Penetrating Abdominal Injuries: A 10-Year Retrospective Analysis of 49 Patients at a German Trauma Center

Introduction

Penetrating abdominal injuries from stabbing or gunshot wounds are uncommon in central Europe. Thus, outside of level 1 trauma centers, most trauma surgeons lack experience in the diagnosis and treatment of penetrating abdominal injuries. In Germany, penetrating injuries show an incidence of 5%; of those, 36% were stab wounds and 12% were gunshot wounds, with abdominal involvement in 30% [1].

In contrast, the incidence of penetrating injuries in other countries such as the USA (38%) or South Africa (80%) is much higher [2,3]. However, the incidence of firearm associated assaults had been increased in Germany by 6% in 2024 compared to 2023. This information is available at https://www.bka.de/DE/AktuelleInformationen/StatistikenLagebilder/PolizeilicheKriminalstatistik/pks_node.html.

Management of penetrating abdominal wounds has been debated for decades. In the past and following international guidelines, all penetrating abdominal wounds required surgical exploration [4]. Since the 1960s, selective non-operative management (SNOM) has been reported as a possible treatment in stable patients to reduce complications [5]. Fabian et al found that only 50% of all abdominal stab wounds penetrate the peritoneum and that only 20–40% of those cause significant damage requiring further surgical treatment [6]. These data resulted in development of selective algorithms for the treatment of penetrating abdominal injuries in patients without signs of shock, peritonitis, or evisceration [7].

Biffl et al proposed guidelines for patients with abdominal stab wounds, recommending initial local wound exploration in stable patients. If there is peritoneal involvement but absence of clinical signs, the patient is admitted for serial clinical assessments [8].

Due to the low incidence of stab and gunshot wounds in Germany, there is no standard approach to treatment of these injuries. Implementation of the SNOM algorithm was shown to be poor in low-volume settings and routine use of laparotomy and laparoscopy are still largely debated even for non-symptomatic patients [9]. With this background, treatment algorithms must be established in European trauma departments to identify penetrating abdominal injuries requiring surgical repair and avoid unnecessary laparotomy with its associated morbidity.

This study sought to evaluate the effectiveness of preoperative diagnostics in predicting intraoperative findings in patients with penetrating abdominal injuries at a level 1 trauma center. The findings are discussed in the context of recent publications, and the indications and significance of laparotomy are elucidated.

Material and Methods

DATA:

Clinical data, including injury details, diagnostics, surgical procedures, and intraoperative and preoperative CT findings, were analyzed. Outcomes assessed were morbidity, mortality, hospital stay, and complications. For hemodynamically stable patients, initial diagnostics included clinical examination, Focused Assessment with Sonography for Trauma (FAST), and CT scan. FAST was defined as positive if intra-abdominal fluid or solid-organ injury was detected. CT findings were classified as equal, less severe, or more severe than laparotomy results, with intraoperative findings as the standard.

STATISTICS:

Statistical analysis was performed using SPSS® 15, with descriptive data presented as means±standard deviation (SD) or frequencies (%). Statistical significance was considered with P values of <0.05. Because of the non-normal distributed data, we used Spearman correlations.

Results

PATIENT DEMOGRAPHICS:

Within 10 years, 49 patients with penetrating abdominal injuries were admitted to our level I trauma center. Due to emergency surgery and lack of preoperative CT, 7 patients were excluded from the study because a comparison from CT with surgery was not possible. Figure 1 demonstrates the annual admission rate of cases in the observed time period.

Thirty-five (83.3%) of the patients were male and the mean age of the patients was 39.9±16.1 years. The cause of injury was stab wound in 35 (83.3%) cases and gunshot wound in 7 (16.7%). Most of the injuries were caused by interpersonal violence (64.3%) and 15 (35.7%) by failed suicide attempt. Additional extra-abdominal injuries were present in 59.5% of the patients (Figure 2A).

At admission, 6 (15%) patients were ventilated with a tracheal tube. Of these, 5 (83.3%) had stab wounds. All patients with documented vital sign were stable at admission according to the German S3 guideline for polytrauma treatment [12], with a pressure >90 mmHG and a frequency of <100/min. The hemoglobin levels at admission ranged from 8.1 to 17.8 g/dl. Women showed nearly normal values, while 61.8% of male patients had significantly lower values. Also, serum lactate was high in 38.9% of our patients, and 50% of these had abdominal injuries. Known illnesses other than penetrating trauma had been documented in 19 (45.2%) patients, and 57.9% of these had an underlying psychiatric disease.

INTRA-ABDOMINAL CT FINDINGS: In 23 cases (54.8%) CT scans showed no intra-abdominal injuries. In 2 cases, intra-abdominal injuries were unclear but most liver injuries were described. Moreover, there was a positive correlation of gunshot wounds with pathological intra-abdominal findings in CT scans (P=0.018, r=0.364). Furthermore, injury characteristics showed differences between the stabwound and gunshotwound patient groups (Figure 3).

INTRAOPERATIVE ABDOMINAL FINDINGS: Intraoperatively, no intra-abdominal injury was found in 19 cases (45.2%). In most cases the bowel was injured, involving the ilium and colon. In 11 cases, there were at least 2 intra-abdominal injuries. Of these, 4 patients had injury of the ilium and colon (36.4%), and 3 others had liver laceration and another injury (Figure 2B; Table 1).

COMPARISON OF THE FINDINGS:

Retrospective preoperative diagnostics via CT and intraoperative finding showed the same number of injuries in 29 cases (69.1%). In 18 (62.1%) of these, no injuries were found during diagnostics or surgery. Of the 2 cases with questionable findings in the preoperative CT, 1 was confirmed and the other was disproved. There were 13 cases with a different diagnosis and from these, the CT showed in 9 cases (69.2%) less injuries and in 4 patients (9.5%) more injured organs than the surgical exposure. The sensitivity and specificity of the CT scan for correctly diagnosed intra-abdominal findings after penetrating (stab or gunshot) injuries were 52.6% and 82.6%, respectively.

If only stab wounds were considered, the CT scan was correct in 80% (28 patients) and showed the same damage as that confirmed by surgical exposure of the abdomen. For patients with gunshot wounds the picture was completely different, as only 1 (14.9%) showed the same injury pattern on CT scan and intraoperatively. In all other cases, the CT scan and intraoperative finding differed from each other. In 42.9% of cases the CT scan showed a higher injury pattern compared to surgical findings. The 2 groups had similar sex distributions.

EXTRA-ABDOMINAL INJURIES: Within our study cohort, we found extra-abdominal injuries in 25 cases (59.5%). Extra-abdominal injuries were found in 57.1% of patients with stab wounds and in 71.4% of patients with gunshot wound. Mostly, the neck and thorax were in 48% of the cases involved. Other Injuries were cutting damage on upper extremities with 36% and skull fractures with 24% (Table 2).

REVISION SURGERY:

A total of 31 revision surgeries were performed in 3 patients (7.1%). One patient required 26 revisions due to complications and died during hospitalization. The second patient underwent 4 additional surgeries, including 3 for infection and perforation, followed by stoma creation, and 1 for adhesiolysis and re-anastomosis 8 months later. The third patient, without an initial lesion, needed surgery for a hernia 8 months after the initial injury.

COMPLICATIONS:

Eleven of the patients (26.2%) had complications, such as pneumonia, acute kidney failure, abscesses, pleural effusion, and pancreatic fistula (Table 3). Complications occurring no more than twice were categorized as ‘others’. The 3 most common complications were anemia (5), respiratory insufficiencies (4), and shock (3); 63.6% of these patients had multiple complications.

The mechanism of injury affected complication rates, as 50% of gunshotinjured patients had 2 or more complications, which was positively correlated with the trauma mechanism (r=0.327, P=0.035)

LENGTH OF INTENSIVE CARE UNIT (ICU) STAY:

Every patient was constantly monitored in the ICU after admission. The median ICU stay was 4.9 days ±3.1 days. Individuals injured by gunshots required a prolonged stay at the ICU, which was 2 days longer on average than for patients with other types of injuries (6.1 days).

HOSPITAL STAYS:

Hospital stay duration varied significantly based on injury type and severity. The average stay was 10.1 days, with a median of 7 days. Patients with stab wounds were discharged 13 days earlier than those with gunshot injuries (7.9 vs 21 days). Statistical analysis showed a negative correlation between stab wounds and hospital stay (p=0.002, r=−0.469), and a positive correlation for gunshot wounds (p=0.002, r=0.469). A longer stay was also positively correlated with discrepancies between CT and intraoperative findings (P=0.006, r=0.293), and negatively when findings matched (p=0.145, r=−0.229). Unsurprisingly, intra-abdominal pathologies on CT were also associated with longer hospital stays.

Discussion

EPIDEMIOLOGY:

Most stab and gunshot wounds in Germany are related to suicide attempts and violent crimes [1], and the incidence of penetrating abdominal injuries is low; therefore, it is not surprising that this study included only 49 patients over 10 years. Registry data document that the percentage of penetrating injuries is only 4% of all injuries in Germany during the last 10 years [13], and compared to the USA this percentage is very low [2,3]. Olaogun et al also described 47.9% of penetrating injuries in a patient cohort in Nigeria, with 60.9% caused by gunshots [14]. In total, the German Federal Criminal Police Office reported 35 799 cases (+6.3%) of violations of the weapons law and 4687 cases (+5.5%) of gun-related incidents in 2023 compared to 2015 [15]. Similar reports are also found in other European nations [16,17]. Overall, the incidence of penetrating injuries is increasing in Europe. Our patients had a mean age of 39.9 (13.4–79.3±16.1) years, which is comparable with the patient collective from Malkomes, Mnguni, and Olaogun, with 36.2±14, 29.2±10.7, and 34.2±10.8 years [14,18,19].

In Germany, most penetrating traumas are caused by stab injuries, whereas in South Africa and many countries outside Europe, most are related to gunshot wounds. In the Ukrainian war, 85.6% of abdominal injuries are from shrapnel wounds and 10.1% are from gunshot wounds [20], and the abdominal injury pattern in war is completely different than in civilian settings, like in our study. Staged treatment with repetitive surgeries near the front line with a short time to surgery can reduce the mortality of injured patients to nearly 10% [21]. Our patient cohort had 83.3% stab injuries and 16.7% gunshot-related trauma, similar to that reported by Beltzer et al (79% vs 21%) [22]. For example, Mnguni et al reported 45.5% stab injuries and 54.6% gunshot injuries [19]. In the literature, 70–90% of patients with penetrating wounds were male [14,19]. In our study, 83.3% were male. Most of our patients were hemodynamical stable at admission. This is also reported in the literature, with only 6.9% of unstable patients reported by Malkomes et al and 11% reported by Mnguni et al [18,19].

Data from the German trauma registry in 2017 [23] showed that only 4.2% of all polytraumatized cases from 2006–2016 had penetrating trauma – 4.6% were suicide attempts and 2.4% were involved in violent crimes. In the abdominal penetrating injuries we examined, a different picture emerged. In our own results, violent crime was evident in 64.3% of cases and failed suicide in 35.7% of cases. Also, Malkomes et al and Pallett et al documented substantially higher rates of violent offenses (60% and 44%, respectively) and greater proportions of suicide intent (17.2% and 8%, respectively) [16,18].

The length of hospital stay is reported to have a relatively large range in various publications. In London, the length of stay was only 3.0 days [16], in Frankfurt it was 7.4 days [18], and it was 10.1 days according to our own data, indicating significantly longer stays in German hospitals. In our own results, the length of stay for stab wounds was 7.9 days. Only Tonus et al (2003) showed a longer length of stay (13.1 days), with data from 1989 to 1998 [24]. This is likely due to improved care options, advancements in medicine over time, and economic factors. In South Africa, the average length of stay was 9.2 (±10.8) days [19] and in Nigeria it was of 12.7 days [14], and Africa overall has more gunshot wounds than stab wounds. Gunshot wounds cause massive destruction. High-velocity projectiles transfer more energy to the abdominal organs, leading to larger injuries around the bullet path due to cavitation [19], which could explain the extended length of stay.

PREOPERATIVE DIAGNOSIS:

Multislice-CT scans are the preferred imaging method for quickly and safely diagnosing trauma patients, with a reported sensitivity and specificity of around 90% [25]. However, diagnostic uncertainty occurs in over 50% of cases [18]. Therefore, decisions regarding surgery or SNOM should not rely solely on CT results but should also consider clinical and laboratory findings. Our study found a lower sensitivity of 52.6%, meaning the CT scans correctly diagnosed about half of the patients, but CT had a high specificity of 82.6%.

CT SCAN VS INTRAOPERATIVE FINDINGS:

In our study, CT scan and intraoperative findings disagreed in 13 patients. In 70%, the CT showed less injuries than the surgery found. This also explains the poor sensitivity, but there was no disadvantage for the patients because the surgeon always performed a systematic inspection of the whole abdomen according to Barret et al [26].

ABDOMINAL INJURIES:

There is a wide range in incidence regarding peritoneum damage by stab injuries. Hildebrand et al reported this is happening only in 30%, in contrast to that Belzer et al, who reported about 50–70% requiring surgery [22,27].

As seen in our data, additional injuries are common (57.14%) and should not be overlooked. Thoracic stab wounds (8 out of 20) can have life-threatening consequences. In our data, the small and large intestines (30.4%) and the liver (75%) were most frequently affected by stab wounds. In the literature, liver injuries are reported to be up to 40% [28], and small- and large-intestine injuries are reported to be 45% [28] and 33.8% [22], respectively.

The injury pattern of gunshot wounds is different from that of stab wounds. A gunshot wound shows massive destruction. Due to shock waves and cavitation, intra-abdominal gunshot injuries often result in multiple injuries with increased morbidity. Consequently, these patients undergo more frequent re-operations [19]. Among our 3 patients with revisions, 2 had injuries from a gunshot, and one of them had to be revised a total of 26 times.

In our own analysis, the small and large intestines were also most frequently injured by gunshot wounds. This correlates with the literature, which reports small-intestine injuries in 50% and large-intestine injuries in 40% [28].

TREATMENT (SNOM, LAPAROSCOPY, OR LAPAROTOMY):

In our study group we did not use conservative treatment. Therefore, we have no data available for that, but Como et al reported that SNOM is a possible treatment for stable patients with no injury detected by CT or focused assessment with sonography for trauma (FAST) [7]. Como et al reported that most of the patients could be discharged after 24 h of observation. Furthermore, it is reported that 23% to 53% of all laparotomies are not necessary and that 2.5% to 41% of these have postoperative complications. Como et al concluded that with SNOM the rate of unnecessary laparotomies and following complications could be lowered and late-onset surgery was not associated with higher mortality, and other studies have reported similar data [19,29].

As an alternative, laparoscopy is possible, and the rate of false-negative surgery is reported to be 1.4%. Also, the rate of postoperative complications should be lower according to the literature [22]. But there is no consensus about which kind of treatment should be standard. A recent case series from the Ukrainian war by Lurin et al founded that laparoscopy is possible in stable patients with combat trauma [30]. We found a 45.2% rate of negative laparotomies and a 26.2% rate of complications, which agrees with the literature.

Conclusions

TAKE-HOME MESSAGES:

(1) Hemodynamic stability is the key determinant of management. Unstable patients with penetrating abdominal injuries require immediate surgical exploration via laparotomy, with no role for delayed or non-operative strategies. (2) Selective non-operative management (SNOM) is safe in carefully selected patients. Hemodynamically stable patients with stab wounds, negative CT and FAST findings, and reliable clinical examination can be managed non-operatively under strict observation without increased morbidity or mortality. (3) Laparoscopy should be considered the initial surgical approach in stable patients. It allows accurate injury assessment, reduces postoperative complications and hospital stay, and can be converted to open laparotomy if necessary. (4) Standardized treatment algorithms are essential. Clearly defined indications for operative and non-operative management help guide clinical decision-making, particularly for less experienced clinicians, and minimize unnecessary laparotomies.

LIMITATIONS:

This study has several limitations. The retrospective nature may have introduced selection bias and limited the completeness of available data. Some relevant clinical variables may not have been consistently documented, potentially affecting the accuracy of our findings. Additionally, the study was conducted at a single center, which may limit the applicability of the results to other settings. Variability in clinical practice over the study period could also have influenced outcomes. Prospective multicenter studies are needed to validate these findings.