Introduction

Retropharyngeal hematoma is a life-threatening condition due to the potential for progressive upper airway obstruction. Spontaneous unprovoked bleedings are extremely rare, and the condition is often accompanied by various precipitating events, such as cervical spinal cord injury, trauma with or without bone injury, and neck surgery [1,2,3,4]. Protection of the airway is the primary strategy, followed by intensive care.

Although retropharyngeal hematoma is a serious condition, there are no clinical guidelines regarding its management, and its clinical characteristics have not yet been fully examined [5]. Patients with retropharyngeal hematoma following spinal cord injury and spinal fracture are commonly observed, and the clinical course and outcomes of these patients are determined by their primary injuries [6]. Thus, the natural clinical course of retropharyngeal hematoma remains unclear.

In this study, we aimed to examine the clinical characteristics and treatment periods of traumatic retropharyngeal hematoma without spinal cord injury or spinal fracture (TREWISS).

Materials and methods

Patients and study design

We performed a national retrospective analysis of patients who presented to the emergency department and were identified by neck computed tomography (CT) scan with a soft tissue swelling of the retropharyngeal space between April 2010 and April 2020. Data were collected from eight hospitals in Japan (Fig. 1). The study was approved by the local ethics committee of St. Luke’s International Hospital (approval number; 19-R129) and all the other centers, and was conducted following STROBE guideline. The inclusion criterion was thickness of the retropharyngeal space > 7 mm at C1–C4 or > 22 mm at C5–C7 on CT images (Fig. 2). The thickness of the retropharyngeal space was measured on sagittal images as the distance between the anterior border of the vertebral body and the air within the pharynx or trachea.

Fig. 1
figure 1

Participating facilities in this national survey

Full size image
Fig. 2
figure 2

Inclusion criterion

Full size image

Exclusion criteria were (1) age < 18 years, (2) cardiopulmonary arrest, (3) other causes of soft tissue swelling besides hematoma, such as cysts and abscesses, and (4) patients with spinal cord injury or spinal fracture. Spinal cord injury was diagnosed based on neurological examination and MRI imaging. Clinical assessments were not standardized and were performed according to each centers’ standard of care.

Data collection and processing

The following parameters were recorded for each study subject: age, sex, body mass index, comorbidities, medications, vital signs on arrival, symptoms, laboratory data, imaging findings in the emergency department, airway management (including medications used when intubating, and whether endotracheal intubation, cricothyrotomy or tracheotomy was performed), interventional radiology procedure, if performed, amount of blood transfusion, duration of mechanical ventilation, duration of illness (in days) at the time of tracheal tube removal, duration of ICU and hospital stay, and outcomes.

Study endpoints

The primary endpoint was to clarify clinical factors associated with intubation requirements. The secondary endpoints were durations of mechanical ventilation, tracheal tube placement, and ICU or hospital stay.

Statistical analysis

Descriptive statistics were used to summarize data on baseline characteristics and outcomes. Categorical variables were counted and presented as proportions. Continuous variables were expressed as medians and interquartile ranges (IQRs). Baseline characteristics were compared between intubated and non-intubated patients using the Wilcoxon test for continuous variables and the chi-square test for binary and categorical variables, as appropriate. Baseline characteristics were also compared between patients discharged to their homes and those transferred to other hospitals, in order to determine factors related to patient outcomes. Statistical analysis was performed using JMP version 14 statistical software (SAS Institute, Inc., Cary, NC, USA). Missing data were excluded from analysis.

Results

From April 2010 to April 2020, a total of 152 patients met the inclusion criterion. Of these, two patients aged < 18 years, one patient with cardiopulmonary arrest, and one patient with severe traumatic brain injury were excluded. Other causes of retropharyngeal hematoma, such as cysts and abscesses, were suspected in 22 patients, and spinal cord injury or spinal fracture were recognized in 104 patients. After excluding these patients, the remaining 22 patients were enrolled in analysis. Among them, 16 patients needed intubation (Fig. 3).

Fig. 3
figure 3

Patient flow in this study

Full size image

Baseline characteristics of the study population

Overall, median patient age was 69 years, and 18 (81.8%) patients were male (Table 1). Twenty-seven percent of them were on antiplatelet or anticoagulant medications. Vital signs, except for respiratory rate, were normal in all cases. Laboratory data showed elevated d-dimer levels, but were otherwise normal. Stridor, tracheal deviation and cervical swelling were observed, although neck purpura was not observed in our cases. The median width of the hematoma on sagittal CT views was 2.2 cm. More than 80% of patients were admitted to the ICU for close follow up. Overall, 13 patients were discharged to their homes, eight were discharged to another hospital for rehabilitation, and one patient died during hospitalization.

Table 1 Baseline characteristics of the study subjects
Full size table

Comparison of clinical characteristics between intubated and non-intubated patients

Patients in the intubated group were significantly older [71 years (57–80) vs. 37 years (25–78), p = 0.011] (Table 1). Stridor, deviation of the trachea, and neck swelling were observed only in intubated patients. Stridor and cervical swelling were significant physical findings in the intubated group [10 (62.5) vs. 0 (0), p = 0.015, and 9 (56.3) vs. 0 (0), p = 0.046, respectively]. Hemoglobin levels and platelet counts were significantly lower in the intubated group [13.0 (9.6–13.7) vs. 15.3 (14.0–16.1), p = 0.004, and 17.9 (11.9–25.0) vs. 24.0 (20.9–30.2), p = 0.043, respectively]. The hematoma on sagittal CT images was significantly wider in the intubated group [2.5 cm (2.0–3.4) vs. 1.2 cm (0.9–1.7), p = 0.002].

Outcomes of intubated patients

All the intubated patients were intubated on the first day of admission (Table 2). Among the intubated patients, cricothyrotomy was performed in three (18.8%) patients and orotracheal intubation was performed in the others. Muscle relaxants were used in only two (12.5%) patients at the time of intubation. Tracheotomy was performed in nine (56.3%) patients, at a median time of three days after hospitalization. Median durations of mechanical ventilation, endotracheal tube placement, and hospital stay were 4, 22 and 30 days, respectively. Fifty percent of patients were successfully discharged to their homes, and one (6.3%) patient died in the hospital. Comparison of clinical characteristics between patients discharged to their home and those transferred to another hospital indicated no significant differences in baseline characteristics between them (Supplemental Table 1).

Table 2 Treatment and outcomes of intubated patients
Full size table

Details of the fatal case

The patient who died while in the hospital was an 80-year-old male with a history of mitral valve replacement who was on warfarin treatment, who presented to the emergency department after falling to the ground. He had an injury on the forehead and CT scan revealed a retropharyngeal hematoma. The patient was intubated in the emergency department and tracheotomy was performed on Day 3. The retropharyngeal hematoma resolved within two weeks; however, his condition was complicated by a psoas muscle hematoma on Day 11 and gallstone pancreatitis on Day 23. The patient subsequently died because of multiple organ failure.

Discussion

Our retrospective study of 22 patients with TREWISS revealed that only 60% of patients were successfully discharged to their homes and one patient (6.3%) died during hospitalization. The width of the hematoma on sagittal CT images was significantly wider in the intubated group, and large hematomas were seen even in patients who were not on antiplatelet or anticoagulant medications. More than 70% of patients were intubated, and among the intubated patients, tracheotomy was performed on around the third day, and the endotracheal tube was placed for about 3 weeks.

Comparison

We reviewed 46 previous case reports of TREWISS (Table 3). Among them, median patient age was 72 years and males constituted 72% of the population; the characteristics of patients in those studies were similar to those of patients in the current study. The mortality rate was 11%, and the cause of death in the patients who died were unrelated to the TREWISS, and included cardiac, pulmonary and gastrointestinal causes. Almost 60% of patients needed intubation, which included orotracheal intubation, nasotracheal intubation, emergency tracheotomy and cricothyrotomy. Overall, tracheotomy was performed in 33% of the patients. The rates of intubation and tracheotomy were also similar to those in the current study. In previous studies, data on the discharge site were available in only a few cases [18–20].

Table 3 Comparison of patient characteristics between previous cases and the current study
Full size table

Mechanism of hematoma

In the current study, we focused on trauma patients without spinal fracture, and the mechanism of hematoma in this category of patients was thought to be injury to the anterior longitudinal ligament, longus colli muscle, or branches of the vertebral and thyroid arteries. For example, in our study, interventional radiology treatment was performed for two patients and both of them had extravasation from branches of the inferior thyroid artery. Anatomically, the space between the vertebral bodies and pharynx is divided into three spaces (the retropharyngeal space, the danger space and the prevertebral space) [7]. When hematomas develop in these spaces, they tend to expand anteriorly because posteriorly located vertebral bodies prevent their spread in that direction, compressing the airway [8]. Hematomas in this area are usually isolated and have no place to drain, and hence, it takes time for them to be absorbed spontaneously. A previous study described the time course of hematoma resorption, showing that it usually resolves completely in 3 to 4 weeks [9]. Our study also showed that endotracheal tube were needed for about 3 weeks, and that patients with a large retropharyngeal hematoma required hospitalization for several weeks for close observation.

Clinical implementation

Cricothyrotomy was performed for three patients in our study, all of whom were difficult to intubate in the emergency department. Some previous studies that have described the time course from trauma to intubation showed that respiratory collapse occurred within several hours after the traumatic event, ranging from 2 to 17 h [10,11,12]. Since retropharyngeal hematoma can progress rapidly, intubation should be completed soon after the diagnosis is made in order to prevent sudden airway obstruction. Older patients and those with lower hemoglobin levels and platelet counts might be candidates for early intubation. Physical findings, such as stridor and neck swelling, are also important determinants of the timing for intubation. A wider hematoma in sagittal CT images also indicates the need for earlier intubation. Six previous studies described the width of the hematoma in intubated patients, which ranged from 1.5 to 3.6 cm [13,14,15,16], which was similar to the intubated cases in our study, in whom the median width of the hematoma was 2.5 cm. In our study, more than half of intubated patients needed tracheotomy, and the median day of illness on which the tracheotomy was performed was Day 3. In previous studies, tracheotomy was performed within four days in all the patients evaluated. [7, 17] Some studies used follow-up imaging or direct observation of the hematoma by laryngoscopy to decide whether to perform tracheotomy [12, 18, 20].

The present study showed that more than 40% of patients needed to be transferred to other hospitals for rehabilitation. We compared baseline characteristics between patients who were discharged to their homes and those who were discharged to other hospitals, and found no significant differences in baseline characteristics between the groups, including in age, intubation, duration of ICU stay, and duration of mechanical ventilation. However, this result might indicate a beta error because of insufficient patient number.

Limitations

Several limitations of our study should be addressed. First, the need for intubation was determined by each physician, and there is no standardized protocol for intubation. Therefore, the precise indications for intubation remain unclear. Some patients also prophylactically received endotracheal intubation. Second, the number of patients in our study was relatively small. Third, data on the time course of hematoma expansion were not obtained. Hence, we could not describe how fast the hematoma expands and when physicians have to make the decision to intubate.

Conclusion

Early intubation and subsequent intensive care are important for patients with TREWISS. Due to the anatomical characteristics of this condition, patients require several weeks of hospitalization and their outcomes are usually not favorable.