Development of cerebral vasospasm following traumatic intracranial hemorrhage: incidence, risk factors, and clinical outcomes

Abstract

OBJECTIVE Limited evidence exists characterizing the incidence, risk factors, and clinical associations of cerebral vasospasm following traumatic intracranial hemorrhage (tICH) on a large scale. Therefore, the authors sought to use data from a national inpatient registry to investigate these aspects of posttraumatic vasospasm (PTV) to further elucidate potential causes of neurological morbidity and mortality subsequent to the initial insult. METHODS Weighted discharge data from the National (Nationwide) Inpatient Sample from 2015 to 2018 were queried to identify patients with tICH who underwent diagnostic angiography in the same admission and, subsequently, those who developed angiographically confirmed cerebral vasospasm. Multivariable logistic regression analysis was performed to identify significant associations between clinical covariates and the development of vasospasm, and a tICH vasospasm predictive model (tICH-VPM) was generated based on the effect sizes of these parameters. RESULTS Among 5880 identified patients with tICH, 375 developed PTV corresponding to an incidence of 6.4%. Multivariable adjusted modeling determined that the following clinical covariates were independently associated with the development of PTV, among others: age (adjusted odds ratio [aOR] 0.98, 95% CI 0.97–0.99; p < 0.001), admission Glasgow Coma Scale score < 9 (aOR 1.80, 95% CI 1.12–2.90; p = 0.015), intraventricular hemorrhage (aOR 6.27, 95% CI 3.49–11.26; p < 0.001), tobacco smoking (aOR 1.36, 95% CI 1.02–1.80; p = 0.035), cocaine use (aOR 3.62, 95% CI 1.97–6.63; p < 0.001), fever (aOR 2.09, 95% CI 1.34–3.27; p = 0.001), and hypokalemia (aOR 1.62, 95% CI 1.26–2.08; p < 0.001). The tICH-VPM achieved moderately high discrimination, with an area under the curve of 0.75 (sensitivity = 0.61 and specificity = 0.81). Development of vasospasm was independently associated with a lower likelihood of routine discharge (aOR 0.60, 95% CI 0.45–0.78; p < 0.001) and an extended hospital length of stay (aOR 3.53, 95% CI 2.78–4.48; p < 0.001), but not with mortality. CONCLUSIONS This population-based analysis of vasospasm in tICH has identified common clinical risk factors for its development, and has established an independent association between the development of vasospasm and poorer neurological outcomes.