Incidence of Cerebral Venous Thrombosis: A Population-Based Study, Systematic Review, and Meta-Analysis

Abstract

Background: Studying the baseline incidence of cerebral venous thrombosis (CVT) prior to COVID-19 and the limitations of how this has been previously reported in the literature will help improve understanding of this disease and how risks may have changed in the post-COVID era. Methods: We examined CVT incidence using linked administrative data in British Columbia, Canada (population 5.2 million). To contextualize our findings, we also examined CVT incidence in the published literature and searched MEDLINE and EMBASE for article titles and abstracts up to Nov 2, 2021 on CVT incidence in adults. We performed abstract screening and full-text review prior to data extraction and explored associations between CVT incidence and year of study, geographic location, and study quality with meta-analyses and meta-regression. A random-effects restricted maximum likelihood model was used. Publication bias was assessed using the Egger tests and using visual inspection of the funnel plot for symmetry. Results: There were 554 unique CVT cases (mean age 50.9 years, 55.4% women) in British Columbia from 2000 to 2017; overall annual incidence was 8.7 (95%CI‚ 8.0–9.4) per million. Incidence increased over time in men across the entire study period, and from 2011 to 2017 in women. We identified 22 other studies on CVT incidence before 2020 (21/23 total studies included in meta-analysis). Annual incidence overall was 12.1 (95% CI‚ 9.9–14.3) per million with significant between-study heterogeneity (I 2 98.8%, Qp-value<0.001). There were no significant associations on meta-regression between incidence and study year, study quality score, or gross national income per capita of the study country. Visual inspection of the funnel plot and a significant Egger test (z=2.8, P <0.01) suggested possible publication bias. Conclusions: Incidence of CVT in Canadian data increased over time but remained lower than in other population-based studies. Significant heterogeneity exists in the literature, which may be subject to publication bias.