METHODS
Admissions for adult patients with cAVM ruptures were identified in the National Inpatient Sample during the period from 2015 to 2019. Complex samples multivariable logistic regression and chi-square automatic interaction detection (CHAID) decision tree analyses were performed to identify significant associations between clinical covariates and the development of vasospasm, and a cAVM–vasospasm predictive model (cAVM-VPM) was generated based on the effect sizes of these parameters.
RESULTS
Among 7215 cAVM patients identified, 935 developed vasospasm, corresponding to an incidence rate of 13.0%; 110 of these patients (11.8%) subsequently progressed to delayed cerebral ischemia (DCI). Multivariable adjusted modeling identified the following baseline clinical covariates: decreasing age by decade (adjusted odds ratio [aOR] 0.87, 95% CI 0.83–0.92; p < 0.001), female sex (aOR 1.68, 95% CI 1.45–1.95; p < 0.001), admission Glasgow Coma Scale score < 9 (aOR 1.34, 95% CI 1.01–1.79; p = 0.045), intraventricular hemorrhage (aOR 1.87, 95% CI 1.17–2.98; p = 0.009), hypertension (aOR 1.77, 95% CI 1.50–2.08; p < 0.001), obesity (aOR 0.68, 95% CI 0.55–0.84; p < 0.001), congestive heart failure (aOR 1.34, 95% CI 1.01–1.78; p = 0.043), tobacco smoking (aOR 1.48, 95% CI 1.23–1.78; p < 0.019), and hospitalization events (leukocytosis [aOR 1.64, 95% CI 1.32–2.04; p < 0.001], hyponatremia [aOR 1.66, 95% CI 1.39–1.98; p < 0.001], and acute hypotension [aOR 1.67, 95% CI 1.31–2.11; p < 0.001]) independently associated with the development of vasospasm. Intraparenchymal and subarachnoid hemorrhage were not associated with the development of vasospasm following multivariable adjustment. Among significant associations, a CHAID decision tree algorithm identified age 50–59 years (parent node), hyponatremia, and leukocytosis as important determinants of vasospasm development. The cAVM-VPM achieved an area under the curve of 0.65 (sensitivity 0.70, specificity 0.53). Progression to DCI, but not vasospasm alone, was independently associated with in-hospital mortality (aOR 2.35, 95% CI 1.29–4.31; p = 0.016) and lower likelihood of routine discharge (aOR 0.62, 95% CI 0.41–0.96; p = 0.031).