Injury patterns and predictors of surgical intervention in road traffic accident patients presenting to the emergency department

Abstract

Background

Road traffic accidents (RTAs) impose substantial clinical and operational burdens on emergency departments (EDs). Identifying the injury patterns and predictors of surgical intervention is essential for trauma system planning and triage optimization. This study aimed to characterize the injury distribution and identify the independent predictors of surgical intervention in RTA patients presenting to a high-volume tertiary trauma center.

Methods

This retrospective cohort study evaluated all patients presenting to the ED following RTAs between January 1 and December 31, 2024. Data were obtained from electronic medical records. Variables included age, sex, initial hemodynamic status, mechanism of injury, Glasgow Coma Scale (GCS), Injury Severity Score (ISS), anatomical injury regions, imaging utilization, transfusion requirement, and hospital length of stay. The primary outcome was the requirement for surgical intervention, defined as any operative procedure or major therapeutic intervention requiring surgical team involvement, encompassing both traditional operating room procedures and life-saving bedside interventions. Secondary outcomes included injury distribution patterns across anatomical regions and clinical outcomes, such as mortality and patient clinical disposition.

Results

Of 4747 included patients, the median age was 26 years, and 74.7% were male. Motorcycle injuries (49.3%) were the leading mechanism. Surgical intervention was required in 470 patients (9.9%). Tibial (23.4%) and femoral (13.8%) fractures were the most frequent operative indications. Surgically managed patients had higher ISS (median 9 vs 2), greater transfusion needs (24.5% vs 0.4%), and longer hospital stays (p < 0.001). Age (OR 1.03), male sex (OR 1.43), hemodynamic instability on ED arrival (OR 4.63), lower GCS (OR 1.13), lower extremity trauma (OR 2.32), head-maxillofacial trauma (OR 1.52), abdominal or pelvic trauma (1.38), motorcycle crashes (compared with vehicle occupants, OR 2.03), and pedestrian crashes (compared with vehicle occupants, OR 1.67) independently predicted surgical intervention.

Conclusion

The need for surgical intervention in RTA patients is primarily driven by physiological status, specific injury mechanisms, and the distribution of anatomical injuries. Incorporating physiological status and mechanism-informed triage pathways into trauma protocols may enhance surgical readiness, optimize resource allocation, and improve care delivery in high-volume urban trauma centers.

Introduction

Road traffic accidents (RTAs) remain a significant global public health problem, resulting in approximately 1.2 million deaths and 50 million injuries each year worldwide [1–3]. These incidents represent a major cause of trauma-related morbidity and mortality worldwide and impose a considerable burden on healthcare systems, as they frequently result in injuries of varying severity that require extensive acute care, multidisciplinary management, and ongoing rehabilitation [4, 5]. According to national surveillance data, more than 1.4 million traffic accidents and over 260,000 injuries were recorded in 2024 [6]. These data underscore the substantial clinical and healthcare burden associated with RTAs.

From the emergency medicine perspective, traffic accidents represent a major cause of trauma-related admissions, often involving multi-trauma patients who require imaging, multidisciplinary consultations, and surgical interventions [5, 7, 8]. Such cases contribute substantially to emergency department (ED) overcrowding and resource strain, with surgical intervention generally required in patients with more severe or anatomically complex injuries, which can be reflected by higher trauma severity indicators such as the Injury Severity Score (ISS) [9–11]. This clinical and operational burden underscores the need for detailed epidemiological and clinical insights to better understand which patients are most likely to require surgical management. Large-scale investigations into the demographic and clinical profiles of RTA patients in EDs remain scarce, particularly in the context of evolving traffic dynamics, such as the rising prevalence of motorcycle use [12]. Moreover, data specifically characterizing the clinical and injury patterns associated with the need for surgical intervention are notably limited.

This study aimed to determine the injury patterns and identify the independent predictors of surgical intervention among RTA patients presenting to the ED of a tertiary trauma center in a major metropolitan city. By identifying factors associated with surgical intervention, we aim to contribute to a deeper understanding of urban trauma epidemiology and inform targeted strategies for prevention and emergency management.

Materials and methods

Study design and settings

This study is a single-center, retrospective cohort study conducted at the Adult ED of Marmara University Pendik Training and Research Hospital, a level 1 trauma center equipped with advanced diagnostic and surgical facilities in Istanbul, Türkiye. The study protocol was approved by Marmara University Clinical Research Ethics Committee (protocol number: 09.2025–25-050). The study was conducted following the principles of the Declaration of Helsinki, and the reporting adhered to the STROBE guidelines for observational studies.

Study participants

All patients who presented to the ED between January 1 and December 31, 2024, due to RTAs were retrospectively reviewed. Patients were included if they presented to the ED following any traffic accident mechanism, including in-vehicle, motorcycle, pedestrian, bicycle, scooter vs incidents, and if complete demographic and outcome data were available in the hospital information system. Patients were excluded if outcome data were missing, if they were discharged against medical advice (DAMA) before completion of their initial evaluation, or if they presented solely for medico-legal documentation without a clinical assessment. All data were obtained from the hospital’s electronic medical record system in accordance with the approved study protocol.

Variables and data collection

The study variables included demographic, clinical, radiological, therapeutic, and outcome parameters. Demographic variables consisted of age, sex, mechanism of injury, and comorbidities, which were assessed using the Charlson Comorbidity Index (CCI) [13]. Clinical variables included the Glasgow Coma Scale (GCS) score obtained at admission. Hemodynamic instability on ED arrival was defined as a systolic blood pressure < 90 mmHg or clinical signs of persistent shock requiring immediate fluid or blood resuscitation. The mechanism of injury was classified as in-vehicle, motorcycle, pedestrian, or other (e.g., bicycle, scooter, or all-terrain vehicle) accidents. Anatomical injury regions were classified as head-maxillofacial-neck, thoracic, abdominal/pelvic, spinal trauma, upper extremity, and lower extremity. Multiple regional injury was defined as trauma involving two or more anatomical regions. The ISS was also calculated to quantify the overall severity of trauma [14].

The use of imaging modalities, including computed tomography (CT) and plain radiography (X-ray), was recorded for each patient. Surgical intervention and transfusion requirements were also documented. The length of hospital stay was categorized as less than 8 hours, between 8 and 24 hours, or more than 24 hours after ED admission.

Outcome variables included discharge, hospitalization in a ward, admission to the intensive care unit (ICU), or in-hospital death. All-cause mortality was defined as any death occurring during hospitalization or within 30 days following ED presentation.

Outcomes

The primary outcome was the requirement for surgical intervention. “Surgical intervention” was defined as any operative procedure or major therapeutic intervention requiring surgical team involvement, including but not limited to fracture fixation, craniotomy, laparotomy, extracorporeal membrane oxygenation (ECMO), and tube thoracostomy. This includes both operating room procedures and life-saving bedside procedures. Secondary outcomes included injury distribution patterns across anatomical regions and clinical outcomes, such as mortality and patient clinical disposition.

Statistical analysis

Statistical analyses were performed using Jamovi software (version 2.6.26, Windows platform). The distribution of continuous variables was assessed visually using histograms and Q–Q plots. Categorical variables were summarized as frequencies and percentages. Numerical variables with normal distribution were expressed as mean ± standard deviation (SD), whereas those without normal distribution were presented as median and interquartile range (IQR). Group comparisons were conducted using the Chi-square test for categorical variables. For continuous variables, the independent samples t-test or Mann–Whitney U test was applied in group comparisons. A univariate logistic regression analysis was initially performed to identify potential predictors of surgical intervention. Variables with p < 0.1 in univariate analysis and those deemed clinically relevant were entered into a multivariate logistic regression model. Multicollinearity among independent variables was assessed before model inclusion. A p-value < 0.05 was considered statistically significant in all analyses.

Results

A total of 4917 patients presented to the Adult ED of Marmara University Pendik Training and Research Hospital with RTA-related injuries between January 1 and December 31, 2024. Of these, six patients were transferred to another facility, 126 patients left the hospital without a documented outcome (DAMA), and 38 patients presented solely for medico-legal documentation. Consequently, a total of 4747 patients with RTA-related injuries were analyzed during the study period, with a median age of 26 years (20–38). The cohort predominantly consisted of male patients (n = 3546, 74.7%), while female patients accounted for 25.3% (n = 1201). Among these cases, 1368 (28.8%) were vehicle occupants (eg, car, truck, or lorry), 2340 (49.3%) were motorcycle riders or passengers, 841 (17.7%) were pedestrians, and 198 (4.2%) were classified as other mechanisms. Overall, 470 patients (9.9%) underwent surgical intervention during hospitalization. Regarding patient outcomes, 3921 patients (82.6%) were directly discharged from the ED, while 243 patients (5.1%) were boarded in the ED (access block) and discharged after extended observation due to inpatient bed unavailability. Additionally, 520 patients (11.0%) were admitted to hospital wards, 50 patients (1.1%) required ICU admission, and 13 patients (0.3%) died in the ED. All-cause mortality rate for the cohort was 0.7% (n = 32). Patient flow and clinical outcomes are summarized in Fig. 1.

Fig. 1.

Flowchart of patient selection and distribution of outcomes

Of the 4747 trauma patients, 470 (9.9%) required a surgical intervention, while 4277 (90.1%) were managed conservatively. Surgically treated patients were older (median age: 29 [21–45] vs. 25 [20–37] years; p = 0.025) and predominantly male (80.6% vs. 74.0%; p = 0.002). Surgical intervention rates differed significantly between injury mechanisms (p < 0.001), with the highest proportions observed in pedestrian (12.1%) and motorcycle (11.7%) injuries, followed by vehicle occupants (5.9%) and other mechanisms (6.2%). Lower GCS scores (median 15 [13–15] vs. 15 [15–15]; p < 0.001) were more frequent in those requiring surgery. Patients who underwent surgery had higher ISS and CCI values compared to those managed without surgery. The median ISS was 9 (9–16) in the surgical group and 2 (2–9) in the non-surgical group (p < 0.001). The median CCI was 0 (0–1) in the surgical group and 0 (0–0) in the non-surgical group (p < 0.001). Anatomical injury patterns such as lower extremity, thoracic, abdominal/pelvic, spinal, and multiple regional injury were each significantly associated with surgical intervention (all p < 0.05). Imaging utilization was higher in the surgical group, with 93.8% undergoing X-ray and 93.4% undergoing CT (both p < 0.001). Patients who underwent surgical intervention demonstrated a markedly more severe clinical profile; they were significantly more likely to present with hemodynamic instability on ED arrival (5.3% vs. 0.3%, p < 0.001), had substantially higher transfusion requirements (24.5% vs. 0.4%, p < 0.001), and experienced longer hospital stays (>8 h and >24 h, both p < 0.001), indicating a significantly greater clinical and resource burden compared with non-surgical patients. Surgical intervention outcomes according to clinical and anatomical variables are summarized in Table 1.

Table 1.

Comparison of demographic, clinical, and anatomical characteristics according to surgical intervention

Variable, n (%)	Total (n = 4747)	Surgical intervention (+) n = 470 (9.9%)	Surgical intervention (−) n = 4277 (90.1%)	OR (95% CI)	p value
Age, years (median, IQR)	26 (20–38)	29 (21–45)	25 (20–37)	1.02 (1.01–1.03)	0.025
Sex (Male)	3546 (74.7%)	379 (80.6%)	3167 (74.0%)	1.46 (1.15–1.85)	0.002
GCS (median, IQR)	15 (15–15)	15 (13–15)	15 (15–15)	0.76 (0.70–0.81)	<0.001
Trauma score, median (IQR)
Injury Severity Score	4 (2–9)	9 (9–16)	2 (2–9)	1.43 (1.38–1.48)	<0.001
Charlson Comorbidity Index	0 (0–0)	0 (0–1)	0 (0–0)	1.56 (1.28–1.90)	<0.001
Anatomical region
Upper extremity trauma	2299 (48.4%)	254 (54.0%)	2045 (47.8%)	1.28 (1.06–1.55)	0.01
Lower extremity trauma	2764 (58.2%)	361 (76.8%)	2403 (56.2%)	2.58 (2.07–3.23)	<0.001
Head–maxillofacial–neck trauma	1989 (41.9%)	225 (47.9%)	1764 (41.2%)	1.31 (1.08–1.58)	0.006
Thoracic trauma	975 (20.5%)	130 (27.7%)	845 (19.8%)	1.55 (1.25–1.93)	<0.001
Abdominal/pelvic trauma	1067 (22.5%)	151 (32.1%)	916 (21.4%)	1.74 (1.41–2.14)	<0.001
Spinal trauma	731 (15.4%)	92 (19.6%)	639 (14.9%)	1.39 (1.09–1.77)	0.008
Multiple regional injuries	2875 (60.6%)	337 (71.7%)	2538 (59.3%)	1.74 (1.41–2.14)	<0.001
Hemodynamic instability on ED arrival	39 (0.8%)	25 (5.3%)	14 (0.3%)	17.04(8.77–33.00)	<0.001
Transfusion requirement	132 (2.8%)	115 (24.5%)	17 (0.4%)	81.2 (48.2–137)	<0.001
X-ray	3769 (79.4%)	441 (93.8%)	3328 (77.8%)	4.34 (2.96–6.36)	<0.001
Computed Tomography	3306 (69.6%)	439 (93.4%)	2867 (67.0%)	6.96 (4.81–10.1)	<0.001
Injury mechanism				—	<0.001
├─ Vehicle occupants	1368 (28.8%)	81 (17.2%)	1287 (30.1%)	—
├─ Motorcycle	2342 (49.3%)	275 (58.5%)	2067 (48.3%)	—
├─ Pedestrian	842 (17.7%)	102 (21.7%)	740 (17.3%)	—
└─ Other	195 (4.1%)	12 (2.6%)	183 (4.3%)	—
Length of stay
Hospital stay > 8 h	816 (17.2%)	451 (96.0%)	365 (8.5%)	45.3 (35.1–58.3)	<0.001
Hospital stay > 24 h	595 (12.5%)	443 (94.3%)	152 (3.6%)	47.0 (36.7–60.2)	<0.001

Among 470 patients who underwent surgical intervention, the lower extremity was the most frequently operated anatomical region in 256 patients (54.5%), followed by the upper extremity in 111 patients (23.6%) and the head–maxillofacial–neck region in 63 patients (13.4%). Thoracic, abdominal/pelvic, and spinal surgeries were performed in 37 (7.9%), 31 (6.6%), and 19 (4.0%) patients, respectively. 45 patients (9.6%) underwent surgical procedures involving two anatomical regions, while one patient (0.2%) required surgery in three distinct anatomical regions. Among all patients who underwent surgical intervention (n = 470), the most common indications were tibial fractures (110 patients, 23.4%) and femoral fractures (65 patients, 13.8%). In the head–maxillofacial–neck region, intracranial hemorrhages constituted 34.9% of surgeries, including epidural hematomas (14.3%) and subdural hematomas (15.9%), while maxillofacial fractures represented 47.6%, most frequently involving the mandible (15.9%) and zygomatic bone (14.3%). In the thoracic region, tube thoracostomy was performed in 81.1% of patients, with pneumothorax (62.2%) and hemothorax (51.4%) being the main indications. In abdominal and pelvic trauma, pelvic injuries (45.2%) and solid organ injuries (38.7%) were the predominant causes for surgery, most commonly involving the spleen (66.7%) among solid organ cases. In upper extremity trauma, operations were most frequently performed for radius fractures (36.9%), ulna fractures (30.6%), and humerus fractures (18.0%). In the lower extremity, tibial fractures (42.9%) and femoral fractures (25.4%) were the leading indications for operative management. The detailed distribution of surgical procedures by anatomical region is presented in Fig. 2.

Fig. 2.

Anatomical distribution of surgical procedures in road traffic accident patients

In multivariate logistic regression analysis, older age, male sex, lower GCS, hemodynamic instability on ED arrival, lower extremity trauma, head–maxillofacial–neck trauma, abdominal/pelvic trauma, motorcycle crashes compared with in-vehicle crashes, and pedestrian crashes compared with in-vehicle crashes were independently associated with an increased likelihood of surgical intervention (all p < 0.05) (Table 2). Hemodynamic instability on arrival was the strongest independent predictor, with unstable patients 4.63 times more likely to require surgical intervention than those who were hemodynamically stable. Each one-year increase in age was associated with a 3% higher probability of requiring surgery, and male patients were 1.43 times more likely to undergo a surgical procedure than females. For clinical severity, each 1-point decrease in GCS corresponded to a 1.13-fold increase in the likelihood of surgery. Regarding anatomical injury patterns, the presence of lower extremity trauma nearly doubled the probability of surgical intervention, and head–maxillofacial–neck trauma increased it by approximately 1.5 times. Abdominal/pelvic trauma was also a significant predictor, increasing the likelihood of surgery by approximately 1.4 times. Among injury mechanisms, motorcycle crashes were associated with approximately a two-fold increase in surgical intervention compared with in-vehicle crashes, and pedestrian crashes increased the probability by approximately 1.7 times.

Table 2.

Multivariate logistic regression analysis for predictors of surgical intervention

Variable	OR (95% CI)	p value
Age	1.03 (1.02–1.03)	<0.001
Male sex	1.43 (1.09–1.86)	0.009
Glasgow Coma Scale	1.13 (1.03–1.25)	0.009
Hemodynamic instability on ED arrival	4.63 (1.93–11.05)	<0.001
Upper extremity trauma	1.01 (0.82–1.24)	0.912
Lower extremity trauma	2.32 (1.82–2.96)	<0.001
Head–maxillofacial–neck trauma	1.52 (1.22–1.88)	<0.001
Thoracic trauma	1.15 (0.89–1.48)	0.295
Abdominal/pelvic trauma	1.38 (1.09–1.74)	0.007
Spinal trauma	0.99 (0.94–1.32)	0.939
Motorcycle crash (vs. vehicle occupant)	2.03 (1.49–2.76)	<0.001
Pedestrian crash (vs. vehicle occupant)	1.67 (1.19–2.31)	<0.001
Other mechanisms (vs. vehicle occupant)	1.15 (0.60–2.19)	0.673

When patients were compared according to mortality, non-survivors were significantly older than survivors (median 33 [21–52] vs 26 [20–37] years, p < 0.001). Male predominance was higher among non-survivors (90.6% vs 74.6%, p = 0.038). Lower GCS scores were markedly more frequent in fatal cases (median GCS 7 [3–14] vs [15–15], p < 0.001). The median ISS and CCI values were significantly higher among non-survivors compared to survivors (both p < 0.001). The ISS was 41 (24–75) in non-survivors and 4 (2–9) in survivors. The CCI was 1 (0–2) in non-survivors and 0 (0–0) in survivors (Table 3). There was no statistically significant difference in mortality between different injury mechanisms (p = 0.743).

Table 3.

Comparison of Charlson comorbidity Index (CCI) scores between survivors and non-survivors

CCI	Survivors (n = 4715)	Non-survivors (n = 32)	Total (n = 4747)
0	3749 (79.5%)	13 (40.6%)	3762 (79.3%)
1	352 (7.5%)	9 (28.1%)	361 (7.6%)
2	256 (5.4%)	2 (6.3%)	258 (5.4%)
3	155 (3.3%)	2 (6.3%)	157 (3.3%)
4	142 (3.0%)	4 (12.5%)	146 (3.1%)
5	61 (1.3%)	2 (6.3%)	63 (1.3%)

Discussion

In this large, single-center cohort of 4747 patients with RTA-related injuries, only one in ten patients required surgical intervention, yet this subgroup represented the most clinically complex segment of the cohort, with higher injury severity, greater transfusion requirements, and longer hospital stays. Surgically managed patients were older, more often male, more likely to present with hemodynamic instability, had lower GCS scores, higher ISS values, more frequent multiregional injuries, and markedly higher transfusion needs, which together underline that the need for surgery is a robust marker of trauma burden and resource utilization. The very low overall all cause 30-day mortality of 0.7% must also be interpreted in the context of this case mix: most patients had relatively low ISS, whereas non-survivors had very high ISS and CCI values, as well as severely depressed GCS scores, indicating that mortality was concentrated in a small but extremely high-risk subgroup. Although mechanisms of injury did not differ significantly between survivors and non-survivors, surgical patients who died were older and more comorbid, suggesting that host factors modulate outcomes even when definitive trauma and surgical care are available.

Several of our findings align with and extend prior work on predictors of surgical intervention and outcomes in traffic-related trauma. Our multivariate analysis identified hemodynamic instability on arrival as the strongest independent predictor of surgical intervention. This finding emphasizes that initial physiological derangement, more so than anatomical injury patterns alone, serves as a primary driver for operative decision-making. In our cohort, the presence of shock on admission effectively prioritized patients for major therapeutic interventions, reinforcing the role of rapid physiological assessment in the high-stakes environment of a busy trauma center. Yamamoto et al. developed the SIM score in a nationwide Japanese trauma registry and showed that age, mechanism of injury, and initial physiological derangement predict the need for emergent laparotomy or thoracotomy in motor vehicle crash victims, with higher scores associated with sharply increasing probabilities of surgery [15]. Our data similarly identify older age, motorcycle and pedestrian mechanism, and markers of physiologic and anatomical severity (hemodynamic instability, lower GCS and specific regional injuries) as independent correlates of surgical treatment, but in a broader clinical definition of surgery that also includes orthopedic and neurosurgical procedures. The findings regarding injury mechanisms offer a specific contribution to the literature by reflecting the evolving trauma landscape of a megacity like Istanbul. Unlike many Western cohorts, where occupant injuries predominate, our data shows a significant surgical burden from ‘vulnerable road users’. The nearly two-fold increase in surgical risk for motorcycle and pedestrian accidents (OR 2.03 and 1.67, respectively) underscores a critical public health issue in Türkiye. This trend is reflected in recent national statistics, which indicate that motorcycle involvement in traffic accidents has surged by approximately 33% over the last year, outpacing traditional traffic safety measures [6]. Our study demonstrates that this epidemiological shift directly translates into an increased surgical workload, echoing a broader global trend where rapid urbanization and the proliferation of commercial motorcycle use are reshaping trauma profiles in middle-income countries. This convergence suggests that simple combinations of readily available variables can consistently stratify surgical need across different trauma systems [16].

The structure of outcomes we observed, with most patients managed non-surgically and a minority requiring surgery or dying, is also consistent with the three-way categorization of non-surgical, surgical, and fatal trajectories described by Oishi et al. for traumatic brain injury after road traffic crashes in Bangladesh [17]. In that study, male sex, older age, motorcycle use, and other injury-related factors increased the probability of both surgical and fatal outcomes [17]. Our findings extend this concept to a heterogeneous trauma population by showing that initial physiological status (hemodynamic instability), age, sex, and mechanism remain important even when all anatomical regions and surgical disciplines are considered together, and by documenting the specific operative indications and anatomical regions that dominate in a busy urban trauma center.

The low mortality observed in our cohort appears to be at the lower end of rates reported from many hospital-based series of road traffic injuries in low- and middle-income settings, where overall case fatality rates are commonly around 2%, while substantially higher proportions, often exceeding 10%, have been reported among patients with traumatic brain injury [17–19]. Although differences in case mix, inclusion criteria, and follow-up periods preclude direct comparison, this pattern is compatible with the established literature showing that care in organized trauma systems and level 1 trauma centers reduces both in-hospital and one-year mortality when compared with non-trauma centers. MacKenzie et al. demonstrated that level 1 trauma center care in the United States is associated with significantly lower in-hospital and 1-year mortality, even after adjustment for injury severity and patient characteristics [20]. Similarly, analyses of national datasets have shown lower mortality in regions with mature trauma systems and 24-hour surgical capability [6]. Our results support these observations in a different context by suggesting that a high-volume trauma center with rapid access to imaging, multidisciplinary teams, and operative services can achieve low early mortality despite managing a large share of the national RTA burden [21, 22].

An important nuance of our findings is the interaction between age, comorbidity, and outcome. Trauma is often framed as a disease of the young, and indeed, the median age in our cohort was 26 years. However, non-survivors were older, had substantially higher CCI scores, and were more frequently represented in higher CCI categories. This pattern is consistent with prior work showing that older adults and patients with significant comorbidity have higher mortality and poorer functional outcomes after traumatic brain injury and road traffic injuries, even when treated in tertiary centers [23–25]. Our data therefore reinforce the notion that older and multimorbid trauma patients constitute a distinct high-risk group, in whom the pathway from surgical indication to postoperative recovery is more fragile.

Taken together, these findings have several implications for clinical practice and trauma system design. First, they emphasize that surgical intervention in RTA patients is not merely a procedural endpoint but a proxy for high morbidity, high resource use, and elevated risk. Early recognition of patients who are likely to require surgery based on age, mechanism, GCS, hemodynamic status, ISS surrogates, and anatomical injury patterns could facilitate rapid triage to level 1 trauma centers, early notification of trauma surgical teams, and proactive allocation of operating room and blood bank resources. In settings where surgical expertise is not available around the clock, risk-stratification tools similar to the SIM score, adapted to local circumstances, could help identify patients who should be directly transported to higher-level centers to avoid delays in life- and limb-saving procedures [15, 26].

Second, from a health system perspective, the volume and representativeness of our cohort are noteworthy. According to recent national statistics, approximately 2% of all RTAs in Türkiye are treated at this ED, suggesting that the study population captures a meaningful fraction of the national urban trauma burden [6]. Within this context, the finding that lower extremity, head or maxillofacial, and abdominal or pelvic injuries, as well as motorcycle and pedestrian crashes, independently predict surgical intervention is clinically important because these patterns mirror the changing urban traffic profile, in which motorcycle and pedestrian injuries account for a large share of serious trauma. They highlight the types of injuries that most frequently drive operative decision-making in a high-volume trauma center. Building on this, optimizing care pathways at this single institution, including improvements in prehospital triage, interhospital transfer policies, and in-house workflows for such high-risk mechanisms, may translate into meaningful reductions in preventable morbidity and mortality at the national level [27, 28]. The low mortality observed in our data, when interpreted alongside national reports indicating an overall traffic-related hospital mortality rate of approximately 1.4%, supports continued investment in multidisciplinary trauma teams, standardized protocols, and rapid access to advanced surgical care as key components of injury control in Türkiye [6].

Third, our findings underscore the need for heightened vigilance in older and comorbid patients who undergo surgery after traffic-related trauma. The combination of advanced age, pre-existing disease burden, and high ISS appears to place patients at particular risk of death despite access to definitive surgical care. In this subgroup, a structured multidisciplinary perioperative management strategy, meticulous hemodynamic optimization, and careful postoperative monitoring are likely to be especially important. Integrating comorbidity and frailty assessments into trauma protocols and perioperative decision-making may help to identify patients who require more intensive monitoring or tailored postoperative pathways [29].

The present study also generates several hypotheses for future research. First, our multivariable model identifies a parsimonious set of variables that predict surgical intervention in a mixed trauma population. Prospective validation of this model in other centers could inform the development of a simple, context-specific prediction rule for use in prehospital or early ED triage. Second, more detailed work is needed to disentangle the relative contributions of mechanism, anatomical injury pattern, physiologic derangements, and comorbidity to both short-term mortality and longer-term outcomes such as disability, quality of life, and return to work.

Finally, given the aging of the population and rising prevalence of chronic diseases, future studies should specifically focus on geriatric and multimorbid trauma patients, examining optimal thresholds for surgery, perioperative risk stratification, and models of multidisciplinary postoperative care.

This study has several strengths. It is based on a large, consecutive cohort of road traffic-injured patients evaluated over a full calendar year at a level 1 trauma center with comprehensive emergency and surgical services. Data were extracted from an electronic medical record system that enabled systematic capture of key clinical variables, including GCS, ISS, and CCI, as well as detailed information on anatomical injury patterns, imaging, transfusions, and operative procedures. The analysis considered both surgical intervention and mortality, allowing simultaneous characterization of clinical burden and lethal risk. The fact that the study center accounts for a significant share of all traffic accidents in Türkiye underscores the relevance of the findings for national trauma planning.

At the same time, several limitations should be acknowledged. The single-center, retrospective design limits generalizability, particularly to rural settings or hospitals without advanced trauma capabilities. Selection and information biases are possible, despite the use of predefined inclusion criteria and standardized data extraction. Our operational definition of surgical intervention aggregates procedures of varying urgency and risk, ranging from emergent life-saving laparotomies to delayed orthopedic fixations. While the lack of granular ‘time-to-surgery’ data is a limitation, our primary objective was to identify patients who require any form of major surgical intervention. ISS, CCI, and other variables were calculated from in-hospital data and, therefore, cannot be directly applied in the prehospital environment without adaptation. Residual confounding is likely, as unmeasured factors such as prehospital time and socioeconomic status were unavailable. Finally, outcomes were limited to in-hospital and 30-day mortality; longer-term functional outcomes, complications, and quality of life were not assessed.

Despite these limitations, the present study provides granular, contemporary evidence from a high-volume trauma center in Türkiye on which traffic-injured patients are most likely to require surgery. These insights can inform both clinical decision-making and trauma system policy and highlight the need for particular attention to older and comorbid patients who undergo surgery after RTAs.

Conclusion

In conclusion, in this large single-center cohort of RTA patients, approximately one in ten required surgical intervention, and these patients represented the most clinically complex subgroup with higher injury severity, resource use, and worse short-term outcomes. Older age, male sex, hemodynamic instability on arrival, lower extremity injuries, head or maxillofacial injuries, abdominal or pelvic injuries, motorcycle crashes, and pedestrian crashes were independently associated with the need for surgery, while mortality clustered in older and comorbid patients with high ISS values. These findings suggest that the indication for surgery is a useful marker of trauma burden that can support early risk stratification and triage to level 1 trauma centers, and they highlight the particular vulnerability of older, multimorbid patients. Further multicenter studies are needed to validate these predictors and to clarify long-term outcomes, especially in this high-risk subgroup.

Abbreviations

CCI

Charlson Comorbidity Index

CT

Computed tomography

ED

Emergency department

GCS

Glasgow Coma Scale

ICU

Intensive care unit

IQR

Interquartile range

ISS

Injury Severity Score

OR

Odds Ratio

RTA

Road traffic accident

DAMA

Discharged against medical advice

Author contributions

**MA:** Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Project administration; Software; Visualization; Writing – original draft (lead). **EK:** Data curation; Formal analysis; Investigation; Methodology; Software; Visualization; Writing – original draft (supporting). **MBK:** Data curation; Formal analysis; Investigation; Methodology; Software; Visualization; Writing – original draft (supporting). **MAY:** Data curation; Formal analysis; Investigation; Writing – original draft (supporting) **AT:** Data curation; Formal analysis; Investigation; Writing – original draft (supporting) **FY:** Investigation; Writing – original draft (supporting) **AD:** Supervision; Validation; Visualization; Writing – review & editing (lead). All authors read and approved the final submitted version of the manuscript.

Funding

There is no funding.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of Marmara University Faculty of Medicine (Approval No: 09.2025–25-050). As this was a retrospective study based on anonymized clinical records, the requirement for informed consent was waived by the Ethics Committee. All data were handled in accordance with institutional privacy policies and ethical standards, and the study was conducted in compliance with the principles of the Declaration of Helsinki.

Consent for publication

All authors approved the submitted version for publication.

Competing interests

The authors declare no competing interests.