SIR 2024
Embolization
Waseem Wahood, MD, MS
Resident
HCA Aventura Hospital; University of Miami
Financial relationships: Full list of relationships is listed on the CME information page.
To understand the utilization and effect of interventional radiology (IR) treatment on trauma patients that received blood transfusion within the first 4 hours of arrival in the national trauma databank.
Materials and Methods:
The National Trauma Databank was queried from 2013 to 2019 for patients who received packed RBCs or whole blood within first 4 hours after arrival. Using the Abbreviated Injury Scale (AIS) pre-dot codes, patients with trauma to the abdomen, extremities, and thorax were included. Patients with AIS severity score >3 for head, neck, spine, or whole body were excluded. IR intervention was identified as those who underwent angioembolization and angiography with stenting, while angiography only and no angiography where considered “no IR intervention”. Cox proportional hazard regression, adjusted for comorbidities and injury severity, was conducted to assess effect of IR intervention in the cohort and subgroup analysis in the setting of non-operative and operative management. Additionally, inverse probability weighted regression adjustment was used, in which the treatment and outcome models were adjusted for ISS score, concurrent surgical intervention, type of trauma, patient and hospital characteristics
Results:
154,441 patients were identified in the database, 13,142 (8.51%) were in the IR group. Majority of patients were non-operative with no IR intervention (n=77,000, 51.07%). 4,084 (31.08%) patients in the IR group were female, compared to 39,712 (28.11%) in the non-IR group. 9,011 (68.57%) in the IR group were White, compared to 86,014 (60.87%) in the non-IR group.
IR intervention was associated with a lower risk of mortality compared to no intervention by 13% (HR: 0.87; p< 0.001). Compared to males, females were associated lower risk of mortality (HR: 0.87; p< 0.001). White patients had an lower risk of mortality by 11% compared to non-white patients (HR: 0.89; p< 0.001). Propensity weighted analysis indicated that IR intervention was associated with lower rates of mortality (ATE: -0.01; p=0.006). Regarding non-surgical patients, IR intervention was associated with lower risk of mortality (HR: 0.69; p< 0.001). Regarding surgical patients, IR intervention was associated with similar risk of mortality among (HR: 0.96; p=0.20).
Conclusion:
Based on our survival analysis in a national trauma databank, IR was associated with lower mortality risk among both non-operative and operative patients. This result remains with propensity weighted analysis. Additionally, there was a higher proportion of White patients receiving IR compared to non-white patients.