The relationship between earthquake risk perception and sleep problems in pregnant women.

Venous thromboembolism (VTE) is a major cause of death in patients receiving systemic therapy for cancer; however, it is a potentially preventable condition. Clinical trials have established that interventions such as low-molecular-weight heparin (LMWH) or direct oral anticoagulants (DOAC) can substantially lower the incidence of VTE in high-risk patients. These treatments are not only cost-effective but also demonstrate significant benefits regarding absolute risk reduction and their efficacy-to-safety ratio. Consequently, the ability to identify patients at risk of VTE in a dynamic, real-time manner can enable life-saving and cost-effective prophylactic measures. The success of such strategies hinges on the accurate stratification of patients by VTE risk and the timely implementation of treatment. Several risk assessment models (RAMs) have been created to detect VTE risk. The Khorana Score, a widely used tool, assesses risk based on cancer type and laboratory values but has notable limitations in sensitivity and discriminatory power. Furthermore, its application is restricted to patients with solid tumors and lymphomas who are receiving chemotherapy. More recently, Li et al. developed EHR-CAT, an updated RAM that integrates 11 predictors that are readily available in clinical practice and demonstrates increased discrimination and coverage in multiple settings. Despite these advances, predicting VTE in cancer patients remains a challenge, particularly for cases occurring more than 6 months after the initiation of systematic treatment. A patient’s medical history is critical for assessing future disease risks, yet existing models, while attempting to use this information through aggregation, often fail to capitalize on the rich longitudinal data within EHRs. Model development is further complicated by irregularities in time and data missingness.