Mortality in cerebral venous thrombosis: Results from the national inpatient sample database

Background: Outcomes of cerebral venous thrombosis (CVT) vary from full recovery to death. Few studies have been performed examining epidemiologic and medical risk factors associated with high mortality in CVT. In this study, we examined the National Inpatient Sample (NIS) to determine the epidemiologic and medical risk factors associated with increased mortality from CVT. Materials and Methods: Using the NIS from 2001 to 2008, patients who suffered from CVT were identified using the ICD-9 codes 437.6 (nonpyogenic thrombosis of intracranial venous sinus), 325 (phlebitis and thrombophlebitis of intracranial venous sinuses) and 671.5 (peripartum phlebitis and thrombosis, cerebral venous thrombosis, thrombosis of intracranial venous sinus). We analyzed the associations of demographic factors, risk factors, comorbidities, complications of CVT, and therapeutic interventions with in-hospital mortality. We performed a multivariate logistic regression analysis to determine which variables were independently associated with in-hospital mortality. Results: 11,400 patients were hospitalized with CVT between 2001 and 2008. Two-hundred and thirty-two (2.0%) suffered in-hospital mortality. Patients 15-49 years old had the lowest mortality rate (1.5%) compared with 2.8% for patients aged 50-64 (p < 0.001) and 6.1% for patients ≥65 years old (p < 0.001). The most common condition associated with CVT was pregnancy/puerperium (24.6%), and these women had a low mortality rate (0.4%). On multivariate analysis, the comorbidity most strongly associated with increased risk of mortality was sepsis (mortality rate 15.6%, OR = 7.5, 95% CI = 4.79-11.53, p < 0.001). Malignancy, underlying autoimmune disease and substance abuse were also independently associated with mortality, but with lower mortality rates (<5%). Complications associated with increased risk of mortality included paralysis (8.0%, OR = 3.4, 95% CI = 3.17-6.96, p < 0.001), intracranial hemorrhage (8.7%, OR = 5.4, 95% CI = 4.38-7.96, p < 0.001), and hydrocephalus (15.0%, OR = 3.2, 95% CI = 5.54-15.11, p = 0.004). Demographic variables associated with decreased mortality on multivariate analysis were male gender (2.1%, OR = 0.62, 95% CI = 0.43-0.87, p = 0.006) and Asian/Pacific Islander race (OR = 0.00, 95% CI = 0-0.27, p < 001). Conclusions: CVT is associated with a low in-hospital mortality rate. Amongst patients suffering CVT, male gender and Asian/Pacific Islander race were independently associated with lower odds of in-hospital mortality when compared to their female and white counterparts, respectively. Septic patients with CVT have the greatest risk of in-hospital mortality. Hydrocephalus, intracranial hemorrhage, and motor deficits are also associated with higher risk of death. Our results build on previous evidence that serves to define a group of patients with CVT at high risk of early death.