Abstract Background Administering antithrombotic therapy (ATT) in patients with infective endocarditis (IE) involves a complex balance of bleeding and thromboembolic risks. Data on outcomes beyond the acute phase remain limited. This retrospective single-center cohort study had two aims: first, to describe the use of anticoagulation during the acute phase of left-sided IE; and second, to examine, without inferring causality, how anticoagulation, as used in routine care, correlated with in-hospital and long-term clinical outcomes, including mortality and neurological events. Methods ATT in patients with left-sided IE was assessed retrospectively and categorized into two groups: any therapy that included anticoagulation (AC) and therapy without anticoagulation (No-AC). Two observational periods were analyzed: the in-hospital phase and the period beginning 3 months after discharge, when 30% of patients had their ATT modified. Vital and neurological status were obtained by standardized telephone follow-up (mean follow-up time 4.2 ± 3.1 years). Log-rank tests, Kaplan–Meier estimates, Cox regression analyses, and matched analyses were used to explore correlations between ATT and these outcomes. Results A total of 504 hospitalized patients (mean age 65 ± 13 years, 25% female) with left-sided IE were included. During inpatient treatment, 83 patients (16%) died, with no relevant difference between AC and No-AC groups. During follow-up, patients in the AC group showed a more favorable value for the combined endpoint of mortality and unfavorable neurological function (P = 0.029) that was driven primarily by higher survival rates (P < 0.001). In Cox regression analyses, higher age, CHA₂DS₂-VA score, EuroSCORE II, Staphylococcus aureus bacteremia, and atrial fibrillation were each linked to a higher hazard of the combined endpoint, whereas AC showed an inverse correlation. Consecutive matched analyses yielded similar results. Conclusion In this retrospective cohort, anticoagulated patients did not show a higher rate of adverse events during hospitalization and had a lower long-term event rate. These findings represent correlations observed in a non-randomized, single-center setting and may partly reflect differences in underlying risk profiles and treatment selection (confounding by indication and residual confounding). Prospective studies are needed to confirm any causal effects and to define more precisely the role of ATT in patients with IE and elevated cardiovascular risk. Graphical Abstract
Posner et al. (Mon,) studied this question.