Venous thromboembolism in patients with traumatic pelvic and acetabular fractures: a retrospective cohort study Authors: Hasan M. Al-Dorzi, Ghadah K. AlHabs, Raseel Alorainy, Hind M. Almajed, Hissah Alsarra, Seham S. Alsalamah, Abdulrahman aldahlawi, Ibrahim Al Babtain Published: 2025 ABSTRACT BackgroundStudies on venous thromboembolism (VTE) in patients with traumatic pelvic and acetabular fractures have shown variable epidemiology, risk factors and outcomes.MethodsThis retrospective study evaluated patients with traumatic pelvic and/or acetabular fractures who were admitted to a Level 1 trauma center in Riyadh, Saudi Arabia between January 2016 and December 2022. We compared patients who had VTE to those who did not.ResultsThe study evaluated 437 patients with (median age: 37 years; 72.8% males). VTE occurred in 42 patients (9.6%), with deep-vein thrombosis and pulmonary embolism occurring at almost equal rates. Patients with VTE were older and more likely to receive tranexamic acid (16.7% vs. 9.9%, p = 0.06), red blood cell (26.2% vs. 13.2%, p = 0.02) and platelet (9.5% vs. 3.8%, p = 0.08) transfusions, and pelvic angioembolization (9.5% vs. 3.0%, p = 0.03). The most common associated injuries were chest injury (28.4%) and abdomen/pelvis injury (23.8%). Pharmacologic thromboprophylaxis was used in 94.1% of patients, predominantly enoxaparin. The significant predictors of VTE were age, admission to the intensive care unit (odds ratio 2.746, 95% confidence interval 1.041–7.240) and hospital length of stay. The timing of pharmacologic thromboprophylaxis was not significantly associated with VTE. Patients with VTE had similar hospital mortality compared with those without VTE (4.8% vs. 4.1%, respectively; p = 0.83) but stayed for a longer period in the hospital (median stay: 21.5 days vs. 12.0 days, respectively; p < 0.0001).ConclusionsVTE was diagnosed in almost 1 in 10 patients with traumatic pelvic and/or acetabular fractures. The need for admission to the intensive care unit, but not the timing of pharmacologic thromboprophylaxis, predicted VTE. Background Pelvic fractures are commonly caused by severe blunt trauma and account for up 6.9% of all traumatic fractures [1, 2]. They are often associated with other injuries [3]. The immediate complications of pelvic fractures include pelvis instability, injury to the adjacent organs and tissues, and extensive hemorrhage from the fractured bone, and/or injured veins and arteries. Pelvic fractures are also associated with venous thromboembolism (VTE), whether deep-vein thrombosis (DVT) or pulmonary embolism (PE) [4, 5]. The incidence of VTE in affected patients varied between studies from < 5% to almost 30% [4, 6, 7]. Most patients develop VTE with the first 7 days after injury [4, 5]. Risk factors for VTE include older age [4, 7], obesity [4, 8], angioembolization [9], associated injuries [7], delayed pharmacologic thromboprophylaxis [5, 6], and delayed surgery [4, 7]. Thromboprophylaxis has been shown to reduce the incidence of VTE in hospitalized patients with temporary lower limb immobilization after traumatic and nontraumatic injuries [10]. A systematic review that included 13 randomized clinical trials demonstrated that low molecular weight heparin (LMWH) decreased the risk of VTE compared to not taking pharmacological agents in these patients [10]. The 2018 guidelines of the American Society of Hematology recommended using pharmacological or mechanical thromboprophylaxis for surgical hospitalized patients depending on the risk of bleeding and suggested either LMWH or unfractionated heparin (UFH) in patients with low bleeding risk [11]. In addition, the guidelines suggested against using inferior vena cava filters as thromboprophylaxis for individuals undergoing major operative management [11]. The optimal timing of pharmacologic thromboprophylaxis is unclear and concerns about fracture-associated hemorrhage often delay the initiation of pharmacologic thromboprophylaxis. However, multiple studies favored early pharmacologic thromboprophylaxis. For instance, a study of 2752 patients with isolated pelvic fractures found that patients who received late thromboprophylaxis (> 48 h) had a higher incidence of VTE (4.3% vs. 2.2%, p = 0.004), and that LMWH was more protective than UFH [6]. Patients with traumatic pelvic injuries have substantial morbidity and mortality [6], making them an important focus for clinical research. Understanding the risk factors for VTE and the practices of thromboprophylaxis are vital to optimize patient management. However, there is limited research on VTE epidemiology, prevention, and outcome in patients with pelvic fractures in Saudi Arabia. This study aims to assess the incidence of VTE in patients with traumatic pelvic fractures, examine thromboprophylaxis practices, and determine VTE-associated outcomes with VTE at a Level 1 trauma center in Riyadh, Saudi Arabia. Methods Study design, patients and setting This study was a retrospective cohort of adult patients with traumatic pelvic and/or acetabular fractures who were admitted to King Abdulaziz Medical City in Riyadh between January 1, 2016, and December 31, 2022. This hospital was a tertiary care hospital and a Level 1 trauma center. Trauma cases were brought to the Emergency Department, where emergency physicians evaluated them. A Trauma Team was activated for severe injuries according to predefined criteria. This team would lead the subsequent evaluation and management. Routine radiologic evaluation included chest X-ray, pelvis X-ray and CT of brain, cervical spine, chest, abdomen, and pelvis. The orthopedic team was consulted for severe musculoskeletal injuries. Patients who needed organ support or close monitoring were admitted to the intensive care unit (ICU). The hospital had evidence-based guidelines for thromboprophylaxis, that included risk stratification and selection of the modality of prophylaxis based on VTE and bleeding risks [12]. Intermittent pneumatic compression devices were routinely applied to patients not receiving pharmacologic prophylaxis, especially in the ICU. Physiotherapists evaluated patients on admission and provided therapy sessions on weekdays tailored to the activity restrictions prescribed by the orthopedic team. In this study, we excluded patients with pathological pelvic fractures. The Institutional Review Board of the Ministry of National Guard Health Affairs approved this study. Due to the nature of the study, consent from patients was waived. Data collection We collected the following data: demographics, comorbidities, admission Glasgow Coma Scale, mechanism of injury, presence of pelvic fracture, acetabular fracture or both, associated injuries (brain, cervical spine, chest, abdomen/pelvis), and pertinent laboratory findings. We also assessed key interventions including intravenous tranexamic acid, blood transfusion, pelvic angioembolization, ICU admission, mechanical ventilation, surgical management of pelvic/acetabular fracture and its timing, pharmacologic thromboprophylaxis (enoxaparin, UFH, and other anticoagulants and timing), and insertion of a vena cava filter. Based on the computed tomographic description, we assessed the severity of pelvic fracture using the Abbreviated Injury Score related to pelvis (1 minor, 2 moderate, 3 severe nonvital, 4 severe vital, and 5 critical) [13]. The primary outcome of this study was the development of VTE during the hospital stay. Evaluation for VTE was at the discretion of the treating medical/surgical team, based on clinical suspicion. Diagnostic investigation included Doppler ultrasound of the lower limbs to assess DVT and CT pulmonary angiography to detect PE. Routine periodic surveillance for DVT was not part of the hospital’s thromboprophylaxis guidelines during the study period. Other studied outcomes were length of ICU and hospital stay, and ICU and hospital mortality. Statistical analysis We calculated the rate of VTE in all patients and in selected subgroups. We categorized patients into two groups: those who were diagnosed to have VTE during hospitalization and those who did not. Continuous data were presented as median with the first and third quartiles while categorical data as frequency with percentage. Differences in continuous data were assessed using the Student’s t-test or Mann Whitney U test depending on the normality of distribution. Differences in categorical data were assessed using the Chi-square test or Fisher’s exact test, as appropriate. Multivariable logistic regression was performed to assess the risk factors of VTE. Independent variables entered in the model were clinically relevant baseline characteristics and early management factors including age, ICU admission, presence of any chronic comorbidity, associated injuries, Abbreviated Injury Scale for pelvis, pelvic/acetabular fracture management (nonsurgical vs. surgical), pelvic angioembolization, mechanical ventilation, receiving tranexamic acid, red blood transfusion in the first day, timing of pharmacologic thromboprophylaxis (early within 2 days vs. later), as well as length of hospital stay. The results were presented as odds ratio with 95% confidence interval. A p-value of 0.05 or less was considered statistically significant. Data were analyzed using SPSS version 24 software (SPSS Inc., Chicago, IL, USA). Results Baseline characteristics The cohort included 437 patients with traumatic pelvic and/or acetabular fractures. Table 1 describes their demographics and clinical characteristics. Their median age was 37.0 years, and the majority were males (72.8%). Common comorbidities included diabetes (21.5%) and hypertension (20.4%). Previous history of VTE was present in only 2 patients (0.5%). The mechanisms of injury were motor vehicle collision in 69.8% and fall in 23.3%. A significant percentage of the pelvic fractures were comminuted (41.0%), displaced (34.1%) and associated with adjacent hematoma (24.3%). Associated injuries included traumatic brain injury (8.5%), cervical spine injury (3.2%), chest injury (28.4%), and abdomen/pelvis injury (23.8%).Table 1Baseline characteristics of the study patientsAll patientsN = 437VTEN = 42No VTEN = 395P-valueAge (years), median (Q1, Q3)37.0 (25.0, 62.0)48.0 (33.3, 62.5)37.0 (24.0, 61.0)0.09Male sex, n (%)Female sex, n (%)318 (72.8)119 (27.2)28 (66.7)14 (33.3)290 (73.4)105 (26.6)0.35Body mass index (kg/m2), median (Q1, Q3)Obese (≥ 30 kg/m2), n (%)26.4 (23.2, 30.7)122 (28.2)27.3 (24.7, 30.3)11 (26.2)26.3 (23.1, 30.7)111 (28.5)0.460.76Any chronic comorbidity, n (%)180 (41.2)22 (52.4)158 (40.0)0.12Hypertension, n (%)89 (20.4)9 (21.4)80 (20.3)0.86Diabetes, n (%)94 (21.5)10 (23.8)84 (21.3)0.70Dyslipidemia, n (%)61 (14.0)3 (7.3)58 (14.7)0.20Chronic kidney disease, n (%)21 (4.8)2 (4.8)19 (4.8)0.99Stroke, n (%)10 (2.3)0 (0.0%)10 (2.5)0.30Myocardial Infarction, n (%)7 (1.6)0 (0.0%)7 (1.8)0.38Heart failure, n (%)7 (1.6)0 (0.0%)7 (1.8)0.38Chronic respiratory disease, n (%)20 (4.6)4 (9.5)16 (4.1)0.11Chronic liver disease, n (%)9 (2.1)1 (2.4)8 (2.0)0.88Peripheral vascular disease, n (%)1 (0.2)0 (0.0%)1 (0.3)0.74Cancer, n (%)7 (1.6)0 (0.0%)7 (1.8)0.38Previous VTE, n (%)2 (0.5)1 (2.4)1 (0.3)0.052Pregnancy, n (%)1 (0.2)0 (0.0%)1 (0.3)0.74GCS admission, median (Q1, Q3)15.0 (15.0, 15.0)15.0 (15.0, 15.0)15.0 (15.0, 15.0)0.65Mechanism of Injury, n (%)Motor vehicle collisionFallOther305 (69.8)102 (23.3)20 (4.6)27 (64.3)12 (28.6)2 (4.8)278 (70.4)90 (22.8)18 (4.6)0.74Type of Fracture, n (%)Pelvic FractureAcetabular FractureBoth152 (34.8)199 (45.5)92 (21.1)14 (33.3)15 (35.7)13 (31.0)138 (34.9)184 (46.6)79 (20.0)0.840.180.01Fracture description, n (%)ComminutedDisplacedAdjacent hematomaAssociated proximal femur fracture179 (41.0)148 (34.1)106 (24.3)25 (5.7)20 (47.6)13 (31.0)12 (28.6)3 (7.1%)159 (40.3)135 (34.4)94 (23.8)22 (5.6)0.360.650.490.68Abbreviated Injury Score (pelvic fracture)Score could not be calculated for 63 patients3.5 ± 0.83.7 ± 0.83.4 ± 0.80.09Associated injuries, n (%)BrainCervical SpineChestAbdomen/Pelvis37 (8.5)14 (3.2)124 (28.4)104 (23.8)3 (7.1)0 (0)16 (38.1)15 (35.7)34 (8.6)14 (3.5)108 (27.3)89 (22.5)1.00.380.140.06Pertinent laboratory findings on admission, median (Q1, Q3)Hemoglobin (g/L)128 (111, 146)121 (104, 134)129 (111, 148)0.03Platelet count (109/L)281 (218, 341)289 (216, 332)280 (218, 342)0.72Partial thromboplastin time (sec)26.5 (24.6, 28.9)26.4 (24.7, 27.7)26.6 (24.6, 29.1)0.30Prothrombin time (sec)11.4 (10.9, 12.0)11.1 (10.5, 11.5)11.4 (10.9, 12.0)0.002International normalized ratio1.1 (1.0, 1.1)1.0 (1.0,1.1)1.1 (1.0, 1.1)0.004D-Dimer (μg/mL)16.4 (9.0, 35.1)16.4 (11.0, 32.7)16.0 (8.8, 35.1)0.99GCS, Glasgow coma scale; Q1, first quartile; Q3, third quartile; VTE, venous thromboembolism Table 1Baseline characteristics of the study patientsAll patientsN = 437VTEN = 42No VTEN = 395P-valueAge (years), median (Q1, Q3)37.0 (25.0, 62.0)48.0 (33.3, 62.5)37.0 (24.0, 61.0)0.09Male sex, n (%)Female sex, n (%)318 (72.8)119 (27.2)28 (66.7)14 (33.3)290 (73.4)105 (26.6)0.35Body mass index (kg/m2), median (Q1, Q3)Obese (≥ 30 kg/m2), n (%)26.4 (23.2, 30.7)122 (28.2)27.3 (24.7, 30.3)11 (26.2)26.3 (23.1, 30.7)111 (28.5)0.460.76Any chronic comorbidity, n (%)180 (41.2)22 (52.4)158 (40.0)0.12Hypertension, n (%)89 (20.4)9 (21.4)80 (20.3)0.86Diabetes, n (%)94 (21.5)10 (23.8)84 (21.3)0.70Dyslipidemia, n (%)61 (14.0)3 (7.3)58 (14.7)0.20Chronic kidney disease, n (%)21 (4.8)2 (4.8)19 (4.8)0.99Stroke, n (%)10 (2.3)0 (0.0%)10 (2.5)0.30Myocardial Infarction, n (%)7 (1.6)0 (0.0%)7 (1.8)0.38Heart failure, n (%)7 (1.6)0 (0.0%)7 (1.8)0.38Chronic respiratory disease, n (%)20 (4.6)4 (9.5)16 (4.1)0.11Chronic liver disease, n (%)9 (2.1)1 (2.4)8 (2.0)0.88Peripheral vascular disease, n (%)1 (0.2)0 (0.0%)1 (0.3)0.74Cancer, n (%)7 (1.6)0 (0.0%)7 (1.8)0.38Previous VTE, n (%)2 (0.5)1 (2.4)1 (0.3)0.052Pregnancy, n (%)1 (0.2)0 (0.0%)1 (0.3)0.74GCS admission, median (Q1, Q3)15.0 (15.0, 15.0)15.0 (15.0, 15.0)15.0 (15.0, 15.0)0.65Mechanism of Injury, n (%)Motor vehicle collisionFallOther305 (69.8)102 (23.3)20 (4.6)27 (64.3)12 (28.6)2 (4.8)278 (70.4)90 (22.8)18 (4.6)0.74Type of Fracture, n (%)Pelvic FractureAcetabular FractureBoth152 (34.8)199 (45.5)92 (21.1)14 (33.3)15 (35.7)13 (31.0)138 (34.9)184 (46.6)79 (20.0)0.840.180.01Fracture description, n (%)ComminutedDisplacedAdjacent hematomaAssociated proximal femur fracture179 (41.0)148 (34.1)106 (24.3)25 (5.7)20 (47.6)13 (31.0)12 (28.6)3 (7.1%)159 (40.3)135 (34.4)94 (23.8)22 (5.6)0.360.650.490.68Abbreviated Injury Score (pelvic fracture)Score could not be calculated for 63 patients3.5 ± 0.83.7 ± 0.83.4 ± 0.80.09Associated injuries, n (%)BrainCervical SpineChestAbdomen/Pelvis37 (8.5)14 (3.2)124 (28.4)104 (23.8)3 (7.1)0 (0)16 (38.1)15 (35.7)34 (8.6)14 (3.5)108 (27.3)89 (22.5)1.00.380.140.06Pertinent laboratory findings on admission, median (Q1, Q3)Hemoglobin (g/L)128 (111, 146)121 (104, 134)129 (111, 148)0.03Platelet count (109/L)281 (218, 341)289 (216, 332)280 (218, 342)0.72Partial thromboplastin time (sec)26.5 (24.6, 28.9)26.4 (24.7, 27.7)26.6 (24.6, 29.1)0.30Prothrombin time (sec)11.4 (10.9, 12.0)11.1 (10.5, 11.5)11.4 (10.9, 12.0)0.002International normalized ratio1.1 (1.0, 1.1)1.0 (1.0,1.1)1.1 (1.0, 1.1)0.004D-Dimer (μg/mL)16.4 (9.0, 35.1)16.4 (11.0, 32.7)16.0 (8.8, 35.1)0.99GCS, Glasgow coma scale; Q1, first quartile; Q3, third quartile; VTE, venous thromboembolism Baseline characteristics of the study patients Male sex, n (%) Female sex, n (%) 318 (72.8) 119 (27.2) 28 (66.7) 14 (33.3) 290 (73.4) 105 (26.6) Body mass index (kg/m2), median (Q1, Q3) Obese (≥ 30 kg/m2), n (%) 26.4 (23.2, 30.7) 122 (28.2) 27.3 (24.7, 30.3) 11 (26.2) 26.3 (23.1, 30.7) 111 (28.5) 0.46 0.76 Mechanism of Injury, n (%) Motor vehicle collision Fall Other 305 (69.8) 102 (23.3) 20 (4.6) 27 (64.3) 12 (28.6) 2 (4.8) 278 (70.4) 90 (22.8) 18 (4.6) Type of Fracture, n (%) Pelvic Fracture Acetabular Fracture Both 152 (34.8) 199 (45.5) 92 (21.1) 14 (33.3) 15 (35.7) 13 (31.0) 138 (34.9) 184 (46.6) 79 (20.0) 0.84 0.18 0.01 Fracture description, n (%) Comminuted Displaced Adjacent hematoma Associated proximal femur fracture 179 (41.0) 148 (34.1) 106 (24.3) 25 (5.7) 20 (47.6) 13 (31.0) 12 (28.6) 3 (7.1%) 159 (40.3) 135 (34.4) 94 (23.8) 22 (5.6) 0.36 0.65 0.49 0.68 Abbreviated Injury Score (pelvic fracture) Score could not be calculated for 63 patients Associated injuries, n (%) Brain Cervical Spine Chest Abdomen/Pelvis 37 (8.5) 14 (3.2) 124 (28.4) 104 (23.8) 3 (7.1) 0 (0) 16 (38.1) 15 (35.7) 34 (8.6) 14 (3.5) 108 (27.3) 89 (22.5) 1.0 0.38 0.14 0.06 GCS, Glasgow coma scale; Q1, first quartile; Q3, third quartile; VTE, venous thromboembolism Venous thromboembolism events VTE occurred in 42 patients (9.6%) on a median of 3 days after admission (the first and third quartiles: 1, 5). 18 patients had DVT alone, 17 PE alone and 7 both DVT and PE. VTE occurred before pelvic fracture surgery in 23 patients and after in 14 (one patient did not have surgery). Patients with VTE tended to be older (median 48.0 vs. 37.0 years, p = 0.09). More patients with VTE had at least one chronic comorbidity (52.4% vs. 40.0% in non-VTE, p = 0.12). Patients with VTE had lower hemoglobin levels (median 121 vs. 129 g/L) and slightly lower INR levels (1.0 vs. 1.1, p = 0.004). The Abbreviated Injury Scale related to pelvis was slightly higher in the VTE group (3.7 ± 0.8 versus 3.4 ± 0.8, p = 0.09). VTE occurred in 5 out of 25 patients (20.0%) with a score of 5 had VTE compared with 29 out of 349 patients (8.3%) with lower scores. Management Key management elements of the study patients are described in Table 2. 10.5% of all patients received IV tranexamic acid, with a higher percentage in the VTE group (16.7% vs. 9.9%, p = 0.06). Blood (26.2% vs. 13.2%, p = 0.02) and platelet (9.5% vs. 3.8%, p = 0.08) transfusions were more common in the VTE group. Pelvic angioembolization was performed more in the VTE group (9.5% vs. 3.0%, p = 0.03). For pelvic fracture management, 128 patients (29.4%) had conservative non-surgical management. The non-VTE group tended to have more conservative management (30.6% vs. 17.1%, p = 0.07). The VTE group showed a trend towards more frequent open reduction and internal fixation (73.2% vs. 58.2%, p = 0.06). ICU admission was required more in the VTE group (38.1% vs. 17.7%, p = 0.002). There was no significant difference in mechanical ventilation rates (19.0% vs. 10.6%, p = 0.10).Table 2Management of patientsCharacteristicsAll patientsN = 437VTEN = 42No VTEN = 395P-valueTrauma managementIntravenous tranexamic acid on admission, n (%)46 (10.5%)7 (16.7%)39 (9.9%)0.06RBC transfusion, n (%)63 (14.4%)11 (26.2%)52 (13.2%)0.02RBC Transfusion in the first 24 h (Unit), median (Q1, Q3)2.0 (2.0, 4.0)2.0 (2.0, 4.0)2.0 (2.0, 5.0)0.49Platelet transfusion19 (4.3%)4 (9.5%)15 (3.8%)0.08Platelet transfusion (Unit), median (Q1, Q3)6.0 (6.0, 6.0)5.0 (4.0, 6.0)6.0 (6.0, 6.0)0.16Pelvic angioembolization, n (%)16 (3.7)4 (9.5)12 (3.0)0.03Admission to the intensive care unit, n (%)86 (19.7)16 (38.1)70 (17.7)0.002Mechanical ventilation, n (%)50 (11.4)8 (19.0)42 (10.6)0.10Pelvic fracture management n (%)Non-Surgical128 (29.4)7 (17.1)121 (30.6)0.07Open reduction internal fixation260 (59.6)30 (73.2)230 (58.2)0.06Closed reduction internal fixation21 (4.8)2 (4.9)19 (4.8)0.99Posterior column screw2 (0.5)0 (0.0)2 (0.5)0.65Ilio-sacral screw15 (3.4)3 (7.3)12 (3.0)0.15Trans-iliac trans-sacral screw1 (0.2)0 (0.0)1 (0.3)0.75Anterior column screw2 (0.5)0 (0.0)2 (0.5)0.65Spinopelvic fixation3 (0.7)0 (0.0)3 (0.8)0.58Days from Injury to surgical treatment5.0 (2.5, 7.0)6.0 (3.0, 12.5)5.0 (2.0, 7.0)0.12Thromboprophylaxis approachNo pharmacologic thromboprophylaxis26 (5.9)1 (2.4)25 (6.3)0.50Any pharmacologic thromboprophylaxis411 (94.1)41 (97.6)370 (93.7)0.50Enoxaparin only278 (63.6)2 (47.6)258 (65.3) < 0.0001Unfractionated heparin only82 (18.8)6 (14.3)76 (19.2)0.54Enoxaparin and unfractionated heparin43 (9.8)14 (33.3)29 (7.3) < 0.0001Other anticoagulants8 (1.8)1 (2.4)7 (1.8)0.56Start of pharmacologic thromboprophylaxisDay 1Day 2 or 3After day 3241 (58.6)117 (28.5)53 (12.9)21 (51.2)15 (36.6)5 (12.2)220 (59.5)102 (27.6)48 (13.0)0.47Vena cava Filter insertion, n (%)13 (3.0)4 (9.5)9 (2.3)0.03CTPA done, n (%)58 (13.3)24 (57.1)34 (8.6) < 0.0001Limb Doppler ultrasound done, n (%)229 (52.4)36 (85.7)193 (48.9) < 0.0001CTPA, CT pulmonary angiography; Q1, first quartile; Q3, third quartile; RBC, red blood cell Table 2Management of patientsCharacteristicsAll patientsN = 437VTEN = 42No VTEN = 395P-valueTrauma managementIntravenous tranexamic acid on admission, n (%)46 (10.5%)7 (16.7%)39 (9.9%)0.06RBC transfusion, n (%)63 (14.4%)11 (26.2%)52 (13.2%)0.02RBC Transfusion in the first 24 h (Unit), median (Q1, Q3)2.0 (2.0, 4.0)2.0 (2.0, 4.0)2.0 (2.0, 5.0)0.49Platelet transfusion19 (4.3%)4 (9.5%)15 (3.8%)0.08Platelet transfusion (Unit), median (Q1, Q3)6.0 (6.0, 6.0)5.0 (4.0, 6.0)6.0 (6.0, 6.0)0.16Pelvic angioembolization, n (%)16 (3.7)4 (9.5)12 (3.0)0.03Admission to the intensive care unit, n (%)86 (19.7)16 (38.1)70 (17.7)0.002Mechanical ventilation, n (%)50 (11.4)8 (19.0)42 (10.6)0.10Pelvic fracture management n (%)Non-Surgical128 (29.4)7 (17.1)121 (30.6)0.07Open reduction internal fixation260 (59.6)30 (73.2)230 (58.2)0.06Closed reduction internal fixation21 (4.8)2 (4.9)19 (4.8)0.99Posterior column screw2 (0.5)0 (0.0)2 (0.5)0.65Ilio-sacral screw15 (3.4)3 (7.3)12 (3.0)0.15Trans-iliac trans-sacral screw1 (0.2)0 (0.0)1 (0.3)0.75Anterior column screw2 (0.5)0 (0.0)2 (0.5)0.65Spinopelvic fixation3 (0.7)0 (0.0)3 (0.8)0.58Days from Injury to surgical treatment5.0 (2.5, 7.0)6.0 (3.0, 12.5)5.0 (2.0, 7.0)0.12Thromboprophylaxis approachNo pharmacologic thromboprophylaxis26 (5.9)1 (2.4)25 (6.3)0.50Any pharmacologic thromboprophylaxis411 (94.1)41 (97.6)370 (93.7)0.50Enoxaparin only278 (63.6)2 (47.6)258 (65.3) < 0.0001Unfractionated heparin only82 (18.8)6 (14.3)76 (19.2)0.54Enoxaparin and unfractionated heparin43 (9.8)14 (33.3)29 (7.3) < 0.0001Other anticoagulants8 (1.8)1 (2.4)7 (1.8)0.56Start of pharmacologic thromboprophylaxisDay 1Day 2 or 3After day 3241 (58.6)117 (28.5)53 (12.9)21 (51.2)15 (36.6)5 (12.2)220 (59.5)102 (27.6)48 (13.0)0.47Vena cava Filter insertion, n (%)13 (3.0)4 (9.5)9 (2.3)0.03CTPA done, n (%)58 (13.3)24 (57.1)34 (8.6) < 0.0001Limb Doppler ultrasound done, n (%)229 (52.4)36 (85.7)193 (48.9) < 0.0001CTPA, CT pulmonary angiography; Q1, first quartile; Q3, third quartile; RBC, red blood cell Management of patients Start of pharmacologic thromboprophylaxis Day 1 Day 2 or 3 After day 3 241 (58.6) 117 (28.5) 53 (12.9) 21 (51.2) 15 (36.6) 5 (12.2) 220 (59.5) 102 (27.6) 48 (13.0) CTPA, CT pulmonary angiography; Q1, first quartile; Q3, third quartile; RBC, red blood cell Pharmacologic thromboprophylaxis was used in 94.1% of all patients, with the majority receiving enoxaparin only (63.6%). The median daily enoxaparin dose was 40 mg, with no difference between patients who developed VTE and those who did not. Only 4.1% of patients received higher doses for thromboprophylaxis. Pharmacologic thromboprophylaxis was initiated on Day 1 for 58.6% of all patients (51.2% of patients with VTE, and 59.5% of patients without VTE), on Day 2 or 3 for 28.5% of all patients (36.6% of patients with VTE, and 27.6% of patients without VTE), and after Day 3 for 12.9% of all patients (12.2% of patients with VTE, and 13.0% of patients without VTE), which indicates no significant difference in the timing of prophylaxis initiation between the groups. A vena cava filter was inserted in 25 patients (5.7%); after VTE diagnosis in 14. Among the other 11 patients, only 1 patient was diagnosed with VTE after filter insertion. Predictors of venous thromboembolism events On multivariable logistic regression analysis (Table 3), the significant predictors of VTE were age (odds ratio per year increment 1.022, 95% confidence interval 1.000–1.045), admission to the ICU (odds ratio 2.746, 95% confidence interval 1.041–7.240) and hospital length of stay (odds ratio per day increment 1.008, 95% confidence interval 1.000–1.016). Presence of comorbidities, non-surgical treatment, pelvic angioembolization, chest injury, cervical spine injury, and abdominal/pelvic injury, did not show significant associations with VTE. No early pharmacologic thromboprophylaxis had an odds ratio for VTE of 2.156 (95% confidence interval 0.776–5.990), but this was not statistically significant (p = 0.14).Table 3Multivariable logistic regression analysis for the predictors of venous thromboembolismOdds ratio95% confidence intervalP-valueAge per year increment1.0221.000–1.0450.045Female versus male sex1.6150.720–3.6230.25Body mass index per unit increment0.9920.934–1.0530.79No Comorbidity vs. presence of at least one co-morbidity0.8040.327–1.9780.64Glasgow Coma Scale at presentation1.1240.931–1.3580.22Traumatic brain injury0.5380.126–2.3000.53Cervical spine injuryNo estimateNo estimate1.0Chest injury1.3340.548–3.2500.41Abdomen/pelvis injury2.0640.851–5.0050.11Abbreviated Injury Scale (pelvis)1.0220.598–1.7470.90Tranexamic acid use1.0480.283–3.8830.94Red blood cell transfusion on day 11.6640.640–4.3260.24Pelvic angioembolization0.7850.161–3.8420.77ICU admission2.7461.041–7.2400.03Nonsurgical vs. any surgical treatment for pelvic/acetabular fracture1.8780.669–5.2730.23Early pharmacologic prophylaxis (within 2 days) vs. no early prophylaxis0.4630.153–1.4040.17Length of hospital admission per day increment1.0081.000–1.0160.04For the 63 patients without the Abbreviated Injury Scale, the missing values were imputed using the cohort median, which was 4The model was statistically significant compared with the null model (chi-square value = 37.3.4, p = 0.03) and had a correct classification rate of 91.3%. The Hosmer and Lemeshow goodness-of-fit test was significant (p = 0.88). Nagelkerke R2 = 0.185 Table 3Multivariable logistic regression analysis for the predictors of venous thromboembolismOdds ratio95% confidence intervalP-valueAge per year increment1.0221.000–1.0450.045Female versus male sex1.6150.720–3.6230.25Body mass index per unit increment0.9920.934–1.0530.79No Comorbidity vs. presence of at least one co-morbidity0.8040.327–1.9780.64Glasgow Coma Scale at presentation1.1240.931–1.3580.22Traumatic brain injury0.5380.126–2.3000.53Cervical spine injuryNo estimateNo estimate1.0Chest injury1.3340.548–3.2500.41Abdomen/pelvis injury2.0640.851–5.0050.11Abbreviated Injury Scale (pelvis)1.0220.598–1.7470.90Tranexamic acid use1.0480.283–3.8830.94Red blood cell transfusion on day 11.6640.640–4.3260.24Pelvic angioembolization0.7850.161–3.8420.77ICU admission2.7461.041–7.2400.03Nonsurgical vs. any surgical treatment for pelvic/acetabular fracture1.8780.669–5.2730.23Early pharmacologic prophylaxis (within 2 days) vs. no early prophylaxis0.4630.153–1.4040.17Length of hospital admission per day increment1.0081.000–1.0160.04For the 63 patients without the Abbreviated Injury Scale, the missing values were imputed using the cohort median, which was 4The model was statistically significant compared with the null model (chi-square value = 37.3.4, p = 0.03) and had a correct classification rate of 91.3%. The Hosmer and Lemeshow goodness-of-fit test was significant (p = 0.88). Nagelkerke R2 = 0.185 Multivariable logistic regression analysis for the predictors of venous thromboembolism For the 63 patients without the Abbreviated Injury Scale, the missing values were imputed using the cohort median, which was 4 The model was statistically significant compared with the null model (chi-square value = 37.3.4, p = 0.03) and had a correct classification rate of 91.3%. The Hosmer and Lemeshow goodness-of-fit test was significant (p = 0.88). Nagelkerke R2 = 0.185 Figure 1 describes VTE rates in selected subgroups. Patients with late surgical interventions (after 7 days) had higher rates of VTE (14/64 [21.9%] vs. 18/209 [8.6%], p = 0.004). Patients who received enoxaparin for thromboprophylaxis had a very low VTE rate (0.7%) compared with 7.3% for patients who received only UFH (p = 0.002).Fig. 1Rates of venous thromboembolism (deep-vein thrombosis and/or pulmonary embolism) in subgroups of patients Rates of venous thromboembolism (deep-vein thrombosis and/or pulmonary embolism) in subgroups of patients Outcomes of patients The outcomes of patients are shown in Table 4. The hospital mortality was similar between the two groups (4.8% for VTE patients vs. 4.1% for non-VTE patients, p = 0.83). Patients with VTE had a longer median duration of ventilation (11.0 days vs. 7.0 days, p = 0.20) and ICU stay (14.0 days vs. 10.0 days, p = 0.23), although these differences were not statistically significant. However, the median hospital length of stay was significantly longer for VTE patients compared to non-VTE patients (21.5 days vs. 12.0 days, p < 0.0001).Table 4Outcomes of the study patientsAll patientsN = 437VTEN = 42No VTEN = 395P-valueHospital death, n (%)18 (4.1%)2 (4.8%)16 (4.1%)0.83Duration of mechanical ventilation (days), median (Q1, Q3)7.50 (3.3, 13.0)11.0 (4.8, 24.3)7.0 (3.0, 12.3)0.20ICU stay* (days), median (Q1, Q3)10.5 (4.0, 17.3)14.0 (4.3, 26.0)10.0 (4.0, 16.3)0.23Hospital length of stay (days), median (Q1, Q3)13.0 (7.0, 22.0)21.5 (12.0, 48.3)12.0 (7.0, 21.0) < 0.0001*For patients who were admitted to the ICU. ICU, intensive care unit; Q1, first quartile; Q3, third quartile; VTE venous thromboembolism Table 4Outcomes of the study patientsAll patientsN = 437VTEN = 42No VTEN = 395P-valueHospital death, n (%)18 (4.1%)2 (4.8%)16 (4.1%)0.83Duration of mechanical ventilation (days), median (Q1, Q3)7.50 (3.3, 13.0)11.0 (4.8, 24.3)7.0 (3.0, 12.3)0.20ICU stay* (days), median (Q1, Q3)10.5 (4.0, 17.3)14.0 (4.3, 26.0)10.0 (4.0, 16.3)0.23Hospital length of stay (days), median (Q1, Q3)13.0 (7.0, 22.0)21.5 (12.0, 48.3)12.0 (7.0, 21.0) < 0.0001*For patients who were admitted to the ICU. ICU, intensive care unit; Q1, first quartile; Q3, third quartile; VTE venous thromboembolism Outcomes of the study patients *For patients who were admitted to the ICU. ICU, intensive care unit; Q1, first quartile; Q3, third quartile; VTE venous thromboembolism Discussion The current study revealed that the incidence of VTE in hospitalized patients with pelvic fracture was approximately 10%, with DVT and PE occurring at similar rates. Pharmacologic thromboprophylaxis was widely used, mostly early after hospital admission. Older age and ICU admission were identified as risk factors for VTE, possibly reflecting the severity of pelvic or other injuries. Patients who had delayed surgery (after 7 days) had a higher VTE rate. Although VTE was not linked to increased mortality, it was associated with a longer hospital stay. In our study of 437 patients, the rate of VTE was approximately 10%, most commonly occurring early, a median of 3 days after trauma. VTE rates in patients with trauma in general and pelvic fractures vary considerably across studies [14]. A systematic review of studies published between 1990 and 2023 reported rates ranging from 0.39% to 32%, and identified pelvic injury including fracture as a risk factor for DVT [14]. Similar variability has been observed in patients with traumatic pelvic fractures. A study of 110 patients with pelvic or acetabular fractures who received mechanical and chemical thromboprophylaxis and had Doppler ultrasound for screening found a 29.1% VTE rate, with acetabular fractures associated with higher DVT incidence (37.1%) than pelvic fractures (18.8%) (p = 0.04) [7]. A study at a single center in the US, symptomatic VTE was diagnosed in 3.8% of patients with pelvic and/or acetabular fractures [15]. A Taiwanese study of 402 patients with traumatic pelvic fractures with doppler ultrasound done on the third postoperative day reported a 3.5% VTE rate, with only 0.25% developing symptomatic PE [16]. Mechanical prophylaxis and early surgery likely contributed to these lower rates [16]. A retrospective U.S. study using a large trauma database showed an overall VTE rate of 0.82% (0.39% for DVT and 0.43% for PE) in patients with pelvic and lower extremity fractures and on thromboprophylaxis, though pelvic open reduction internal fixation alone had a higher rate of 1.70% [17]. The reason of variation in reported incidence of VTE is likely related to differences in study design, including the use of population-based administrative databases in some studies, and in patient characteristics, including trauma severity. Our finding of nearly equal distribution of DVT and PE contrasts with most other studies, in which PE rates were typically lower (e.g., 0.25% for PE vs. 3.48% for DVT in one study) [17]. Most VTE events after trauma usually occur in the first 1–3 weeks, with similar patterns in patients with pelvic and/or acetabular fractures [15, 18]. One study found that 63.4% of 82 patients with pelvic and acetabular fractures who had preoperative DVT had them within 7 days after injury, and 81.7% patients within 10 days [4], which is not different from our findings. In our study pharmacologic thromboprophylaxis was used in 94.1% of all patients, predominantly early, which aligns with the current evidence and clinical practice guidelines [19, 20]. Early pharmacologic thromboprophylaxis is associated with reduced VTE rates compared to delayed administration [6]. For example, a study of 2752 patients with isolated pelvic fractures found higher VTE rates in patients who received late thromboprophylaxis (> 48 h) versus earlier (4.3% vs. 2.2%, p = 0.004), and greater protection with LMWH than UFH [6]. However, not all studies agree as one study found no relationship between VTE prophylaxis timing and VTE incidence [9]. We did not find a significant association between pharmacologic thromboprophylaxis timing and VTE incidence, which could be related to sample size. A systematic review of studies published up to October 2020 found 3 studies (3107 patients with pelvic and/ or acetabular fractures) demonstrated higher rates of symptomatic DVT, higher VTE risk (odds ratio 1.9, 95% confidence interval 1.2–3.2) and death (odds ratio 4.0, 95% confidence interval 1.5–11) with delayed prophylaxis [21]. Studies also showed no increase in bleeding complications in early pharmacologic thromboprophylaxis, even in patients with solid organ injuries or angioembolization [9, 22]. Our findings favor LMWH over UFH for thromboprophylaxis which is in line with the current guidelines and other studies [19]. Standard dosing may be insufficient in trauma. Emerging strategies, such as anti-Xa–guided dosage, weight-based dosing, and thromboelastography, may enhance efficacy [20]. In select patients with pelvic fractures, direct oral anticoagulants may offer an alternative with evidence suggesting lower VTE compared to LMWH [23]. In our study, age, ICU admission and length of hospital stay were independent risk factors for VTE. ICU admission often reflects the severity of injuries and is often associated with prolonged immobility. The Adjusted Injury Scale for pelvis tended to be higher in the patients with VTE, which also suggests that pelvic trauma severity, and potentially the associated immobility, may have contributed to VTE. The association between length of hospital stay and VTE may be bidirectional, with extended stays increasing VTE risk and VTE and its treatment increasing hospital stay. We also found that patients who had only LMWH compared with those who received only UFH for prophylaxis and those who had earlier surgical management versus delayed surgical management had lower rates of VTE on univariate analysis. Other studies found older age [4, 7], angioembolization [9], associated injuries [7], delayed pharmacologic thromboprophylaxis [5, 6], and delayed surgery [4, 7], as risk factors for VTE in patients with pelvic fractures. ICU admission and delayed surgery can prolong immobility, which together with trauma-associated hypercoagulable state and inflammation increase the risk of venous thrombosis [24]. In our study, VTE was not associated with increased hospital mortality, which was close to 4%. It should be noted that trauma patients with pelvic fractures have increased mortality compared to other trauma patients [3, 25]. VTE has been associated with increased mortality in patients with trauma (hazard ratio 1.18, p < 0.001) [8]. While VTE is a serious complication, its impact on mortality can be mitigated with appropriate care. In addition, VTE was associated with longer hospital stay, which suggests that VTE incurs the need for additional and longer treatment and recovery time. The limitations of this study include its retrospective design and being conducted at a single center, which may limit generalizability of our findings. Although we did note other traumatic injuries, we did not calculate injury severity scores. In our hospital, especially in the ICU, intermittent pneumatic compression devices were routinely applied to patients not receiving pharmacologic prophylaxis. Such practice was not consistently documented in the medical records, which limited our ability to study this variable. Many patients received different heparins during hospital stay for thromboprophylaxis, which did not allow us to enter this variable in the logistic regression analysis. We also did not assess adherence to anticoagulants, bleeding events, and factors that may have contributed to delays in surgical interventions. These may have affected VTE incidence and other outcomes. The association between ICU admission and VTE may be related to trauma severity or complications acquired during ICU or hospital stay. These and other potential confounders were not accounted for in this study. Additionally, the sample size may also have reduced the statistical power to detect associations. Conclusions In conclusion, VTE occurred in about 10% of hospitalized patients with traumatic pelvic and/or acetabular fracture and so remains a notable risk, especially among those who required admission to the ICU, which is likely a marker of more severe trauma. While not associated with higher mortality, VTE was associated with longer hospital stays. These findings underscore the need for more tailored risk assessments and improved VTE prevention strategies in this patient population to reduce morbidity, resource use and prolonged hospitalization.