Patient characteristics, presentation, causal microorganisms, and overall mortality in the NatIonal Danish endocarditis stUdieS (NIDUS) registry.

Abstract

Background: Most knowledge on infective endocarditis (IE) comes from large IE cohorts that include patients from tertiary hospitals, leading to referral bias and retrospective population-based studies. This highlights the need for a more detailed characterization of IE in unselected patient cohorts.

Methods: In the National Danish Endocarditis Studies (NIDUS) registry, all hospitalizations in Denmark from 2016 to 2021 with an IE diagnosis were reviewed and validated using electronic medical records (EMR) by healthcare professionals under the supervision of IE experts. Episodes meeting the European Society of Cardiology 2015 modified diagnostic criteria for possible or definite IE were included.

Results: We screened 4390 unique patients, of whom 3557 (81%) were included in the NIDUS registry. Of the 3557 unique patients, 2832 (79.6%) were classified as definite IE and 725 (20.4%) as possible IE. The age was 73.7 years, and most patients were men (68.3%). In total, 689 (19.4%) underwent surgery during hospitalization. The most frequent comorbidities were diabetes (23.7%), heart failure (18.7%), and chronic kidney disease (17.4%). Most patients presented with fever (61.1%), followed by dyspnea (33.0%) and myalgias (27.0%). Sepsis was found in 828 (23.3%) patients, while 378 (10.6%) had signs of embolization at admission. Positive blood cultures were identified in 3191 (89.7%) patients, and the most frequent microbiological etiology was Staphylococcus aureus (31.9%). The in-hospital mortality was 17.3%, and the one-year mortality rate was 31.3%.

Conclusion: The NIDUS registry provides comprehensive, granular, and nationwide data on a cohort of patients with infective endocarditis, revealing that when selection is not restricted to tertiary hospitals or voluntary registries, some important differences emerge. Patients with IE are on average older, have a similar burden of comorbidities, and less often undergo surgery. Minimizing selection bias with the use of a national registry provides a clearer picture of IE as it occurs in real-world clinical settings.