Long-term mortality in patients who survive surgery for infective endocarditis versus the background population: a nationwide study

BACKGROUND: Valvular surgery for infective endocarditis (IE) can improve survival but carries substantial risk. Limited data exist on long-term outcomes for patients who survive surgery compared with a background population. We aim to compare long-term mortality and morbidity in patients with IE, who survive 90-days following valvular surgery to a matched Danish background population.

METHODS: Using Danish registries, we identified patients who survived >90 days post-valvular surgery for first-time IE (2010-2023). Each patient was matched 1:3 with controls from the background population by age, sex and selected comorbidities. All-cause mortality was assessed at 5 and 10 years using the Kaplan-Meier estimator and the multivariate Cox model. Further, we examined the time spent in hospital during the first year following the index date.

RESULTS: We identified 1050 patients (77.5% male, median age 65.8 years) surgically treated for IE and 3150 controls. The most common pathogens were: Viridans group streptococci (44.6%), Staphylococcus aureus (23.1%), Enterococci (17.9%). Patients with IE had higher absolute 5-year (20.1% vs 12.9%, p=0.001) and 10-year (38.5% vs 27.9%, p<0.001) mortality compared with controls. Adjusted 5-year and 10-year mortality rates were also higher in patients with IE (5-year HR=1.49 (95% CI 1.24 to 1.79) and 10-year HR 1.38 (95% CI 1.19 to 1.60)). Patients with IE experienced more frequent and longer hospitalisations within the first year postsurgery, as 36.8% patients with IE were hospitalised within the first year following index compared with 17.5% in the matched controls. 9.3% of patients with IE were hospitalised for >14 days compared with 3.0% in matched controls. 60 patients with IE (5.7%) died within the first year following index compared with 51 (1.6%) in the matched controls.

CONCLUSION: Patients with IE who underwent valve surgery had a higher crude mortality than their controls from the background population. After adjusting for confounders, 5-year and 10-year mortality rates remained higher in the IE patient population.