Key points are not available for this paper at this time.
Background and Objectives: Infective endocarditis (IE) remains associated with high mortality, and real-world (RW) patients often differ from trial populations. We evaluated predictors of complications and mortality, the trial-eligibility gap, and temporal trends in guideline adherence across two periods (Period 1 P1: 2011–2016 vs. Period 2 P2: 2017–2025) in a Romanian county hospital. Materials and Methods: We conducted a retrospective analysis of consecutive adult patients with definite IE. Patients were categorized as trial-eligible (TE) or RW according to predefined criteria. The composite endpoint included acute heart failure, cardiogenic or septic shock, embolic events, infectious complications, need for renal replacement therapy, and in-hospital mortality. We evaluated guideline adherence using a predefined quality indicator (QI) score ≥ 3. We identified independent predictors of outcome using multivariable logistic regression. Results: Among 206 patients (mean age 63.0 ± 14.8 years; 70.4% male), blood cultures were positive in 64.1%, with Staphylococcus aureus accounting for 14.1%. Vegetations were documented in 72.8%, and cardiac surgery was performed in 26.2%. Overall, at least one event from the composite endpoint occurred in 61.6%, and mortality was 32.5%. TE patients represented 63.1% of the cohort. Guideline adherence improved over time (QI ≥ 3: from 18.3% in P1 to 25.4% in P2, p = 0.32). In P2, the composite endpoint (66.8% vs. 42.9%, p = 0.002) and embolic events (31.8% vs. 8.2%, p < 0.001) were more frequent, whereas mortality remained unchanged (31.8% vs. 34.7%, p = 0.844). Sepsis at admission and left ventricular ejection fraction (LVEF) < 50% independently predicted adverse outcomes; model discrimination was acceptable, with an area under the curve (AUC) of 0.77. Conclusions: RW IE showed high complication rates and a persistent trial gap; improved guideline adherence was offset by greater clinical severity.
Pop et al. (Fri,) studied this question.