Venous Thromboembolism Prophylaxis in Hospitalized Medical Patients: A Narrative Review of Risk Stratification, Treatment Duration, and Bleeding Trade-Offs

Venous thromboembolism (VTE) remains a leading preventable cause of morbidity and mortality in hospitalized medical patients, yet decisions about who should receive prophylaxis -and for how longremain far from settled.This review synthesizes evidence from 22 sources, including randomized trials, prospective and retrospective cohorts, systematic reviews, and major society guidelines, addressing risk stratification, in-hospital anticoagulant prophylaxis, extended-duration direct oral anticoagulants (DOACs), the associated bleeding trade-offs, and real-world implementation.The evidence broadly supports pharmacologic prophylaxis during hospitalization for appropriately selected, acutely ill medical patients with acceptable bleeding risk, with the strongest randomized evidence favoring low-molecular-weight heparin during the inpatient period.That benefit comes with trade-offs worth weighing carefully, since even well-supported regimens carry a measurable increase in bleeding risk alongside their protective effect.Commonly used risk-assessment tools have real limitations as well: the Padua and IMPROVE scores offer a useful clinical framework, but head-to-head comparison shows both discriminate poorly between patients who will and won't develop VTE, and real-world implementation data reveal frequent misclassification and inconsistent dose administration.Extended-duration anticoagulation after discharge follows a similar pattern across the major DOAC trials in this space -fewer thrombotic events, but more bleeding, without a clear survival benefit.Taken together, this evidence points toward a risk-adapted rather than protocol-driven approach: assess VTE and bleeding risk at admission, initiate appropriate inpatient prophylaxis, reassess as the clinical picture changes, confirm that prescribed doses are actually being given, and reserve extended anticoagulation after discharge for carefully selected patients rather than applying it as a default.Meaningful progress in this area will likely come less from another static risk score and more from models that can track mobility, bleeding risk, and treatment response dynamically over the course of a hospital stay.