of injury was classified using ICD9 External Cause of Injury Codes. For each mechanism of injury, the injury was also broadly categorized as blunt, penetrating, or other.

Procedures

We identified nerve repair, suture, decompression, or graft using ICD9 Procedure Codes (31.6, 31.61, 31.62, 31.63, 31.64, 31.69, 31.7, 31.71, 31.72, 31.73, 31.75, and 31.79). An individual patient could have multiple operations.

Data Analysis

Collected data were summarized and analyzed with R version 4.0.2 (Vienna, Austria 2020) and Prism 9 for macOS (San Diego, CA 2022). Descriptive statistics were utilized to analyze baseline demographic and injury characteristics. Two multivariable logistic regressions were created to predict (1) need for mechanical ventilation and (2) operative management. A linear regression model was created to estimate length of ICU stays. All three models corrected for age, race, gender, and discharge disposition (i.e., deceased at discharge). The 343 patients that were declared dead on arrival were excluded from these models. Variables included were ISS, GCS, mechanism of injury, and hospital interventions (i.e., mechanical ventilation, tracheostomy, surgical intervention).

Cohort Characteristics

The final study cohort included 6,890 patients with laryngeal trauma from 2007 to 2015. Patient demographics are presented in Table 1. Patients were predominately male (85%) and median age 40, and interquartile range of 28–52. The majority of injuries were classified as severe with ISS greater than 15 (51%). 58% of patients were characterized as mild on the Glasgow Coma Scale.

Table 1.

Demographic information and injury characteristics of laryngeal trauma patients

Patient demographics and hospital presentationn = 6,890
Age, years 
 Median 40 
 IQR 28–52 
Race, n (%)  
 White 4,030 (58) 
 Black 1,541 (22) 
 Asian 131 (1.9) 
 Other 806 (12) 
 Not known 382 (5.5) 
Glasgow Coma Scale, n (%) 
 Mild (13–15) 4,025 (58) 
 Moderate (9–12) 296 (4.3) 
 Severe (3–8) 2,569 (37) 
Injury Severity Score, n (%) 
 Mild (<9) 1,340 (19) 
 Moderate (9–15) 1,748 (25) 
 Severe (>15) 3,535 (51) 
Common concomitant injuries, n (%) 
 Facial fractures (total) 1,860 (27) 
 Mandible 1,075 (16) 
 Nasal bone 805 (12) 
 Malar/maxillary 790 (11) 
 Intracranial Injury 1,448 (21) 
 Rib and sternum fracture 1,325 (19) 
 Skull base and skull vault fractures 950 (14) 
 Cervical fracture 866 (13) 
Most common causes of injury, n (%) 
 Motor vehicle accident 1,807 (26) 
 Assault by firearms and explosives 809 (12) 
 Assault by cutting and piercing instrument 583 (8.5) 
 Unarmed fight or brawl 503 (7.3) 
 Self-inflicted injury by cutting instrument 445 (6.5) 
 Assault by blunt object 358 (5.2) 
Patient demographics and hospital presentationn = 6,890
Age, years 
 Median 40 
 IQR 28–52 
Race, n (%)  
 White 4,030 (58) 
 Black 1,541 (22) 
 Asian 131 (1.9) 
 Other 806 (12) 
 Not known 382 (5.5) 
Glasgow Coma Scale, n (%) 
 Mild (13–15) 4,025 (58) 
 Moderate (9–12) 296 (4.3) 
 Severe (3–8) 2,569 (37) 
Injury Severity Score, n (%) 
 Mild (<9) 1,340 (19) 
 Moderate (9–15) 1,748 (25) 
 Severe (>15) 3,535 (51) 
Common concomitant injuries, n (%) 
 Facial fractures (total) 1,860 (27) 
 Mandible 1,075 (16) 
 Nasal bone 805 (12) 
 Malar/maxillary 790 (11) 
 Intracranial Injury 1,448 (21) 
 Rib and sternum fracture 1,325 (19) 
 Skull base and skull vault fractures 950 (14) 
 Cervical fracture 866 (13) 
Most common causes of injury, n (%) 
 Motor vehicle accident 1,807 (26) 
 Assault by firearms and explosives 809 (12) 
 Assault by cutting and piercing instrument 583 (8.5) 
 Unarmed fight or brawl 503 (7.3) 
 Self-inflicted injury by cutting instrument 445 (6.5) 
 Assault by blunt object 358 (5.2) 

Patients may have multiple concomitant injuries.

Concomitant Injuries

The most common concomitant injuries were facial fractures (27%) and intracranial injuries (21%). These injuries are nonexclusive. Other common injuries are included in Table 1.

Injury Characteristics

59% of injuries were categorized as blunt injuries. The most common cause of injury was motor vehicle accident (26%). Other causes include assault with various weapons, which are also displayed in Table 1. The mechanism and intent of injuries are detailed in Table 2.

Table 2.

Detailed characteristics of mechanism of injury for patients presenting dead on admission (DOA), patients receiving mechanical ventilation (MV), patients receiving tracheostomy (Trach), and patients receiving surgical repair

All (n = 6,890)DOA (n = 343)MV (n = 2,823)Trach (n = 1,632)Surgery (n = 992)
Injury classification, n (%) 
 Blunt 4,067 (59) 174 (51) 1,401 (50) 787 (48) 417 (42) 
 Penetrating 2,086 (30) 130 (38) 1,116 (40) 736 (45) 511 (52) 
 Other/unspecified 737 (11) 39 (11) 306 (11) 109 (6.7) 64 (6.5) 
Intent of injury, n (%) 
 Unintentional 3,574 (52) 169 (49) 1,318 (47) 735 (45) 371 (37) 
 Assault 2,347 (34) 112 (33) 967 (34) 630 (39) 415 (42) 
 Self-inflicted 840 (12) 53 (15) 476 (17) 238 (15) 189 (19) 
 Other 129 (2) 9 (2.6) 62 (2.2) 29 (1.8) 17 (1.7) 
Mechanism, n (%) 
 Cut/pierce 1,093 (16) 31 (9.0) 519 (18) 313 (19) 354 (36) 
 Struck by, against 1,060 (15) 4 (1.2) 242 (8.6) 175 (11) 140 (14) 
 Firearm 993 (14) 99 (29) 597 (21) 423 (26) 157 (16) 
 MVT1 occupant 985 (14) 61 (18) 410 (15) 206 (13) 73 (7.4) 
 Fall 590 (8.6) 15 (4.4) 186 (6.6) 106 (6.5) 58 (5.8) 
 MVT motorcyclist 587 (8.5) 56 (16) 262 (9.3) 119 (7.3) 52 (5.2) 
 Suffocation 340 (4.9) 32 (9.3) 158 (5.6) 33 (2.0) 20 (2.0) 
 MVT pedestrian 159 (2.3) 21 (6.1) 67 (2.4) 25 (1.5) 7 (0.7) 
 Pedal cyclist 132 (1.9) 3 (0.9) 28 (1.0) 19 (1.2) 12 (1.2) 
 MVT pedal cyclist 85 (1.2) 2 (0.6) 26 (0.9) 20 (1.2) 9 (0.9) 
 Machinery 61 (0.9) 1 (0.3) 31 (1.1) 23 (1.4) 14 (1.4) 
 Natural/environmental, Bites and stings 34 (0.5) 0 (0) 15 (0.5) 12 (0.7) 10 (1.0) 
 Other 771 (11) 18 (5.2) 282 (10) 158 (9.7) 86 (8.7) 
All (n = 6,890)DOA (n = 343)MV (n = 2,823)Trach (n = 1,632)Surgery (n = 992)
Injury classification, n (%) 
 Blunt 4,067 (59) 174 (51) 1,401 (50) 787 (48) 417 (42) 
 Penetrating 2,086 (30) 130 (38) 1,116 (40) 736 (45) 511 (52) 
 Other/unspecified 737 (11) 39 (11) 306 (11) 109 (6.7) 64 (6.5) 
Intent of injury, n (%) 
 Unintentional 3,574 (52) 169 (49) 1,318 (47) 735 (45) 371 (37) 
 Assault 2,347 (34) 112 (33) 967 (34) 630 (39) 415 (42) 
 Self-inflicted 840 (12) 53 (15) 476 (17) 238 (15) 189 (19) 
 Other 129 (2) 9 (2.6) 62 (2.2) 29 (1.8) 17 (1.7) 
Mechanism, n (%) 
 Cut/pierce 1,093 (16) 31 (9.0) 519 (18) 313 (19) 354 (36) 
 Struck by, against 1,060 (15) 4 (1.2) 242 (8.6) 175 (11) 140 (14) 
 Firearm 993 (14) 99 (29) 597 (21) 423 (26) 157 (16) 
 MVT1 occupant 985 (14) 61 (18) 410 (15) 206 (13) 73 (7.4) 
 Fall 590 (8.6) 15 (4.4) 186 (6.6) 106 (6.5) 58 (5.8) 
 MVT motorcyclist 587 (8.5) 56 (16) 262 (9.3) 119 (7.3) 52 (5.2) 
 Suffocation 340 (4.9) 32 (9.3) 158 (5.6) 33 (2.0) 20 (2.0) 
 MVT pedestrian 159 (2.3) 21 (6.1) 67 (2.4) 25 (1.5) 7 (0.7) 
 Pedal cyclist 132 (1.9) 3 (0.9) 28 (1.0) 19 (1.2) 12 (1.2) 
 MVT pedal cyclist 85 (1.2) 2 (0.6) 26 (0.9) 20 (1.2) 9 (0.9) 
 Machinery 61 (0.9) 1 (0.3) 31 (1.1) 23 (1.4) 14 (1.4) 
 Natural/environmental, Bites and stings 34 (0.5) 0 (0) 15 (0.5) 12 (0.7) 10 (1.0) 
 Other 771 (11) 18 (5.2) 282 (10) 158 (9.7) 86 (8.7) 

1Motor vehicle trauma.

Outcomes

Patient outcomes are detailed in Table 3. 343 (5.0%) patients presented without signs of life to the Emergency Department. For the 6,547 remaining patients, 510 (7.8%) were deceased at discharge and 3,538 (54%) patients were discharged home.

Table 3.

Status of patient at discharge for those receiving mechanical ventilation (MV), tracheostomy (Trach), and surgical intervention

All (n = 6,547)MV (n = 2,823)Trach (n = 1,632)Surgery (n = 992)
Discharge status, n (%) 
 Discharged home 3,538 (54) 1,231 (44) 731 (44) 567 (57) 
 Discharged to rehab 616 (9.4) 400 (14) 285 (17) 115 (12) 
 Discharged under home care 379 (5.8) 208 (7.4) 212 (13) 128 (13) 
 Discharged to skilled nursing facility or similar care facility 360 (5.5) 250 (8.9) 171 (10) 72 (7.3) 
 Discharged to inpatient hospital service 193 (2.9) 124 (4.4) 84 (5.1) 39 (3.9) 
 Left against medical advice 89 (1.4) 37 (1.3) 19 (1.2) 12 (1.2) 
 Discharged to psychiatric hospital 71 (1.1) 41 (1.5) 15 (0.9) 18 (1.8) 
 Other 783 (12) 156 (5.5) 49 (3.0) 26 (2.6) 
 Discharged to hospice 8 (0.1) 5 (0.2) 4 (0.2) 2 (0.2) 
 Deceased 510 (7.8) 371 (13) 62 (3.8) 13 (1.3) 
All (n = 6,547)MV (n = 2,823)Trach (n = 1,632)Surgery (n = 992)
Discharge status, n (%) 
 Discharged home 3,538 (54) 1,231 (44) 731 (44) 567 (57) 
 Discharged to rehab 616 (9.4) 400 (14) 285 (17) 115 (12) 
 Discharged under home care 379 (5.8) 208 (7.4) 212 (13) 128 (13) 
 Discharged to skilled nursing facility or similar care facility 360 (5.5) 250 (8.9) 171 (10) 72 (7.3) 
 Discharged to inpatient hospital service 193 (2.9) 124 (4.4) 84 (5.1) 39 (3.9) 
 Left against medical advice 89 (1.4) 37 (1.3) 19 (1.2) 12 (1.2) 
 Discharged to psychiatric hospital 71 (1.1) 41 (1.5) 15 (0.9) 18 (1.8) 
 Other 783 (12) 156 (5.5) 49 (3.0) 26 (2.6) 
 Discharged to hospice 8 (0.1) 5 (0.2) 4 (0.2) 2 (0.2) 
 Deceased 510 (7.8) 371 (13) 62 (3.8) 13 (1.3) 

For all patients, the 343 patients that were dead on arrival are excluded in the total.

Interventions

Of the patients with signs of life at arrival (n = 6,547), 2,823 (43%) patients received mechanical ventilation. 1,632 (25%) patients received a tracheostomy and 992 (15%) patients received surgical repair on their initial hospitalization. There were 532 (8.1%) patients who had open operative interventions of their laryngeal fracture and 237 (3.6%) patients had their laryngeal lacerations sutured.

Prediction Models

Prediction model characteristics are summarized in Table 4. Figure 1 displays forest plots of the logistic regressions. Figure 2 displays receiver operating characteristic (ROC) curves for logistic models. Figure 3 displays the predicted versus observed values of the linear regression.

Table 4.

Predictive model variables for 3 models predicting mechanical ventilation, surgical repair, and length of ICU stay

OR2.5%97.5%p value
Logistic regression mechanical ventilation 
 (Intercept) 0.70 0.53 0.93 <0.01 
 Male 1.10 0.94 1.28  
 Age 1.00 1.00 1.01  
 ISS 1.50 1.39 1.62 <0.0001 
 GCS 1.98 1.86 2.11 <0.0001 
 Deceased at discharge 1.86 1.49 2.34 <0.0001 
 Cutting/piercing 2.23 1.89 2.64 <0.0001 
 Firearm 3.46 2.88 4.15 <0.0001 
 Suffocation 1.34 1.03 1.75 <0.01 
 Striking injury 0.63 0.52 0.75 <0.0001 
 MVT to occupant 1.16 0.98 1.38  
Logistic regression surgical repair 
 (Intercept) 0.03 0.02 0.04 <0.0001 
 Male 1.12 0.89 1.42  
 Age 1.00 0.99 1.00  
 ISS 1.28 1.15 1.43 <0.0001 
 GCS 0.89 0.81 0.98 <0.01 
 Deceased at discharge 0.19 0.10 0.33 <0.0001 
 Cutting/piercing 4.46 3.58 5.58 <0.0001 
 Firearm 0.86 0.67 1.11  
 Suffocation 0.90 0.52 1.48  
 Striking injury 1.61 1.25 2.06 <0.0001 
 MVT to occupant 0.63 0.46 0.84 <0.001 
 Mechanical ventilation 1.42 1.18 1.71 <0.0001 
 Tracheostomy 7.71 6.46 9.24 <0.0001 
OR2.5%97.5%p value
Logistic regression mechanical ventilation 
 (Intercept) 0.70 0.53 0.93 <0.01 
 Male 1.10 0.94 1.28  
 Age 1.00 1.00 1.01  
 ISS 1.50 1.39 1.62 <0.0001 
 GCS 1.98 1.86 2.11 <0.0001 
 Deceased at discharge 1.86 1.49 2.34 <0.0001 
 Cutting/piercing 2.23 1.89 2.64 <0.0001 
 Firearm 3.46 2.88 4.15 <0.0001 
 Suffocation 1.34 1.03 1.75 <0.01 
 Striking injury 0.63 0.52 0.75 <0.0001 
 MVT to occupant 1.16 0.98 1.38  
Logistic regression surgical repair 
 (Intercept) 0.03 0.02 0.04 <0.0001 
 Male 1.12 0.89 1.42  
 Age 1.00 0.99 1.00  
 ISS 1.28 1.15 1.43 <0.0001 
 GCS 0.89 0.81 0.98 <0.01 
 Deceased at discharge 0.19 0.10 0.33 <0.0001 
 Cutting/piercing 4.46 3.58 5.58 <0.0001 
 Firearm 0.86 0.67 1.11  
 Suffocation 0.90 0.52 1.48  
 Striking injury 1.61 1.25 2.06 <0.0001 
 MVT to occupant 0.63 0.46 0.84 <0.001 
 Mechanical ventilation 1.42 1.18 1.71 <0.0001 
 Tracheostomy 7.71 6.46 9.24 <0.0001 
EstimateStd.Errorp value
Linear regression for length of ICU stay (days) 
 (Intercept) 1.81 0.82 2.20 <0.01 
 Male 0.02 0.44 0.04  
 Age 0.04 0.01 4.14 <0.0001 
 ISS 0.86 0.22 3.99 <0.0001 
 GCS −0.11 0.05 −2.46  
 Deceased at discharge −2.57 0.83 −3.09 <0.001 
 Cutting/piercing −0.18 0.45 −0.40  
 Firearm 0.90 0.42 2.16 <0.01 
 Suffocation −0.85 1.08 −0.79  
 Striking injury −0.67 0.52 −1.28  
 MVT to occupant 0.09 0.49 0.19  
 Surgical repair −0.42 0.31 −1.36  
 Days on ventilation 0.68 0.01 45.04 <0.0001 
 Time to tracheostomy (days) 0.32 0.04 8.05 <0.0001 
EstimateStd.Errorp value
Linear regression for length of ICU stay (days) 
 (Intercept) 1.81 0.82 2.20 <0.01 
 Male 0.02 0.44 0.04  
 Age 0.04 0.01 4.14 <0.0001 
 ISS 0.86 0.22 3.99 <0.0001 
 GCS −0.11 0.05 −2.46  
 Deceased at discharge −2.57 0.83 −3.09 <0.001 
 Cutting/piercing −0.18 0.45 −0.40  
 Firearm 0.90 0.42 2.16 <0.01 
 Suffocation −0.85 1.08 −0.79  
 Striking injury −0.67 0.52 −1.28  
 MVT to occupant 0.09 0.49 0.19  
 Surgical repair −0.42 0.31 −1.36  
 Days on ventilation 0.68 0.01 45.04 <0.0001 
 Time to tracheostomy (days) 0.32 0.04 8.05 <0.0001 

For the logistic regressions, we report the odds ratios (ORs) and its 95% confidence interval.

For the linear regression, the estimated coefficient is reported, where a positive value is a positive association.

Fig. 1.

Forest plots of the two logistic regression prediction models. a depicts on a logarithmic scale the prediction model for mechanical ventilation. b depicts the model for surgical intervention.

Fig. 1.

Forest plots of the two logistic regression prediction models. a depicts on a logarithmic scale the prediction model for mechanical ventilation. b depicts the model for surgical intervention.

Close modal
Fig. 2.

Receiver operating characteristic (ROC) curves for the two logistic regressions models.

Fig. 2.

Receiver operating characteristic (ROC) curves for the two logistic regressions models.

Close modal
Fig. 3.

Residuals plot for the linear regression of length of ICU stay.

Fig. 3.

Residuals plot for the linear regression of length of ICU stay.

Close modal

Mechanical Ventilation

The logistic regression for mechanical ventilation corrected for age, sex, race, injury severity, and discharge disposition. Significant variables included increased risk for ventilation with cutting/piercing injuries (odds ratio [OR] 2.23, 95% CI: [1.89–2.64]) and firearm/explosives (OR 3.46, 95% CI: [2.88–4.15]. Striking injuries were less likely to require mechanical ventilation (OR 0.63, 95% CI: [0.52–0.75]).

Surgical Intervention

The logistic regression predicting the surgical intervention on initial hospitalization, after correcting for age, sex, race, injury severity, and discharge disposition, found cutting/piercing injuries (OR 4.46, 95% CI: [3.58–5.58]) and striking injuries (OR 1.61, 95% CI: [1.25–2.06]) more likely to undergo surgical repair. Motor vehicle trauma (MVT) to the occupant was less likely to undergo surgical repair (OR 0.63, 95% CI: [0.46–0.84]).

Length of ICU Stay

A linear regression model was used to predict length of ICU stay. Surgical intervention was not significantly associated with changes in ICU stay length but decreased time to tracheostomy was associated with decreased ICU stay.

Laryngeal trauma is a relatively uncommon injury. Our study places these injuries at 0.1% (1 in 1,000 cases) of the trauma cases reported in the NTDB. This rate is higher than previous estimates of 1 in 5,000 emergency room visits or 1 in 137,000 inpatient admissions [3, 6, 12]. Our study ranges from 2007 to 2015 compared to prior studies, which is a larger, more inclusive sample. Five percent of patients with laryngeal trauma arrived without signs of life, likely indicating a complicated trauma. Nevertheless, our mortality rate is 7.8%, which is higher than what has been previously reported. This is likely due to the fact that our study includes a much larger national cohort, as opposed to the other small single institutional studies [5, 6].

This is the first study to identify rates of various mechanism of injury. We see, as expected, motor vehicle accident to be the most common cause. We also identify the significant rates of laryngeal trauma from assault with firearms, sharp instruments, or blunt objects. Also, self-inflicted injuries through sharp objects or strangulation account for 12% of these patients. This study helps identify situations where it is important to remain clinically vigilant for signs of laryngeal injury. Past reports have shown that minimal external bruising and reported symptoms can mask underlying edema, hematoma, or injury [5, 10, 19].

Regarding management, each patient is unique in clinical stability, extent of injury, and access to the hospital resources. There is robust literature discussing interventional versus expectant management for patients, but priorities include stabilizing the airway and preventing future stenosis [1, 3, 9‒12, 20]. Another complicating factor includes extent of intracranial injury, which we evaluate with GCS scores. Higher GCS scores, indicating better neurologic function, are more likely to receive mechanical ventilation and less likely to receive surgical repair. We can attribute these trends to use of mechanical ventilation for patients with less severe neurologic injury and with expectations to recover. Regarding surgical intervention, GCS can be better understood as a proxy for complexity of trauma and the need for repair. Nevertheless, we identify trends that can better predict hospital course for these laryngeal trauma patients. Injury from MVT to the occupant is less associated with surgical intervention, despite being the most common cause of injury. This is likely related to the complex nature of MVT and the need for other emergent management. There is a positive association for surgical intervention with patients presenting with striking and piercing/cutting injuries, and in trauma settings, patients should be evaluated with higher suspicion for earlier operative management.

In terms of ICU stay, surgical intervention is not significantly correlated with length of stay. Surgical intervention likely requires stabilization, monitoring, and management of other concomitant injuries. We do see that there are shorter ICU stays with earlier time to tracheostomy (Table 4). There is conflicting evidence on the tracheostomy versus intubation in larynx injury [1, 20, 21]. Tracheostomy better manages airway edema and prevention of additional iatrogenic injury while intubation is more rapid and can be safely performed in cases without obvious laryngeal framework disruption [5, 8, 22]. Additionally, our analysis shows that those receiving tracheostomy are more likely to receive a surgical repair, which suggests the severity of laryngeal framework disruption.

Limitations of this study are associated with the NTDB data, which is constrained to the initial hospitalization. We are unable to extract follow-up visits and later procedures. In addition, the data are not granular enough to understand individual characteristics of each injury which would inform management and operative planning. We are also unable to attain long-term outcomes and functional impairment of these patients. These questions would warrant further studies and investigation. Lastly, the NTDB inherently has a selection and information bias from large trauma centers that report these cases. Some of these concerns are mitigated by the NTDB efforts to increase accessibility, standardization, and interhospital comparisons.

This study identifies 6,890 patients of laryngeal trauma from 7.3 million records, which is the largest cohort studied to date. The study provides a population-based data regarding injuries that have been primarily studied in smaller sample sizes at single institutions. Beyond motor vehicle trauma, patients presenting with self-inflicted or assault injuries to the face, sternum, or neck warrant a thorough investigation for laryngeal trauma as delayed management can severely impact voice and airway later. We identified piercing/cutting and striking injuries as significant injury mechanisms to require surgical intervention. Shortened time to tracheostomy was associated with decreased ICU stay, and this highlights the importance of securing a safe, long-lasting airway.

A waiver was obtained from our institution’s Institutional Review Board because this study was conducted using a de-identified dataset. This study protocol was reviewed and approved by University of Pennsylvania Institutional Review Board, Protocol #848781, Review Board (IRB #6). Consent was not required as this is a de-identified national database provided by the American College of Surgeons.

The authors declare that there is no conflict of interest.

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Katherine Xu, Emma De Ravin, and Karthik Rajasekaran: conception, acquisition, analysis, interpretation of data, manuscript preparation, and final approval. Christian Fritz, Harman Parhar, and Alvaro Moreira: manuscript preparation and final approval.

The data that support the findings of this study are available in the American College of Surgeons National Trauma Data Bank. Further inquiries can be directed to the corresponding author.

1.
Schaefer
SD
.
Management of acute blunt and penetrating external laryngeal trauma
.
Laryngoscope
.
2014
;
124
(
1
):
233
44
.
2.
Juutilainen
M
,
Vintturi
J
,
Robinson
S
,
Bäck
L
,
Lehtonen
H
,
Mäkitie
AA
.
Laryngeal fractures: clinical findings and considerations on suboptimal outcome
.
Acta Otolaryngol
.
2008
;
128
(
2
):
213
8
.
3.
Bent
JP
,
Silver
JR
,
Porubsky
ES
.
Acute laryngeal trauma: a review of 77 patients
.
Otolaryngol Head Neck Surg
.
1993
109
3 Pt 1
441
9
.
4.
Sandhu
GS
,
Nouraei
SAR
.
Laryngeal and esophageal trauma
7 ed
Cummings otolaryngology-head and neck surgery
Philadelphia
2021
. p.
939
51
.
5.
Schaefer
SD
.
The acute management of external laryngeal trauma. A 27-year experience
.
Arch Otolaryngol Head Neck Surg
.
1992
;
118
(
6
):
598
604
.
6.
Jewett
BS
,
Shockley
WW
,
Rutledge
R
.
External laryngeal trauma analysis of 392 patients
.
Arch Otolaryngol Head Neck Surg
.
1999
;
125
(
8
):
877
80
.
7.
Kleinsasser
NH
,
Priemer
FG
,
Schulze
W
,
Kleinsasser
OF
.
External trauma to the larynx: classification, diagnosis, therapy
.
Eur Arch Otorhinolaryngol
.
2000
;
257
(
8
):
439
44
.
8.
Gussack
GS
,
Jurkovich
GJ
.
Treatment dilemmas in laryngotracheal trauma
.
J Trauma
.
1988
;
28
(
10
):
1439
44
.
9.
Trone
TH
,
Schaefer
SD
,
Carder
HM
.
Blunt and penetrating laryngeal trauma: a 13-year review
.
Otolaryngol Head Neck Surg
.
1979
;
88
(
3
):
257
61
.
10.
Maran
AGD
,
Murray
JAM
,
Stell
PM
,
Tucker
A
.
Early management of laryngeal injuries
.
J R Soc Med
.
1981
;
74
(
9
):
656
.
11.
Harris
HH
,
Tobin
HA
.
Acute injuries of the larynx and trachea in 49 patients (Observations over a 15-year period)
.
Laryngoscope
.
1970
;
80
(
9
):
1376
84
.
12.
Mendelsohn
AH
,
Sidell
DR
,
Berke
GS
,
John
MS
.
Optimal timing of surgical intervention following adult laryngeal trauma
.
Laryngoscope
.
2011
;
121
(
10
):
2122
7
.
13.
Haider
AH
,
Saleem
T
,
Leow
JJ
,
Villegas
CV
,
Kisat
M
,
Schneider
EB
.
Influence of the national trauma data bank on the study of trauma outcomes: is it time to set research best practices to further enhance its impact
.
J Am Coll Surg
.
2012 May
214
5
756
68
.
14.
Hashmi
ZG
,
Kaji
AH
,
Nathens
AB
.
Practical guide to surgical data sets: national trauma Data Bank (NTDB)
.
JAMA Surg
.
2018
;
153
(
9
):
852
3
.
15.
National Trauma Data Bank® (NTDB®) | ACS [Internet]. [cited 2022 Aug 17]. Available from: https://www.facs.org/quality-programs/trauma/quality/national-trauma-data-bank/.
16.
Baker
SP
,
Oʼneill
B
.
The injury severity score: an update
.
J Trauma Acute Care Surg
.
1976
;
16
(
11
):
885
.
17.
Rating the severity of tissue damage: I. The abbreviated scale
.
JAMA
.
1971
;
215
(
2
):
277
80
.
18.
Teasdale
G
,
Jennett
B
.
Assessment of coma and impaired consciousness. A practical scale
.
Lancet
.
1974
;
304
(
7872
):
81
4
.
19.
O’Keeffe
LJ
,
Maw
AR
.
The dangers of minor blunt laryngeal trauma
.
J Laryngol Otol
.
1992
;
106
(
4
):
372
3
.
20.
Grewal
H
,
Rao
PM
,
Mukerji
S
,
Ivatury
RR
.
Management of penetrating laryngotracheal injuries
.
Head Neck
.
1995
;
17
(
6
):
494
502
.
21.
Jalisi
S
,
Zoccoli
M
.
Management of laryngeal fractures: a 10-year experience
.
J Voice
.
2011 Jul
25
4
473
9
.
22.
Gussack
GS
,
Jurkovich
GJ
,
Luterman
A
.
Laryngotracheal trauma: a protocol approach to a rare injury
.
Laryngoscope
.
1986 Jun
96
6
660
5
.