Background: Infective endocarditis (IE) is a morbid condition with high mortality. We investigated predictors of in-hospital mortality and embolic phenomena in a contemporary Australasian cohort. Methods: We identified all patients with IE admitted between January 2017 and 30th April 2020 (40 months). Patient characteristics, risk factors and clinical outcomes were retrospectively collected and analysed. Results: One hundred and seventy-two consecutive patients (mean age: 56.8 AE 17.9 years, male: 63%, 114/172) were included. Causative organisms were Staphylococcus aureus (44%, 75/172), Enterococcus faecalis (15%, 26/172), Streptococcus mitis (6%, 10/172) and Staphylococcus epidermidis (3%, 6/172). In-hospital mortality was 15% (25/172). Embolic complications were found among 57% (98/172) of patients, the most common being stroke (23%, 40/172), septic pulmonary emboli (17%, 29/172), splenic and/or renal emboli (17%, 26/172) and peripheral limb emboli (15%, 25/172). Sixty (35%, 60/172) patients underwent cardiac surgery. On multivariable analysis, independent predictors of in-hospital mortality were: increased age (odds ratio: 1.064, per year older, P = 0.001), ICU admission independent of cardiac surgery (OR 9.81, P < 0.001), moderate or severe LV impairment (OR 5.19, P = 0.012) and any sign of embolic phenomena to limbs (OR 5.02, P = 0.006). Multivariable predictors of embolic complications were S. aureus bacteraemia (OR 3.22, P = 0.001) and large vegetation >10 mm (OR 3.04, P = 0.002).
Conclusion:We demonstrate predictors of in-hospital mortality and embolic phenomena in our cohort. Though age remains a consistent predictor of mortality, surprisingly, signs of embolic phenomena to the limbs was established as an independent predictor of mortality. The mechanism of this is unclear and warrants further evaluation. ‡ Multivariable model: Area under receiver-operating characteristic curve (AU-ROC) = 0.847 (95% CI 0.765-0.929), P < 0.001.Variables with p < 0.05 on multivariable analysis are presented in bold.