Introduction
Penetrating thoracoabdominal injuries remain a challenging diagnostic and therapeutic problem for trauma surgeons. The key surgical dilemma is determining which cavity should be explored first: the chest or the abdomen. This decision is particularly critical in hemodynamically unstable patients, in whom delayed control of the dominant life-threatening injury may adversely affect outcomes [1].
Initial management is often guided by hemodynamic status, suspected source of bleeding, chest tube output, imaging findings, and the mechanism of injury. However, the clinical picture may be misleading, and the initially selected operative approach may not always correspond to the cavity requiring urgent exploration. Therefore, understanding the sequence of operative interventions and the reasons for inappropriate initial cavity selection is essential for improving decision-making in penetrating thoracoabdominal trauma.
Aim
The aims of this study were: (1) to describe the clinical characteristics of patients with penetrating thoracoabdominal injuries requiring emergency surgical treatment; (2) to analyze the sequence of operative interventions according to the initial operative approach, chest-first or abdomen-first; and (3) to evaluate the appropriateness of initial cavity selection and identify potential pitfalls.
Material and methods
Study design and setting
This retrospective cohort study was conducted at Kyiv City Clinical Hospital No. 17 over a 23-year period, from January 2000 through December 2023. The study included adult patients with penetrating thoracoabdominal injuries involving thoracic and/or abdominal structures, including diaphragmatic, pulmonary, vascular, and intra-abdominal organ injuries, who underwent emergency surgical management. Patients managed without surgical intervention were not included. The study was reported in accordance with the STROBE guidelines for observational studies.
For the purpose of this analysis, patients were stratified according to the initial operative approach. Patients in whom the first surgical access was thoracic were assigned to the chest-first group, whereas patients in whom the first surgical access was abdominal were assigned to the abdomen-first group. The primary outcome was the appropriateness of initial cavity selection, defined as successful control of the dominant life-threatening injury through the initially selected surgical access without an unplanned urgent need to interrupt the procedure and proceed to the opposite cavity. Secondary outcomes included the need for exploration of the second cavity, estimated blood loss, damage control surgery, postoperative complications, and mortality.
Data sources
Clinical data were extracted retrospectively from hospital medical records, including emergency department notes, eFAST reports when available, chest radiographs, operative reports, perioperative anesthetic charts, intensive care unit records, and discharge summaries. Extracted variables included demographic characteristics, mechanism of injury, hemodynamic status on admission, diagnostic findings, initial operative approach, sequence of surgical access, injured structures, estimated blood loss, emergency department thoracotomy, damage control surgery, postoperative course, and in-hospital mortality.
Study population and eligibility criteria
A total of 111 consecutive adult patients aged 18 years or older with penetrating thoracoabdominal injuries met the inclusion criteria. Eligible patients had penetrating trauma involving the thoracoabdominal region and underwent emergency surgical management for suspected or confirmed injury to thoracic, diaphragmatic, vascular, or intra-abdominal structures. Diaphragmatic injury, a transdiaphragmatic wound tract, or combined thoracic and abdominal involvement was considered suggestive of thoracoabdominal trauma.
Patients declared dead on arrival, patients without signs of life for more than 5 minutes before hospital admission, patients with blunt trauma, and patients managed without surgical intervention were excluded from the study.
All patients were managed in the polytrauma department according to Advanced Trauma Life Support principles. Perioperative care included hemodynamic resuscitation, bleeding control, prevention of hypothermia, and damage control strategies when clinically indicated.
Admission classification and initial management
On admission, patients were classified as hemodynamically stable or hemodynamically unstable according to vital signs, clinical signs of shock, and response to initial resuscitation. Initial assessment and management were performed according to Advanced Trauma Life Support principles.
Resuscitation was initiated immediately and included intravenous fluid administration, blood products when indicated, active warming, and prevention of hypothermia. In patients with suspected ongoing hemorrhage, permissive hypotension was used until surgical bleeding control when clinically appropriate.
When the patient’s physiological condition allowed, preoperative diagnostic assessment included eFAST, chest radiography, and chest computed tomography. In unstable patients, diagnostic evaluation was limited to rapid bedside assessment, and the decision regarding the initial operative approach was based on hemodynamic status, suspected source of bleeding, chest tube output, eFAST findings, and clinical signs of thoracic or abdominal injury.
Antibiotic prophylaxis with second- or third-generation cephalosporins and tetanus prophylaxis were administered according to institutional protocols.
Outcome measures
The primary outcome was the appropriateness of initial cavity selection, defined as successful control of the dominant life-threatening injury through the initially selected surgical access without an unplanned urgent need to interrupt the procedure and proceed to the opposite cavity. Secondary outcomes included the need for exploration of the second cavity, sequence of operative interventions, associated thoracic and abdominal injuries, estimated blood loss, emergency department thoracotomy, damage control surgery, postoperative complications, missed injuries, and in-hospital mortality. In-hospital mortality was defined as death occurring during the index hospitalization.
Statistical analysis
Data were analyzed retrospectively using R software (version 4.3.3). Continuous variables were reported as mean ± standard deviation (SD) if approximately normally distributed or as median with interquartile range (IQR) otherwise. Categorical variables were reported as counts and percentages.
For key proportions, 95% confidence intervals (CIs) were calculated using the Wilson score method. Patients were stratified according to the initial operative approach into chest-first and abdomen-first groups. Between-group comparisons were performed for available categorical outcomes that could be summarized as 2 × 2 contingency tables. Because of small cell counts, Fisher’s exact test was used for these comparisons. Continuous variables were not compared between groups because patient-level data were not available for all variables.
All statistical tests were two-sided, and a p-value < 0.05 was considered statistically significant.
Results
Over the 23-year study period, 111 patients with penetrating thoracoabdominal injuries met the inclusion criteria. The cohort included 102 male patients (91.9%) and 9 female patients (8.1%). The mean age was 32.65 ±6.72 years, with a range from 18 to 65 years.
The injury was right-sided in 40 (36.0%) patients and left-sided in 71 (64.0%) patients. The mechanisms of injury were stab wounds in 99 (89.2%) patients, gunshot wounds in 7 (6.3%) patients, and penetrating injuries involving impacted foreign bodies in 5 (4.5%) patients. On admission, 80 (72.1%) patients were hemodynamically stable, whereas 31 (27.9%) patients were hemodynamically unstable because of hemorrhage. Baseline characteristics of the cohort are presented in Table 1.
Table 1
Baseline characteristics of the cohort
Diagnostic assessment included eFAST, chest radiography, and chest computed tomography when the patient’s physiological condition allowed. Emergency department thoracotomy was performed in 18 (16.2%) patients.
According to the initial operative approach, 69 patients were assigned to the chest-first group and 42 patients to the abdomen-first group. A single-cavity operation was sufficient in 74 patients, whereas 37 patients required sequential or repeated access to another cavity.
Thoracotomy alone was performed in 40 (36.1%) patients, laparotomy alone in 34 (30.6%) patients, thoracotomy followed by laparotomy in 26 (23.4%) patients, laparotomy followed by thoracotomy in 8 (7.2%) patients, repeated thoracotomy in 2 (1.8%) patients, and thoracotomy followed by repeated thoracotomy and laparotomy in 1 (0.9%) patient. Surgical procedures and sequences of access are summarized in Table 2.
Table 2
Surgical procedures and interventions
Operative sequence was further analyzed according to the initial operative approach. Compared with the abdomen-first group, patients in the chest-first group more frequently required additional operative access and exploration of the opposite cavity. However, the rate of incorrect initial cavity selection was not significantly different between the two groups (Table 3).
Table 3
Operative sequence according to the initial operative approach
Values are presented as n (%). Between-group comparisons were performed using Fisher’s exact test.
Among patients who required combined or repeated access, the initial surgical procedure was interrupted in 6 cases because urgent exploration of the opposite cavity was required. This corresponded to 5.4% of the total cohort and 16.2% of patients who required combined or repeated access. Interruption of the initial procedure occurred during laparotomy in 3 cases and during thoracotomy in 3 cases. These cases were mainly related to misinterpretation of chest radiography or chest tube output, including blocked chest tubes and clotted hemothorax.
In the thoracotomy-alone group, diaphragmotomy was performed in 24 of 40 patients (60.0%). In the laparotomy-alone group, diaphragmotomy was performed in 2 of 34 patients (5.9%). This maneuver allowed transdiaphragmatic assessment of the opposite cavity and helped avoid unnecessary additional surgical exploration in selected patients.
The distribution of injured structures is presented in Table 4.
Table 4
Injured structures in thoracoabdominal injuries
One missed injury was identified: an esophageal injury in a critically unstable patient with a gunshot wound. Damage control surgery was performed in 2 patients: 1 patient with exsanguinating liver trauma and 1 patient with esophageal injury requiring exploration of three cavities.
The overall in-hospital mortality rate was 0.9%, with 1 death among 111 patients.
Discussion
Penetrating thoracoabdominal injuries remain a complex diagnostic and therapeutic challenge because the surgeon must rapidly determine which cavity should be explored first: the chest or the abdomen. Although many penetrating thoracic injuries can be managed with thoracic drainage alone, thoracoabdominal trauma is different because life-threatening injuries may involve either cavity or both cavities simultaneously.
In critically unstable patients, the choice of the initial operative approach is especially important. One of the largest series of thoracoabdominal injuries was reported by Asensio et al. in 2002 and included 254 patients. The authors showed that mortality was higher in patients who required exploration of both cavities [2]. This emphasizes that the need for dual-cavity exploration is a marker of injury severity and highlights the importance of careful initial surgical decision-making.
The reported mortality rate in thoracoabdominal injuries ranges from 13% to 40% [3, 4] and depends on the mechanism of injury, injury severity, and whether the trauma occurs in a civilian or military population. In our study, the in-hospital mortality rate was 0.9%. This low mortality rate may be partly explained by the predominance of stab wounds in our cohort, whereas many previously published series included a higher proportion of gunshot wounds. In addition, all patients included in this study were admitted alive and underwent emergency surgical management, which should be considered when interpreting survival.
The central question of this study was whether penetrating thoracoabdominal injuries should be approached as “belly first” or “chest first”. Our findings suggest that this question cannot be answered by a universal rule. According to the initial operative approach, 69 patients were managed in the chest-first group and 42 patients in the abdomen-first group. Single-cavity exploration was more common in the abdomen-first group than in the chest-first group, whereas additional operative access and opposite-cavity exploration were more frequent in the chest-first group. However, this should not be interpreted as evidence that one strategy is superior to the other. Rather, it reflects the clinical complexity of thoracoabdominal trauma and the fact that patients selected for chest-first intervention often had signs suggesting thoracic bleeding or immediate thoracic threat.
Importantly, incorrect initial cavity selection occurred in only 6 patients, representing 5.4% of the total cohort. The rate of incorrect initial cavity selection did not differ significantly between the chest-first and abdomen-first groups. This finding supports the concept that both approaches may be safe when selected according to the patient’s physiological condition, suspected source of bleeding, imaging findings, chest tube output, and mechanism of injury. Therefore, the key decision is not whether the chest or abdomen should always be prioritized, but whether the initially selected cavity corresponds to the dominant life-threatening injury.
Diagnostic assessment plays a major role in this decision. Chest radiography was the most frequently used diagnostic modality in our series, especially during the earlier years of the study period. However, eFAST has become a first-line tool for rapid assessment of the abdominal, pleural, and pericardial cavities [5–7]. When the patient’s physiological condition allows, chest computed tomography remains the most informative diagnostic modality for stable patients with suspected thoracoabdominal trauma. Nevertheless, imaging may be misleading, and clinical reassessment remains essential.
In our series, thoracoscopy was performed in 42 hemodynamically stable patients. Diaphragmatic injury was visualized in 14 of these patients. Although diaphragmotomy does not always allow complete inspection of the opposite cavity, in selected cases it may provide sufficient transdiaphragmatic assessment. Diaphragmotomy was performed in 26 patients and helped avoid unnecessary additional surgical exploration. No missed injuries or complications were observed in this subgroup.
In some patients, surgery in one cavity must be interrupted in order to proceed urgently to the opposite cavity. In our cohort, this occurred in 6 cases. The main reasons were misinterpretation of chest radiography or chest tube output, including blocked chest tubes and clotted hemothorax. These cases illustrate that the initial operative strategy should remain flexible. A surgeon should be prepared to change the operative plan immediately when intraoperative findings or the patient’s physiology indicate that the dominant source of bleeding lies elsewhere.
Close communication with the anesthesiology team is essential during these procedures. Effective resuscitation, bleeding control, prevention of hypothermia, and correction of coagulopathy are critical while surgery is being performed in a critically unstable patient. In unstable patients, surgical management should follow damage control principles [6, 7]. In our study, 2 patients required damage control surgery: one patient with exsanguinating liver trauma and one patient with an esophageal injury in whom three cavities were explored.
This study has several limitations. It was retrospective and covered a long 23-year period, during which diagnostic tools and trauma management strategies evolved. In addition, some comparisons were based on aggregated data, and the chest-first and abdomen-first groups were not randomized. Therefore, differences between these groups should be interpreted descriptively and not as proof of superiority of one approach over the other.
Penetrating thoracoabdominal injuries may be regarded as a diagnostic and therapeutic “black box” with two potential entry points. One approach may be correct and the other may not; however, in selected cases, both cavities must be explored in order to control bleeding, repair injuries, and save the patient’s life. The optimal strategy is not a fixed chest-first or abdomen-first rule, but rapid identification of the most likely life-threatening cavity with readiness for immediate conversion when the initial assessment proves incorrect.
Conclusions
Penetrating thoracoabdominal injuries represent a complex surgical problem in which neither a universal chest-first nor abdomen-first strategy can be recommended. The initial operative approach should be determined by the patient’s hemodynamic status, suspected source of bleeding, imaging findings, chest tube output, and mechanism of injury.
In this cohort, both chest-first and abdomen-first approaches were associated with a low rate of incorrect initial cavity selection and very low in-hospital mortality when applied according to clinical indications. However, some patients required urgent conversion to the opposite cavity, emphasizing the need for continuous reassessment and surgical flexibility.
The key principle in managing penetrating thoracoabdominal trauma is rapid identification of the dominant life-threatening injury, prompt control of bleeding, and readiness to explore the second cavity when required.
