A Stitch in Time Saves Clots: Venous Thromboembolism Chemoprophylaxis in Traumatic Brain Injury.

BACKGROUND: Venous thromboembolism chemoprophylaxis (VTE-CHEMO) is often delayed in patients with traumatic brain injury because of the concern for intracranial hemorrhage (ICH) progression. We hypothesize that (1) late time to VTE-CHEMO (≥48 h) is associated with higher incidence of VTE, and (2) VTE-CHEMO use does not correlate with ICH progression. MATERIALS AND METHODS: This is a multiinstitutional retrospective study of patients with traumatic brain injury admitted between 2014 and 2016. Inclusion criteria were head Abbreviated Injury Code ≥2, ICH present on initial head computed tomography, and two or more head computed tomography scans after admission. The primary outcome was VTE, and the secondary outcome was ICH progression. Patients were classified as receiving VTE-CHEMO early (<48 h) or late (≥48 h). Multivariable analysis with Cox proportional hazards regression was performed. RESULTS: Overall, 1803 patients were included. Patients with VTE (n = 137) were more likely to have spinal cord injury, blunt cerebrovascular injury, pelvic or femur fractures, and missed VTE-CHEMO doses. After multivariable regression, body mass index >30 (hazard ratio [HR], 1.05; P = 0.002), Injury Severity Score (HR, 1.004; P < 0.001), pelvic or femur fractures (HR, 1.05; P < 0.0001), spinal cord injury (HR, 1.28; P = 0.02), and missed VTE-CHEMO doses (HR, 1.08; P = 0.01) were significant predictors of VTE. In those who required neurosurgery, late VTE-CHEMO predicted VTE (HR, 1.21; P = 0.0001). Overall, 32% patients experienced ICH progression, which did not correlate with VTE-CHEMO use or timing. CONCLUSIONS: This multicenter study highlights benefits from early VTE-CHEMO and identifies high-risk groups who may benefit from more aggressive prophylaxis. These data also emphasize risk to patients by withholding VTE-CHEMO.

Examining misses in reading aloud repeated words.

This study examined a novel task in which participants read aloud passages shown two words per line on a computer screen. There were four different passages, all of which included unrelated sentences, with each sentence containing one test word. The passages differed only in the text type (prose, scrambled) and in the identity of the test word (the, one). The word the is a common function word, whereas one is a less common content word. The test word was repeated in half of the sentences at the end of one line and at the start of the following line. Many more misses in reading aloud occurred on the than on one, especially for prose passages; almost all misses involved repeated words. These results were interpreted in terms of hypotheses and models that have been proposed for the letter-detection task. Specifically, it is concluded that reading aloud is influenced by structural processes that differentiate between function and content words.