to investigate clinical properties of course and outcomes of infective endocarditis (IE) depending on source of infection, to find predictors of mortality in a Moscow general hospital.

We included in this study 176 patients with definite and possible infective endocarditis (the Duke criteria), admitted in our hospital in 2010-2017. Patients were divided in three groups according to source of infection. All patients underwent standard clinical and laboratory assessment, echocardiography, blood culture test combined with blood PCR with sequencing. Inhospital and 1-year outcome were evaluated.

Among 176 patients with IE 65.3 % were men (median age 57 [35-72] years), most patients (n=149, 84.7 %) had native valve IE. Etiological factor was identified in 127 (72.2 %) cases. Gram-positive infective agents prevailed (54 %). Surgery in active phase of the disease was performed in 30 (17 %) patients. Among patients with healthcare-associated IE (n=76, 43.9 %) prevailed those older than 60 years, with high Charlson comorbidity index, with culture-negative IE, and complicated clinical course (mainly progressing heart failure). Patients with intravenous drug use associated IE (n=50, 28.4 %) had low Charlson index, association with hepatitis C viral infection, involvement of tricuspid valve with big vegetations, high frequency of embolic complications, and low inhospital mortality. Group of patients with community acquired IE (n=50, 28.4 %) more often had uncommon causative microorganisms, and had better long-term outcome. In-hospital mortality was 30.1 % (n=53) mostly due to sepsis with multi-organ failure, and heart failure. Risk factors of inhospital death were history of cardiovascular diseases, old age, kidney damage, methicillin-resistant Staphylococcus aureus (MRSA) infection, uncontrolled infection, and embolic events. Risk factors of 1-year mortality were history of stroke, and heart failure as IE complication. Independent predictors of in-hospital death were MRSA infection (odds ratio [OR] 50.32, 95 % confidence interval [CI] 1.66-213.92; p=0.002), persistent infection (OR 18.6, 95 %CI 5.37-64.40; p=0.001), duration of fever >7 days after initiation of antibacterial therapy (OR 13.41, 95 %CI 3.51-51.24; p=0.001); and of death during first year - history of cerebral infarction (OR 4.39, 95 %CI 1.32-14.70; p=0.016)), and heart failure as IE complication (OR 8.1, 95 %CI 1.97-67.09; p=0.016). Among patients subjected to surgery there were no fatal outcomes during 1 year after hospital discharge, while among conservatively treated patients were 21 (14.4 %) deaths (p<0.009).

Main clinical features of IE course in patients urgently admitted to a general hospital was dominance of healthcare-associated  IE among patients, who were older than 60 years with severe comorbidities. These patients had more complications and worse outcome. Modeling of prognosis identified uncontrolled infection as key factor of unfavorable outcome. Surgery significantly reduced long-term mortality.