Abstract
Background
Patients with traumatic brain injury (TBI) are at increased risk for venous thromboembolism (VTE) due to immobilization and a hypercoagulable state. The administration and optimal timing for chemoprophylaxis remain unclear because of concerns about safety and efficacy.
Methods
We conducted a retrospective review of adult patients admitted with TBI to a level I trauma center between January 2019 and May 2020. Patient demographics, Glasgow Coma Scale, mobility score, mechanism of injury, comorbidities, timing of chemoprophylaxis and physical therapy, VTE events, and intracranial hemorrhage complications were analyzed. Multivariate logistic regression was performed to identify predictors of VTE.
Results
Among 405 patients, 161 received chemoprophylaxis. Eight patients experienced VTE. Receipt of prophylaxis in ≤48 hours was not associated with reduced VTE risk (P = 0.81). However, a 5-point increase in admission mobility score was independently associated with an 84% lower risk of VTE (P = 0.02). Intracranial hemorrhage expansion occurred in 25 patients, with no significant difference between those who did and did not receive prophylaxis.
Discussion
Higher admission mobility scores predict lower VTE risk in TBI. Early prophylaxis timing was not statistically associated with VTE. These findings underscore the need for prospective studies to determine safe and effective chemoprophylaxis strategies.
Keywords: Chemoprophylaxis, deep vein thrombosis, intracranial hemorrhage, traumatic brain injury, venous thromboembolism
Over 214,000 traumatic brain injury (TBI)-related hospitalizations and over 69,000 TBI-related deaths were reported in 2020 and 2021, respectively.1 Patients with TBI have a significant risk of developing deep vein thrombosis (DVT) and pulmonary embolism (PE).2 This is secondary to the hypercoagulable state that results from the primary brain injury and the extensive periods of immobilization. Initiation of chemoprophylaxis can limit this hypercoagulable state and decrease the risk of venous thromboembolism (VTE). However, there is insufficient evidence to support recommendations regarding the agent, dose, or timing of pharmacologic prophylaxis, and these recommendations come with the warning of an increased risk for intracranial hemorrhage expansion. We hypothesized that patients hospitalized for TBI who had initiation of DVT prophylaxis 48 hours or earlier after admission would have a lower incidence of thromboembolic events.
METHODS
The Baylor Scott & White Research Institute institutional review board approved this study (024–432) and waived the requirement for informed consent. We retrospectively reviewed all patients treated for TBI at Baylor Scott & White Medical Center – Temple, a level I trauma center, between January 2019 and May 2020, including all patients aged 18 years or older who were admitted for TBI.
Patient baseline demographic data on arrival included age, sex, Glasgow Coma Scale (GCS), and medical comorbidities. Patients’ electronic medical records were reviewed for timing and administration of chemical VTE prophylaxis with either subcutaneous heparin or low-molecular-weight heparin. The mechanism of injury (fall, motor vehicle accident, assault, or struck by vehicle as pedestrian), mobility score on admission, time to physical therapy/occupational therapy consult, use of sequential compression device or thromboembolism-deterrent stockings, length of hospital stay, surgical procedures performed, discharge disposition, and length of follow-up were recorded. Brain bleeding complications, intracranial hemorrhage expansion, wound complications, VTE events, and other significant events were identified through the review of imaging and radiology reports. Patients did not receive routine screening for VTE, and VTE events were identified when clinically suspected during the patient’s acute hospitalization. Patients were considered to have had expansion of their intracranial hemorrhage if the radiology report indicated that the size of the intracranial hemorrhage had increased.
We performed a bivariate analysis between patients who did and did not have a VTE event. We used Fisher’s exact test for categorical values, given that every variable had at least one cell count that was ≤4. We used the Kolmogorov-Smirnov test to assess normality for continuous variables and a Wilcoxon rank sum test for categorical variables that did not fit a normal distribution. We performed a multivariate logistic regression with variables of interest designed to predict the occurrence of a thromboembolic event. All analyses were performed with SAS 9.4 (SAS Institute, Cary, NC).
RESULTS
A total of 405 patients were admitted to our hospital for TBI during the study period, and 161 patients had DVT prophylaxis during their hospital admission, a prophylaxis rate of 40%. One patient who had DVT prophylaxis was diagnosed with a PE upon admission and excluded from the final analysis. There were 8 and 152 patients who did and did not have VTE events, respectively. Patients who had a VTE event had a lower median GCS score (3, interquartile range [IQR] 3–10) and a lower median mobility score (6, IQR 6–14) compared to patients who did not have a VTE event (15, IQR 8–15; P = 0.01 for GCS and 13, IQR 9–18; P = 0.003, for mobility score). Complete data for the cohorts are presented in Table 1.
Table 1.
Physical and clinical characteristics of patients who did and did not have thrombotic events
| Variable | Thromboembolic event (N = 8) | No thromboembolic event (N = 152) | P value |
|---|---|---|---|
| Female gender | 4 (50%) | 54 (36%) | 0.46 |
| Age (years): Median (IQR) | 52.7 (37.9–62.4) | 64.4 (39.8–80.5) | 0.17 |
| Body mass index (kg/m2): Median (IQR) | 26.6 (25.9–36.7) | 26.0 (22.7–30.4) | 0.20 |
| Comorbidities | |||
| Prescribed anticoagulation medication at time of admission | 1 (12%) | 28 (18%) | 1.00 |
| History of cerebrovascular accident | 0 | 13 (9%) | 1.00 |
| Congestive heart failure | 0 | 8 (5%) | 1.00 |
| Cirrhosis | 0 | 2 (1%) | 1.00 |
| Chronic obstructive pulmonary disease | 1 (12%) | 6 (4%) | 0.31 |
| Dementia | 0 | 16 (11%) | 1.00 |
| Diabetes mellitus | 0 | 29 (19%) | 0.35 |
| Do not resuscitate order | 0 | 4 (3%) | 1.00 |
| Alcohol use | 0 | 10 (7%) | 0 |
| Functional dependency | 1 (12%) | 43 (28%) | 0.45 |
| Hypertension | 2 (25%) | 69 (45%) | 0.30 |
| Mental | 2 (25%) | 25 (16%) | 0.62 |
| History of myocardial infarction | 0 | 3 (2%) | 1.00 |
| Renal | 0 | 3 (2%) | 1.00 |
| Substance-related | 0 | 3 (2%) | 1.00 |
| Smoking | 1 (12%) | 26 (17%) | 1.00 |
| Mechanism of injury | 0.16 | ||
| Fall | 2 (25%) | 87 (57%) | |
| Motor vehicle accident | 5 (62%) | 44 (29%) | |
| Assault | 0 | 11 (7%) | |
| Struck by vehicle as a pedestrian | 1 (12%) | 10 (7%) | |
| Glasgow Coma Scale: Median (IQR) | 3 (3–10) | 15 (8–15) | 0.01 |
| Mobility score on admission | 6 (6–14) | 13 (9–18) | 0.003 |
| Received physical therapy or occupational therapy consultation | 8 (100%) | 146 (96%) | 1.00 |
| Timing of DVT prophylaxis | 0.46 | ||
| ≤48 after admission | 4 (50%) | 54 (36%) | |
| >48 hours after admission | 4 (50%) | 98 (64%) | |
| Type of DVT prophylaxis | 0.13 | ||
| Unfractionated heparin | 6 (75%) | 141 (93%) | |
| Low-molecular-weight heparin | 2 (25%) | 11 (7%) | |
| Surgical procedure during hospital admission | 3 (38%) | 61 (40%) | 1.00 |
DVT indicates deep vein thrombosis.
A multivariate logistic regression designed to predict occurrence of a VTE event found that initiation of DVT prophylaxis 48 hours or less after admission was not associated with prevention of a VTE event (adjusted odds ratio [aOR] 2.05; 95% confidence interval [CI] 0.32–12.94; P = 0.45). We also found that a 5-point increase in mobility scores was independently associated with a decreased occurrence of a VTE event (aOR 0.16; 95% CI 0.03–0.75; P = 0.02). GCS and mobility scores were found to be colinear, so GCS was excluded from the multivariate model in favor of the mobility score to improve fit. The injury mechanisms “assault” and “struck by” were also combined into a single category. Complete data for the multivariate logistic regression are presented in Table 2. The c-statistic of the multivariate logistic regression was 0.86.
Table 2.
Multivariate logistic regression to predict occurrence of deep vein thrombosis
| Variable | Unit | Odds ratio | 95% confidence interval | P value | |
|---|---|---|---|---|---|
| Age | 10 years | 1.06 | 0.69 | 1.63 | 0.80 |
| Body mass index | 3 kg/m2 | 01.23 | 0.80 | 1.89 | 0.35 |
| Injury mechanism | |||||
| Fall (reference) | N/A | 1.00 | |||
| Motor vehicle accident vs fall | 1.87 | 0.22 | 16.23 | 0.57 | |
| Assault or struck by vehicle vs fall | 0.95 | 0.06 | 15.55 | 0.97 | |
| Mobility score at admission | 5-point increase | 0.16 | 0.03 | 0.75 | 0.02 |
| VTE prophylaxis timing | ≤48 h vs >48 h | 2.05 | 0.32 | 12.94 | 0.45 |
| VTE prophylaxis medication | LMWH vs unfractionated heparin | 1.60 | 0.14 | 18.57 | 0.70 |
LMWH indicates low-molecular-weight heparin; VTE, venous thromboembolism.
Overall, 25 patients were found to have intracranial hemorrhage expansion, with 8 of those patients also requiring surgery in the form of a craniotomy or craniectomy for management. Eleven patients received VTE prophylaxis, and 14 received no VTE prophylaxis prior to the hemorrhage expansion, with no significant difference found between groups in regard to risk of intracranial hemorrhage expansion or neurosurgical intervention.
DISCUSSION
We found that the timing of DVT prophylaxis was not associated with the occurrence of a VTE event, but, after controlling for age, body mass index, mechanism of injury, timing of VTE prophylaxis, and type of VTE prophylaxis, a 5-point increase in the mobility score at admission was associated with an 84% reduction in thrombotic events during hospitalization.
The association between mobility score and VTE events is consistent with the known impact of immobility and critical illness on VTE.3,4 Older age, male gender, Injury Severity Score > 15, and a concomitant lower extremity injury have also been found to be strong predictors of developing DVT in patients hospitalized for TBI.5 Given the association between GCS and mobility score, we included only mobility score in our multivariate logistic regression.
The optimal strategy for preventing VTE in patients with TBI remains a subject of ongoing debate due to the competing risks of thrombosis and hemorrhagic progression. The Brain Trauma Foundation currently provides a level 3 recommendation, based on limited or conflicting evidence, for the use of low-molecular-weight heparin or low-dose unfractionated heparin in combination with mechanical prophylaxis, specifically compression stockings, for the prevention of DVT in TBI.6 This recommendation reflects the lack of consensus among the community about the safety and efficacy of DVT prophylaxis in TBI patients.
The rate of intracranial hemorrhage expansion was found to be just above 6%, with no statistical significance between patients who had and had not received DVT chemoprophylaxis. Several factors have been found to increase the risk of intracranial hemorrhage expansion, including the presence of a subdural or epidural hematoma >8 mm, a contusion or intraventricular hemorrhage >2 cm, multiple contusions per lobe, a subarachnoid hemorrhage with abnormal computed tomography angiogram, and evidence of progression at 24 hours.7 Current evidence suggests that early initiation of VTE prophylaxis, which is typically within 24 to 72 hours after confirmation of stable intracranial hemorrhage, reduces the risk of VTE without significantly increasing the risk of intracranial hemorrhage expansion.8–14 In a large multicenter analysis, Ratnasekera et al found that patients who never received VTE prophylaxis had decreased odds of intracranial hemorrhage expansion, though without significant statistical difference.15 Thus, evidence suggests that stability of intracranial hemorrhage in TBI patients must be ensured prior to initiation of prophylactic agents to avoid increased risks of hemorrhage expansion and increased patient complications.
Beyond the safety of the use of DVT chemoprophylaxis in TBI patients, timing of prophylaxis also appears to affect the efficacy of treatment. We found that the timing of DVT prophylaxis was not associated with the occurrence of a VTE event. Many studies suggest that delayed administration of chemoprophylaxis is associated with significantly higher rates of VTE, especially if withheld for >7 days.8 A systematic review that included 17 studies, of which only one was a randomized controlled trial, showed that administering VTE chemoprophylaxis within 24 hours of injury in stable TBI does not lead to progressive intracranial hemorrhage, while delaying chemoprophylaxis for 72 hours and 96 hours after injury increases the risk of VTE by twofold and threefold, respectively.16 Limited randomized studies exist in the literature; thus, more trials are needed to establish strong, evidence-based recommendations.
This study had several limitations. As a retrospective cohort analysis, it is subject to potential selection and documentation bias, and the heterogeneity of the study population may have introduced unmeasured confounders. The observed rate of VTE events in patients hospitalized for TBI was 5%. The occurrence of VTE is notably lower than the range reported in previous studies, which documented rates up to 30% in select populations, especially those with a hospital stay >7 days.5,17 However, our rate aligns more closely with large multicenter cohorts and systematic reviews reporting rates between 2.8% and 12%.8,15,18 This variability is likely attributable to differences in patient selection, injury severity, diagnostic intensity, and prophylaxis protocols. Also, detection was only identified during the acute hospitalization, underestimating the true incidence of VTE by not including occurrences in the rehabilitation phase. An additional limitation was our definition of expansion of intracranial hemorrhage, which only considered whether patients had an increase in their intracranial hemorrhage without additional quantitative measurements. Our low event rate (n = 8) may have increased the risk of a type II error and affected the statistical power to detect associations between prophylaxis timing and VTE occurrence.
Another statistical limitation was the potential colinear relationship between mobility and GCS scores. Since mobility score had a stronger statistical association with VTE events in the bivariate analysis compared to GCS score, we only included the mobility score in our multivariate logistic regression. In our multivariate logistic regression, we combined patients who were assaulted and struck by a vehicle into one cohort in the mechanism of injury variable due to low cell count. We believe that assault and being struck by a vehicle were a similar mechanism of injury that was different from falls and motor vehicle accidents. Nevertheless, combining patients who experienced assault and patients who were struck by a vehicle is an additional limitation of our study. The findings of our study should be interpreted with caution, and prospective studies with standardized protocols are needed to better define the optimal timing of chemoprophylaxis in TBI.
In summary, higher admission mobility scores were associated with a lower risk of VTE in patients with TBI. While the timing of chemoprophylaxis was not significantly associated with VTE in our cohort, these results reinforce the need for further studies to clarify optimal prophylaxis strategies and support evidence advocating early initiation to minimize thromboembolic risk.
ACKNOWLEDGMENTS
The authors acknowledge the statistical support of Chandni Raiyani, MPH.
DISCLOSURE STATEMENT
The authors report no funding or conflict of interest. This study was departmentally funded.
References
- 1.Centers for Disease Control and Prevention . TBI data; 2023. https://www.cdc.gov/traumaticbraininjury/data/. Accessed May 7, 2024.
- 2.Kaufman HH, Satterwhite T, McConnell BJ, et al. Deep vein thrombosis and pulmonary embolism in head-injured patients. Angiology. 1983;34(10):627–638. doi: 10.1177/000331978303401001. [DOI] [PubMed] [Google Scholar]
- 3.Stawicki SP, Grossman MD, Cipolla J, et al. Deep venous thrombosis and pulmonary embolism in trauma patients: an overstatement of the problem? Am Surg. 2005;71(5):387–391. doi: 10.1177/000313480507100504. [DOI] [PubMed] [Google Scholar]
- 4.Cole KL, Nguyen S, Gelhard S, et al. Factors associated with venous thromboembolism development in patients with traumatic brain injury. Neurocrit Care. 2024;40(2):568–576. doi: 10.1007/s12028-023-01780-8. [DOI] [PubMed] [Google Scholar]
- 5.Ekeh AP, Dominguez KM, Markert RJ, McCarthy MC.. Incidence and risk factors for deep venous thrombosis after moderate and severe brain injury. J Trauma. 2010;68(4):912–915. doi: 10.1097/TA.0b013e3181b21cad. [DOI] [PubMed] [Google Scholar]
- 6.Carney N, Totten AM, O’Reilly C, et al. Guidelines for the management of severe traumatic brain injury, fourth edition. Neurosurgery. 2017;80(1):6–15. doi: 10.1227/NEU.0000000000001432. [DOI] [PubMed] [Google Scholar]
- 7.American College of Surgeons Committee on Trauma . ACS TQIP best practices in the management of traumatic brain injury; 2015. https://www.facs.org/media/mkej5u3b/tbi_guidelines.pdf. Accessed July 2024.
- 8.Abdel-Aziz H, Dunham CM, Malik RJ, Hileman BM.. Timing for deep vein thrombosis chemoprophylaxis in traumatic brain injury: an evidence-based review. Crit Care. 2015;19(1):96. doi: 10.1186/s13054-015-0814-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Pastorek RA, Cripps MW, Bernstein IH, et al. The Parkland Protocol’s modified Berne-Norwood criteria predict two tiers of risk for traumatic brain injury progression. J Neurotrauma. 2014;31(20):1737–1743. doi: 10.1089/neu.2014.3363. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Scudday T, Brasel K, Webb T, et al. Safety and efficacy of prophylactic anticoagulation in patients with traumatic brain injury. J Am Coll Surg. 2011;213(1):148–153. doi: 10.1016/j.jamcollsurg.2011.03.015. [DOI] [PubMed] [Google Scholar]
- 11.Phelan HA. Pharmacologic venous thromboembolism prophylaxis after traumatic brain injury: a critical literature review. J Neurotrauma. 2012;29(10):1821–1828. doi: 10.1089/neu.2011.2277. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Al-Dorzi HM, Al-Yami G, Al-Daker F, Alqirnas MQ, Alhamadh MS, Khan R.. The association of timing of pharmacological prophylaxis and venous thromboembolism in patients with moderate-to-severe traumatic brain injury: a retrospective cohort study. Ann Thorac Med. 2022;17(2):102–109. doi: 10.4103/atm.atm_174_21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Byrne JP, Witiw CD, Schuster JM, et al. Association of venous thromboembolism prophylaxis after neurosurgical intervention for traumatic brain injury with thromboembolic complications, repeated neurosurgery, and mortality. JAMA Surg. 2022;157(3):e215794. doi: 10.1001/jamasurg.2021.5794. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Mohseni S, Talving P, Lam L, Chan LS, Ives C, Demetriades D.. Venous thromboembolic events in isolated severe traumatic brain injury. J Emerg Trauma Shock. 2012;5(1):11–15. doi: 10.4103/0974-2700.93102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Ratnasekera A, Kim D, Seng S, et al. Early VTE prophylaxis in severe traumatic brain injury: a propensity score weighted EAST multicenter study. J Trauma Acute Care Surg. 2023;95(1):94–104. doi: 10.1097/TA.0000000000003985. [DOI] [PubMed] [Google Scholar]
- 16.Spano PJ 2nd, Shaikh S, Boneva D, Hai S, McKenney M, Elkbuli A.. Anticoagulant chemoprophylaxis in patients with traumatic brain injuries: a systematic review. J Trauma Acute Care Surg. 2020;88(3):454–460. doi: 10.1097/TA.0000000000002580. [DOI] [PubMed] [Google Scholar]
- 17.Silva VT, Iglesio R, Paiva WS, Siqueira MG, Teixeira MJ.. Profilaxia de trombose venosa profunda em doentes com traumatismo cranioencefálico [Deep vein thrombosis prophylaxis in patients with traumatic brain injury]. Acta Med Port. 2015;28(2):250–255. [PubMed] [Google Scholar]
- 18.Wu YT, Chien CY, Matsushima K, et al. Early venous thromboembolism prophylaxis in patients with trauma intracranial hemorrhage: analysis of the prospective multicenter Consortium of Leaders in Traumatic Thromboembolism study. J Trauma Acute Care Surg. 2023;95(5):649–656. doi: 10.1097/TA.0000000000004007. [DOI] [PubMed] [Google Scholar]