Time to resolution of radiologically detected hemothorax in trauma patients: A retrospective observational study

Abstract

BACKGROUND

Traumatic hemothorax is a common complication of chest trauma; however, the timeline for its resolution, even with chest tube thoracostomy, remains unclear.

AIM

To determine the time to resolution of the hemothorax to ensure safe discharge based on chest radiography (CXR) findings.

METHODS

A retrospective observational study was conducted at Hamad General Hospital, Qatar, from June 2014 to October 2019, including all patients with hemothorax diagnosed via computed tomography (CT) following chest trauma. Based on the initial imaging study, the hemothorax was divided into right, left, and bilateral.

RESULTS

The study included 422 patients. Of the total, 57.82% (n = 244/422) resolved their hemothorax within three days of admission. Among these, 44 patients required chest tube insertion (CTI) and 200 were cleared without it. Between days 3 and 7, an additional 16.83% (n = 71 /422) of cases were resolved, of which 28 required chest tubes. By days 8 to 14, another 11.37% (n = 48/422) were cleared, with 15 patients requiring chest tubes. After 14 days, 13.98% (n = 59/422) of patients still had hemothorax, 14 of whom required CTI.

CONCLUSION

This study showed that a subset of patients continued to experience retained hemothorax despite early tube thoracostomy. Patients with a larger hemothorax, particularly on the left side, showed prolonged resolution times. Regular imaging such as CXR or CT is recommended for up to 14 days post-intervention. After this period, outpatient follow-up is generally safe, although some patients may still have a persistent hemothorax beyond two weeks.

Key Words: Chest trauma; Hemothorax; Chest tube; Injury; Retained; Motor vehicle crashes; Imaging

Core Tip: Traumatic hemothorax often requires chest tube insertion for resolution. However, the duration of follow-up for small or retained hemothorax (RH) remains uncertain. We aimed to determine the appropriate follow-up duration by using chest radiography (CXR) or computed tomography (CT). The study included 422 patients. Of these, 75% resolved their hemothorax within 7 days and 11% on days 8 to 14 post-admission. A subset of patients continued to experience RH despite early tube insertion. Patients with larger hemothoraces, particularly on the left side, exhibited prolonged resolution times. Regular imaging using CXR or CT is recommended for up to 14 days post-intervention.

INTRODUCTION

Thoracic injuries are common in patients with blunt trauma and rank as the third most prevalent cause of trauma after head and limb injuries. Thoracic trauma entails significant morbidity and mortality, accounting for approximately 25% of trauma-related fatalities, second only to head trauma[1]. The estimated annual incidence of thoracic trauma in the United States is 300000 individuals. Within the emergency department (ED), chest radiography (CXR) serves as the primary diagnostic modality for patients with blunt chest trauma, detecting significant injuries in 6.3% to 12.4% of cases[2]. Currently, chest computed tomography (CT) is the standard imaging tool in ED settings, with a substantial increase in its utilization for the initial assessment of patients with chest trauma[3].

Thoracic injuries commonly lead to the accumulation of blood in the pleural space, known as hemothorax, with an estimated incidence of approximately 21.8%[4]. Traditional hemothorax management involves chest tube placement as advised by the guidelines of the Eastern Association for the Surgery of Trauma guidelines[5]. Despite the effective management of hemothorax with tube thoracostomy (chest tube), this intervention presents potentially serious complications such as iatrogenic injury, retained hemothorax (RH), and empyema in up to 22% of patients[6]. Even after chest tube removal, some patients may continue to exhibit RH and require further intervention or hospital admission.

A retrospective study[7] observed a 3% resolution of RH during a clinical follow-up 1 to 2 weeks after discharge in patients with blunt chest trauma. Asymptomatic patients were discharged from the trauma service. Another study[8] reported a diagnosis of hemothorax in 7.4% of outpatients with at least one rib fracture within 14 days. Moreover, approximately 30.8% of patients with RH are initially managed through observation, with 82.2% exhibiting successful management without additional intervention[9]. A recent Canadian study[10] reported an 11.8% incidence of delayed hemothorax, which causes functional decline and prolonged impairment.

Given the absence of Level I data, a protocolized approach was proposed to enhance patient outcomes and cost-effectiveness in the management of posttraumatic RH[11]. Careful consideration of the potential benefits and limitations of all available therapy modalities is essential for individual patients[9]. The time required to resolve RH after chest tube thoracostomy remains inadequately understood, highlighting the need for further research to address this knowledge gap.

This study aimed to evaluate the timeline required to resolve traumatic hemothorax after proper intervention, and to consider an appropriate time for safe discharge and/or follow-up of these patients.

MATERIALS AND METHODS

In this retrospective study, we analyzed the data retrieved from the medical records and trauma registry of the Level I Trauma Center at Hamad General Hospital. This study included adult trauma patients who exhibited hemothorax on their first CT scan upon admission to the hospital between June 2014 and October 2019. The exclusion criteria were pediatric patients and individuals without confirmed chest trauma, as indicated by CT imaging (only blunt chest trauma, not penetrating chest trauma).

Our study focused on various parameters, including demographic information (age, sex, and nationality), mechanism of injury, associated injuries (chest, spine, pelvis, abdomen, and extremities), Glasgow coma score, injury severity score, chest abbreviated injury score (AIS), intubation, intensive care unit (ICU) admission, CT findings of hemothorax, pneumothorax, lung contusion, rib fracture, chest tube insertion (CTI), last CXR before discharge, and latest CXR on initial and subsequent visits to the outpatient department. The primary measurement was the duration of the hemothorax resolution.

RH was defined as residual pleural blood exceeding 500 mL in volume, blood occupying more than one-third of the thoracic cavity, or any residual blood that could not be drained 72 hours after thoracotomy treatment[12].

Statistical analysis

As appropriate, data are presented as proportions, medians, or mean ± SD. The patients were divided into two groups based on the intervention: Chest tube or conservative treatment. Baseline demographic characteristics, management, and outcomes were compared according to management (CTI vs conservative treatment) using the student’s-t test for continuous variables and Pearson χ² test for categorical variables. For skewed continuous data, a non-parametric Mann-Whitney U test was performed. Two-tailed P values of < 0.05. were considered significant. Data analysis was conducted using the Statistical Package for Social Sciences version 18 (SPSS Inc., Chicago, IL United States).

RESULTS

A total of 439 patients were included in the study. After excluding 17 patients who died (within the 1^st week), 422 patients were included in the analysis. Figure 1 shows a flowchart of the hemothorax management including classification based on laterality, hemothorax clearance time, and outcomes. The mean age was 38.56 years (SD 15.06), and 91% were males. Motor vehicle crashes were the most common mode of injury (36.5%), followed by falls (27.7%) and pedestrian collisions (12.8%). Most patients had polytrauma, with head (28.2%), abdominal (34.4%), spinal (41.7%), and pelvic (23.5%) injuries. The mean ISS was 20.3. Pneumothorax was present in 63% of patients, 48.1% had lung contusion, and 87.2% had rib fractures. Endotracheal intubation was required in 30.1% of patients, whereas chest tubes were inserted in 23.9%. The median ICU length of stay was 5 days (2-14), with a mortality rate of 1.2%. Of the total, 152 were right-sided, 160 were left-sided, and 127 cases were bilateral hemothorax. Table 1 shows a description of trauma patients diagnosed with hemothorax on admission.

Figure 1  Flowchart of hemothorax management: Classification based on laterality, clearance time, and outcomes.

Table 1 Descriptive analysis of trauma patients diagnosed with hemothorax on admission, n (%).

Demographic information	Value
Mean age (SD) years	38.6 (15.06)
Males	384 (90.9)
Mechanism of injury
Motor vehicle crashes	154 (36.5)
Fall	117 (27.7)
Pedestrian	54 (12.8)
Associated injuries
Head	119 (28.2)
Abdomen	145 (34.4)
Spine	176 (41.7)
Pelvis	99 (23.5)
Injury scores; mean (SD)
Chest AIS	3.1 (0.26)
Head AIS	3.6 (0.99)
Abdomen AIS	2.9 (1.08)
Spine AIS	2.2 (0.68)
ISS	20.3 (11.26)
Other chest injuries
Pneumothorax	266 (63.0)
Lung contusion	203 (48.1)
Rib fracture	368 (87.2)
Interventions
Intubation	127 (30.1)
Chest tube insertion	101 (23.9)
Outcomes	Days
ICU LOS; Median (IQR)	5 (2-14)
Mortality (after clearance)	5 (1.2)

AIS: Abbreviated injury score; ISS: Injury severity score; ICU: Intensive care unit; LOS: Length of stay; IQR: Interquartile range.

Overall hemothorax clearance

Of all the hemothorax cases, 57.82% were cleared within 0-3 days, and 18% required chest tubes. Results showed that 16.83% were cleared between 3-7 days, with 39.4% requiring chest tubes. Furthermore, 11.37% were cleared between 8-14 days, with 31.25% requiring chest tubes, while 13.98% persisted over 14 days, with 23.7% requiring chest tubes. Figures 2, 3 and 4 show examples of three cases of traumatic hemothorax.

Figure 2 Image of hemothorax and chest tube. A: Initial computed tomography (CT) scan of a polytrauma patient with right hemothorax; B: Chest X-ray post-CT scan-chest tube was inserted with a resolution of hemothorax.

Figure 3 Chest X-ray and computed tomography of hemothorax. A: Initial chest X-ray in trauma room showing left hemothorax; B: Computed tomography (CT) scan showing mild to moderate left-sided pleural effusion with underlying consolidation and contusion in the lingula and left lower lobe; C: Left chest tube insertion post CT scan; D: After a 1-week X-Ray chest showing left chest tube was removed, however a small left hemothorax with the pleural thickening remains detected. Lungs are clear.

Figure 4 Images of bilateral hemothorax and its clearance. A: Initial computed tomography scan on admission showing bilateral hemothorax with segmental compression atelectasis; B: Follow up after 1 month-chest X-Ray costophrenic angles are clear (without chest tube insertion). It shows also thoracic spine fixation.

Right-sided hemothorax

A total of 152/439 traumatic hemothorax cases were identified. Three patients died within the first three days and were excluded from the study. In the surviving group, 64.4% of patients had pneumothorax, 43.6% had lung contusion, and 83.9% had rib fractures. A total of 105/149 (70.5%) hemothorax cases were cleared radiologically with a follow-up CT scan or chest X-ray between 0-3 days post-hospital admission. The median hemothorax size was 11 mm (range, 1-60 mm). While 17/105 (16.20 %) patients required a chest tube, the remaining patients were treated conservatively. Clearance without a chest tube occurred in 88/105 (83.80%) cases.

Of the remaining 44/149 (29.53%) cases, 19/149 (12.7%) were cleared radiologically within the 4^th to 7^th days post-trauma. The hemothoraces in this group were larger than those in the earlier group, with a median size of 17 mm (range: 2-60 mm). Moreover, CTI was required in 8/19 (42.10%) patients, while the remaining 11/19 (57.90%) patients did not receive CTI.

Right-sided hemothorax persisted in 25/149 (16.78 %) cases after 1 week, with only 14/149 (9.4%) cases resolved between 8 and 14 days. Furthermore, 3/14 (21.43%) cases required a CTI, and11/14 (78.57%) cases were cleared without it. The median size of the hemothorax in this group was 18 mm (2-60 mm). After 14 days, 11/149 (7.4%) patients with hemothorax persisted on imaging follow-up, and only 2/11 (18.2%) patients received CTI. Furthermore, 9/11 (81.8%) cases were subsequently cleared on the follow-up imaging scans.

Left-sided hemothorax

In total, 160/439 cases of left-sided traumatic hemothorax were identified. Four patients died and were excluded from the study. In this group, 60.3% of patients had pneumothorax, 50% had lung contusion, and 85.3% had rib fractures. Of 156 cases, 79/156 (50.6%) were cleared radiologically between 0-3 days from hospital admission. The median hemothorax size was 11 mm (range, 1-55 mm). While 10/79 (12.66%) patients required a chest tube, the remaining patients were treated conservatively with no chest tube. Clearance without a chest tube was observed in 69/79 (87.34%) cases.

In the remaining 77/156 (49.0 %) cases, 31/156 (19.9 %) hemothoraces were cleared radiologically within day 4-7 days post-trauma. The hemothoraces in these cases were larger than those in the earlier group, with a median size of 17 mm (range, 1-55 mm). Furthermore, CTI was needed in 11/31 (35.48%) patients, while the remaining 20/31 (64.52%) patients did not receive chest tubes.

Left-sided hemothorax persisted in 46/156 (29.5%) cases after 1 week, and only 16/156 (10.3%) cases resolved between 8 and 14 days. Moreover, 7/16 (43.75%) cases required a chest tube, while 9/16 (56.25%) cases cleared the hemothorax without tubes. The median hemothorax size was 18 mm (1-55 mm). After 14 days, 30/156 (19.2%) cases persisted during follow-up, and only 5/30 (16.67%) cases were treated with CTI. Furthermore, 25/30 (83.33%) cases were cleared up on follow-up.

Bilateral hemothorax

A total of 117/439 patients with bilateral traumatic hemothorax were identified. Ten patients died and were excluded from the study. Nearly all patients had severe polytrauma, which required the insertion of bilateral chest tubes. In this group, 65% of cases had pneumothorax, 51.3% had lung contusion, and 94% had rib fractures. Table 2 shows a comparative analysis of chest injuries and tube thoracostomy in right, left, and bilateral hemothorax patients. Of the bilateral cases, 60/117 (51.3%) were cleared radiologically between 0-3 days post-hospital admission. The median hemothorax size was 8 mm (2-53 mm) on the right side and 9 mm (1-37 mm) on the left side. While 17/60 (28.33%) patients required CTI, the remaining were treated conservatively. Clearance without CTI occurred in 43/60 (71.67%) cases.

Table 2 Comparative analysis of chest injuries and tube thoracostomy in right, left, and bilateral hemothorax patients (n = 422), n (%).

	Hemothorax right (n = 149; 35.3)	Hemothorax left (n = 156; 37.0)	Bilateral hemothorax (n = 117; 27.7)	P value
Chest injuries
Pneumothorax	96 (64.4)	94 (60.3)	76 (65.0)	0.661
Lung contusion	65 (43.6)	78 (50.0)	60 (51.3)	0.388
Rib fracture	125 (83.9)	133 (85.3)	110 (94.0)	0.032
Clearance by days
0-3 days	105 (70.5)	79 (50.6)	60 (51.3)	0.004 for all
> 3-7 days	19 (12.7)	31 (19.9)	21 (17.9)
> 7-14 days	14 (9.4)	16 (10.3)	18 (15.4)
> 14 days	11 (7.4)	30 (19.2)	18 (15.4)
Clearance with tube thoracostomy (chest tube) insertion	n = 30 (20.1)	n = 33 (21.2)	n = 38 (32.5)
Chest tube 0-3 days	17 (56.7)	10 (30.3)	17 (44.7)	0.389 for all
Chest tube > 3-7 days	8 (26.7)	11 (33.3)	9 (23.7)
Chest tube > 7-14 days	3 (10.0)	7 (21.2)	5 (13.2)
> 14 days	2 (6.7)	5 (15.2)	7 (18.4)
Clearance without chest tube insertion	n = 119 (79.9)	n = 123 (78.8)	n = 79 (67.5)
0-3 days	88 (73.9)	69 (56.1)	43 (54.4)	0.007 for all
> 3-7 days	11 (9.2)	20 (16.3)	12 (15.2)
> 7-14 days	11 (9.2)	9 (7.3)	13 (16.5)
> 14 days	9 (7.6)	25 (20.3)	11 (13.9)

In the remaining 57/117 (48.7%) patients, 21/117 (17.9%) hemothoraces were cleared within 4-7 days post-trauma. The hemothoraces in these cases were larger, with a median size of 8 mm (2-53 mm) on the right side and 11 mm (2-37 mm) on the left side. A chest tube was inserted in 9/21 (42.85%) cases, while 12/21 (57.15%) patients did not undergo CTI.

Bilateral hemothorax persisted in 36/117 (30.76%) cases after 1 week, with 18/117(15.4%) resolved of hemothorax between days 8 and 14, 5/18 (27.78%) cases requiring CTI, and 13/18 (72.22%) cases cleared without CTI. The median hemothorax size was 8 mm (2-53 mm) on the right side and 10.5 mm (2-36 mm) on the left side.

After 14 days, 18/117 (15.4 %) cases with hemothorax persisted on follow-up, 7/18 (38.89%) were treated with CTI, and 11/18 (61.11%) cases were cleared on the follow-up imaging scans. Table 3 demonstrates the time to clearance of hemothorax in patients with and without tube thoracostomy regardless of laterality in hemothorax.

Table 3 Time to clearance of hemothorax in patients with and without tube thoracostomy regardless of laterality in hemothorax.

	Total clearance (n = 539)	Clearance with TT (n = 139)	Clearance without TT (n = 400)	P value
Time of clearance
Clearance 0-3 days	304 (56.4)	61 (43.9)	243 (60.8)	0.001 for all
Clearance 3-7 days	92 (17.1)	37 (26.6)	55 (13.8)
Clearance > 7-14 days	66 (12.2)	20 (14.4)	46 (11.5)
> 14 days	77 (14.2)	21 (15.1)	56 (14.0)

TT: Tube thoracostomy.

DISCUSSION

Hemothorax is characterized by the accumulation of blood in the pleural space or a pleural fluid hematocrit exceeding 50%[13]. In the context of evaluating thoracic injury in the ED, CXR has traditionally been used as the first-line diagnostic tool. However, for the assessment of traumatic hemothorax and the identification of additional injuries, chest CT has emerged as the preferred imaging modality, as approximately 14% of patients can be expected to have an occult hemo/pneumothorax found on CT after a normal CXR[3].

RH is defined as blood that occupies at least one-third of the pleural space and cannot be drained by thoracostomy within 72 hours or as clots of at least 500 mL in volume[14]. RH can develop as early as 24 hours after chest tube placement[15]. The incidence of RH is estimated to range from 5% to 30%, with a significant proportion of patients requiring surgical evacuation[16]. Although these RH may spontaneously reabsorb 4-6 weeks after trauma, excessive RH may lead to additional complications[15]. Reports show several risk factors for RH including a high chest AIS score, low hematocrit rate, initial hemothorax volume (with an increased risk of 15% every 100 mL/day of hemothorax), low Glasgow Coma Scale, number of chest tubes placed, and length of time with thoracostomy. The median reported size of hemothorax was 353 mL (233-397), and as described by Prakash et al[17], the development of an RH increased in moderate and large initial hemothorax with 44.3% and 39.1% RH risk rates, respectively. The mean age of the study cohort was approximately 38.5 years, which was consistent with previous findings[18].

In our study, hemothorax clearance was evaluated radiologically using CT or CXR. Each group was further stratified based on the time taken for radiological clearance within specific intervals following diagnosis and use of CTI. Earlier clearance was associated with smaller hemothorax sizes and lower chest tube requirements. Conversely, delayed hemothorax clearance or persistence was linked to larger hemothorax sizes and increased chest tube requirements.

The size of the hemothorax plays a crucial role in determining the appropriate management strategy, thus influencing the necessity of CTI[19]. Notably, in our study, the median hemothorax sizes for patients who cleared the condition within 72 h were the same on the right and left sides (11 mm), with a small bilateral decrease in hemothorax size (8-9 mm).

Our study revealed that the majority of hemothorax cases (58%) were resolved within the first 72 hours. Given the pivotal role of hemothorax size in decision-making regarding CTI, only 44/244 (18%) cases required early intervention in the form of CTI due to significant hemothorax size within three days of diagnosing traumatic hemothorax, and 82 % of patients cleared the hemothorax without the need for CTI. Hemothorax resolution is also influenced by various factors, including patient age, comorbidities, and associated injuries[20]. In our study, not all patients achieved hemothorax clearance within the first 72 hours.

As previously discussed, the size of the hemothorax is an important factor in resolution[20]. As the size increases, the chances of RH increase (17), which is consistent with our study where patients who did not clear their hemothorax within 3 days had a larger hemothorax (median size of 17 mm) as opposed to those who cleared in the first 72 hours (median size of 11 mm). In our study, 71/422 (16.83%) cases cleared the hemothorax between 4-7 days, resulting in approximately 74.65% of cases cleared by the end of the 1^st week.

After 7 days, approximately 1/4 of all the patients (25.35%) failed to resolve on follow-up imaging despite appropriate management with either conservative treatment or with CTI. One of the reasons for the delayed resolution in these cases may be related to the fact that these patients were mostly polytrauma patients[21] with high ISS and Chest AIS scores, despite having nearly the same hemothorax sizes which resolved within a week.

This study also reported a higher prevalence of persistent hemothorax after 14 days (14%). Notably, it was more prevalent in patients with left-sided hemothorax (19.2%). This observation suggests a potential correlation between the persistence of hemothorax and the severity of associated injuries, particularly in the context of left-sided traumatic hemothorax. However, the available data are insufficient to conclusively support this hypothesis.

The criteria used for discharge of these patients included being clinically asymptomatic on room air, stable vital signs, and a last chest radiograph before discharge showing only minimal blunting of the CP angle (in case the hemothorax was not entirely resolved and still visible on imaging with no clinical signs or symptoms). Of note, our hospital has no strict protocol for follow-up of these patients. Nevertheless, these patients are generally administered on a case-by-case basis every 1-2 weeks until complete resolution of the hemothorax on outpatient visits.

None of these patients required additional intervention during this period, which is consistent with the findings of a follow-up hemothorax study[7] also showing no intervention during the follow-up of these patients.

Limitations

This study had several limitations. Because of the small sample size, identifying significant predictors was challenging. Also, this was a retrospective study and data collection was limited to the amount of information available on patient charts. Therefore, many aspects could not be explained based on the available information. The decision to insert a chest tube was mainly clinical; however, based on clinical judgment, a chest tube was inserted in all patients who required it. Moreover, we do not have an explanation for why a subset of hemothoraces did not resolve and persisted beyond 14 days.

Finally, the monocentric nature of this study should be considered, and our results should be confirmed by further multicenter studies conducted on larger populations.

Recommendation

Patients with RH were discharged based on clinical assessment (i.e., asymptomatic, maintaining saturation on room air, good incentive, and satisfactory radiography before discharge, including minimal hemothorax, not warranting drainage). Follow-ups were performed weekly after discharge, especially in the first two weeks. Subsequently, follow-ups continued at the clinician’s discretion and judgment. We did not report any complications in patients discharged with a returned hemothorax (i.e., symptomatic presentation or requirement for hemothorax drainage).

CONCLUSION

RH persisted in a subset of patients, despite early CTI intervention. Our findings showed that patients with large hemothoraces tended to experience prolonged left-sided as opposed to right-sided hemothoraces. We recommend regular imaging, such as CXR or CT, for up to 14 days (2 weeks) postintervention to monitor patients. After this period, outpatient follow-up is considered appropriate, although a small subset of patients may still exhibit persistent RH beyond 14 days. Further studies with larger sample sizes are required to identify early predictors of interventions for RH and to improve patient outcomes.

ACKNOWLEDGEMENTS

The authors thank the Medical Research Center and the Qatar National Trauma Registry at HMC, Qatar.