Hello, my name is Blake Hauser and I'm an MD-PhD student at Harvard Medical School in my third year working with the Computational Neuroscience Outcomes Center at Brigham and Women's Hospital. And today I'm going to talk to you about venous thromboembolism following pediatric traumatic spinal injuries and how it is associated with injury severity and longer hospital stays. No disclosures. Venous thromboembolism, or VTE, can cause significant morbidity and mortality in hospitalized patients. However, incidence of VTE in pediatric patients is far lower than in adults. VTE may occur disproportionately, though, in patients with limited mobility, such as following traumatic spinal injury, or TSI. We aim to determine what clinical presentation characteristics were associated with VTE among TSI patients, whether injury type and severity were associated with VTE among TSI patients, and how the hospital course of TSI patients with VTE compared to that of patients without VTE. We conducted a retrospective cohort study analyzing data from initial hospitalization using the National Trauma Data Bank, which is publicly available, years 2011 through 2014. We extracted data on all pediatric patients with TSI, which was defined based on ICD-9 code for spinal fracture and spinal cord injury. We then performed multiple amputation for missing data. We conducted multivariable analysis using linear and logistic regression models, and we adjusted these models for demographics, comorbidities, and accompanying injuries. We ultimately had a cohort of 22,752 pediatric TSI patients. 192 of these patients, or 0.8%, experienced VTE during initial hospitalization. Overall, the demographic distributions are remarkably similar across the groups of patients with TSI who did or did not develop VTE. This trend holds for age, sex, race, ethnicity, and insurance status. Looking at individual variables more closely, subtle trends are evident. Patients who developed VTE after presenting with TSI had a slightly higher mean age than patients who did not develop VTE. Additionally, a majority of pediatric patients presenting with TSI were male, with similar proportions of male and female patients across the VTE and non-VTE groups. Private insurance coverage, followed by Medicaid coverage, were the most prevalent types of insurance held by pediatric patients presenting with TSI. A slightly lower proportion of patients who developed VTE had private insurance coverage, while a slightly higher proportion of these patients had no-fault automobile coverage. Note that these trends are in comparison to TSI patients who did not develop VTE. This could reflect differences in the prevalence of motor vehicle accident-related injuries among these two groups of patients. We then looked more closely, specifically at mechanism of injury. All available mechanism of injury codes were considered, so some patients are counted under multiple categories in this analysis, though that is applicable to a minority of patients. Motor vehicle accidents were the most prevalent of mechanism of injury by far overall. A larger proportion of patients with both TSI and VTE had a motor vehicle accident-related mechanism of injury as compared to patients with TSI but without VTE. Injury to other-slash-homicide attempt and injury to self-slash-suicide attempt were also represented more frequently in patients with VTE. Sports-related injuries and falls, on the other hand, were more prevalent in patients with TSI but no VTE. This is suggestive of a possible relationship between increased injury severity and VTE, but further analysis was needed. We next examined the association between select demographic characteristics, concomitant injuries, and comorbidities in VTE in patients with TSI. Many factors related to severity of injury at presentation were significantly associated with VTE in these patients. These factors included lower extremity injury and abdominal injury in comparison to patients where these body regions were uninjured, as well as moderate Glasgow Coma Score, or GCS, and severe GCS in comparison to patients with a GCS of 15. Epidural hematoma was also significantly associated with the development of VTE in patients with TSI. This association did not exist for subdural hematoma, subarachnoid hemorrhage, or cerebral contusion. Interestingly, obesity was also significantly associated with the development of VTE in patients with TSI. This is consistent with findings in other studies, though most of these were conducted in adults rather than in a pediatric population. We then examined associations specifically between characteristics of spinal injuries and VTE in patients with TSI. Despite an association between VTE and both lower extremity injuries and abdominal injuries, the association between lumbar spine injury and the development of VTE was not significant. This association was also insignificant for cervical or thoracic spinal injuries. However, there was a significant association between spinal injury at multiple levels and the development of VTE. Additionally, there was a strong significant association between spinal cord injury and developing VTE. Both of these factors likely represent and reflect increased injury severity. Additionally, they may correspond to decreased patient mobility. This is likely to be particularly true of spinal cord injury. This lack of mobility could put patients at risk for the development of VTE, even in the absence of other risk factors. We next examined the relationship between VTE, hospital course, and discharge disposition among TSI patients. Despite an association with increased injury severity, developing VTE was not associated with an increased likelihood of ICU stay during initial hospital admission. However, developing VTE was associated with adverse discharge. Here we defined adverse discharge as a discharge to a rehabilitation facility or a discharge home with rehabilitative services, as opposed to a discharge home without rehabilitative services. The analysis shown was adjusted for injury severity, so this association was found to be significant even with that adjustment. Interestingly, VTE appears to be protective against death or discharge to a hospice facility in comparison to TSI without VTE. However, this association could be a byproduct of the fact that death and discharge to hospice were both relatively uncommon events in this cohort, experienced by only approximately 4% of patients. It is also possible that VTE develops more frequently after longer periods of hospitalization, which could be precluded by death early in the course of hospitalization for a subset of patients. We performed an adjusted linear regression to evaluate the relative length of stay of patients with TSI who developed VTE in comparison to patients with TSI who did not develop VTE. Ultimately, the possibility that patients with VTE develop this complication following a longer hospital stay is borne out by the fact that the average length of stay for patients with VTE in an adjusted model is around 19 days longer than that of patients with TSI, but without VTE. Of course, it's not possible to say from this analysis whether longer hospital stays disproportionately resulted in VTE or whether patients with VTE had to stay longer as a result of developing a severe complication. The National Trauma Data Bank only includes information from the initial hospitalization. As a result, we were unable to study the long-term outcomes for these patients. Additionally, this analysis was conducted retrospectively, which can make determining causality very difficult from these data. Despite these limitations, using the National Trauma Data Bank gave us access to a sizable national cohort of patients, which facilitated the study of a relatively uncommon complication in the pediatric population. We believe that our findings merit further investigation, particularly with respect to the long-term outcomes of patients who experience VTE. We also believe that these findings should be corroborated by prospective investigations. In conclusion, VTE occurs in a low percentage of pediatric patients hospitalized with traumatic spinal injury. Injury severity is broadly associated with an increased likelihood of developing VTE during initial hospitalization, and this relationship is particularly strong for spinal cord injury. Patients with VTE experience significantly increased length of stay and requirement for rehabilitative services, even when adjusting for injury severity. And we believe that all of these findings merit further study to better understand how this affects patients in a hospital setting. Thank you so much for listening.

venous thromboembolism

pediatric traumatic spinal injuries

injury severity

hospital stays

retrospective cohort study