Predicting Stroke After Trauma

PIs: Heather J. Fullerton and Christine K. Fox

STATUS: Data Analysis
FUNDING: American Heart Association

Although uncommon relative to stroke in older adults, stroke in the young remains an important challenge because it can lead to lifelong disabilities at an early age.  Trauma to cervical or cerebral arteries is a well known cause of stroke in the young; physicians in trauma centers have long faced the challenge of preventing these strokes.  Among patients presenting to trauma centers, the short term risk of stroke is at least 0.18%:  ≈1.5% have evidence of BCVI on vascular imaging, and 12% of those with BCVI will go on to have a stroke (1.5% x 12%=0.18%). Investigators have struggled to (1) identify trauma patients at high-risk for BCVI; (2) develop screening criteria for vascular imaging to detect BCVI; (3) determine the most sensitive and specific imaging modality for BCVI screening; and (4) determine the optimal intervention for primary stroke prevention in patients found to have BCVI. The last objective—optimal treatment—has been the least controversial.  Antithrombotic therapy (i.e., anti-platelets like aspirin, or anticoagulants, like warfarin or heparin) has long been considered standard of care for primary or secondary stroke prevention in the setting of arterial dissection (spontaneous or traumatic) and BCVI in general. Recent observational data suggest that aspirin may be as effective as anticoagulation, making it increasingly favored as a first line therapy given its greater ease of dosing and administration.  Both forms of antithrombotics are relatively safe, with bleeding complications in <1%, although they are considered contraindicated in up to 30% of patients diagnosed with BCVI in a trauma center: those with severe traumatic brain injury, visceral trauma, or multiple orthopedic injuries.

We performed preliminary searches of KPNC databases to estimate the size of the trauma cohort that will be identified for the first two aims of this study.  From 1/1997 through 12/2008, there were a total of 613,854 subjects under 50 years of age with at least one outpatient visit or hospital admission for trauma (an average of 55,805 subjects per study year).  This value includes out-of-system medical encounters.  Of these, 257,257 were children (<20 years of age) and 356,597 were adults.  By extending the cohort through 12/2010, our total cohort should include >700,000 subjects.

To estimate the number of stroke outcomes within this cohort, we searched electronic admissions databases for stroke discharge diagnoses within 12 months of the first trauma encounter.  We considered stroke codes in both the primary diagnostic position (which are more specific for an acute stroke diagnosis) and any diagnostic position (which are more sensitive for a stroke diagnosis); the actual number of stroke outcomes is likely to be between these values.    These data suggest that we will identify between 150 and 400 ischemic stroke outcomes within the 1997-2008 cohort, including between 6 and 49 outcomes for children.   Assuming that all 613,854 subjects had a full 12 months of follow-up after their trauma-related medical encounter, the estimated rate of ischemic stroke within 12 months of trauma would be at least 0.02% (150/613,854) and as high as 0.07% (400/613,854).  
 
A total of 96 subjects (5 children and 91 adults) received a principal discharge diagnosis for hemorrhagic stroke (intracerebral hemorrhage [ICH] or subarachnoid hemorrhage [SAH]) within 12 months of the first trauma encounter.  However, subjects may have received an ICH/SAH code because of intracranial hemorrhage as a direct consequence of trauma rather than as an indirect consequence of trauma to vessels (i.e., rupture of a traumatic pseudoaneurysm or post-traumatic arteriovenous fistula).  Only the latter will be included as post-traumatic strokes in our study, and we will only be able to distinguish the two possibilities through chart review.  However, based on our clinical experience, we anticipate that there will be relatively few post-traumatic hemorrhagic strokes compared to post-traumatic ischemic strokes.  Hence, we conservatively estimate that the total number of post-traumatic stroke cases (ischemic and hemorrhagic) in the 1997-2008 cohort will be between 175 and 450.   For the proposed study cohort that extends through 12/2010, the number of stroke cases will be between 200 and 525.