Atrial fibrillation as an independent risk factor for venous thromboembolism in intracerebral hemorrhage patients: a multicenter retrospective cohort study

Background Venous thromboembolism (VTE) represents a major and potentially fatal complication among patients with intracerebral hemorrhage (ICH); however, accurate risk stratification remains difficult in this population. The present study sought to determine independent predictors of VTE in patients with ICH using data derived from a large multicenter critical care database. Methods A retrospective cohort study was performed using the eICU Collaborative Research Database. Adult patients with a diagnosis of ICH were included. VTE events were identified through an integrated multi-modal approach incorporating diagnostic coding, clinical text mining, and records of therapeutic anticoagulation. Data on baseline demographics, comorbid conditions, laboratory parameters, and thromboprophylaxis practices were collected. Univariate and multivariate logistic regression models were applied to identify independent risk factors for VTE. Sensitivity analyses were conducted using alternative VTE definitions to assess the robustness of the findings. Interaction analyses and additional sensitivity analyses excluding anticoagulation-only VTE cases were performed to further evaluate the robustness of the primary findings. Results A total of 785 patients with ICH were analyzed, of whom 53 (6.75%) developed VTE during hospitalization. In univariate analyses, nine variables were significantly associated with VTE occurrence, including atrial fibrillation (AF) (OR = 5.75, 95% CI: 2.95–11.21, p < 0.001), chronic obstructive pulmonary disease (COPD) (OR = 6.81, 95% CI: 2.67–17.38, p < 0.001), and use of mechanical ventilation (OR = 2.53, 95% CI: 1.38–4.64, p = 0.003). After adjustment for potential confounders in multivariate analysis, atrial fibrillation emerged as the only independent risk factor for VTE (adjusted OR = 3.84, 95% CI: 1.75–8.41, p < 0.001). The incidence of VTE was substantially higher among patients with AF compared with those without AF, representing an unadjusted 4.6-fold higher VTE incidence (24.2% vs. 5.3%). This association remained consistent across multiple sensitivity analyses, including analyses excluding anticoagulation-only VTE cases (adjusted OR = 3.99, 95% CI: 1.02–15.55, p = 0.046). Conclusion Atrial fibrillation was the only independent risk factor for VTE after multivariable adjustment, conferring an approximately four-fold increase in VTE risk, and should be interpreted as a clinical risk marker rather than a direct causal factor. Patients with ICH complicated by AF constitute a particularly high-risk subgroup and may benefit from intensified surveillance and individualized thromboprophylaxis strategies.