Abstract
INTRODUCTION: Cerebral venous thrombosis (CVT) is an uncommon form of stroke with relatively low mortality but higher incidence in younger adults.1–3 Previous work has suggested decreased overall stroke hospitalization volumes, but preserved CVT hospitalization volumes and increased CVT mortality during the COVID‐19 pandemic.4,5 We sought to provide updated incidence and trend data for cerebral venous thrombosis (CVT) in the United States from 2016‐2020, examine the impact of the COVID‐19 pandemic on CVT, and identify predictors of in‐hospital mortality. METHODS: Validated ICD‐10 codes were used to identify patients with CVT in the National Inpatient Sample (NIS) between 2016 and 2020. The NIS is part of the Healthcare Cost and Utilization Project (HCUP) and is maintained the Agency for Healthcare Research and Quality. The NIS provides a stratified nationally representative 20% sample of all hospital discharges in the United States, excluding rehabilitation and long‐term acute care hospitals. Annual updates to the NIS are released approximately 20 months after the conclusion of the data year. Sample weights were applied to generate nationally representative estimates, and census data were used to compute incidence rates. The first wave of the COVID‐19 pandemic was defined as January‐May 2020. Predictor variables for mortality were selected based upon previous studies of incidence and outcomes of CVT and biological plausibility.6–8 Multivariable logistic regression was conducted using all predictor variables that achieved p<0.10 in univariable regression. Trend analysis was completed using Joinpoint regression. RESULTS: From 2016 to 2020, the incidence of CVT increased from 24.34 per 1,000,000 population per year (MPY) to 33.63 per MPY (Annual Percentage Change (APC) 8.6%; p<0.001). CVT incidence was higher in women than men (37.07 per MPY vs 30.10 per MPY) and the rate of increase was also higher in women (APC 10.1% vs APC 6.8%). Racial differences in incidence rate increases were noted, with incidence increasing by 9.8% annually for White patients, 16.1% for Black patients, and 6.7% for Hispanic patients. All‐cause in‐hospital mortality was 4.9% [95% CI 4.5‐5.4]. On multivariable analysis, use of thrombectomy, increased age, atrial fibrillation, stroke diagnosis, infection, presence of prothrombotic hematologic conditions, and male sex were associated with in‐hospital mortality. CVT incidence was similar comparing the first 5 months of 2020 and 2019 (31.37 vs 32.04; p=0.322) with no difference in median NIHSS (2 [IQR 1‐10] vs. 2 [1‐9]; p=0.959) or mortality (4.2% vs. 5.6%; p=0.176). Mortality was 6.7% [2.3‐17.9] in patients with both CVT and COVID (vs. no COVID 5.5% [4.6‐6.6]; p=0.705). CONCLUSION: CVT incidence increased in the US from 2016 to 2020 while mortality did not change. CVT incidence was higher in women and Black patients. Increased age, prothrombotic state, stroke diagnosis, infection, atrial fibrillation, male sex, and use of thrombectomy were associated with in‐hospital mortality following CVT. During the first wave of the COVID‐19 pandemic, CVT volumes and mortality were similar to the prior year.