Existing guidelines recommend a duration of antibiotic treatment of endocarditis of 4-6 weeks with one or two types of intravenously administered antibiotics. The long hospitalization increases several risks for the patient, including mental strain and increased loss of function. Furthermore, it poses a significant burden on health systems. Current guidelines fail to use available data collected from patients (echo, temperature, CRP, leukocytes, procalcitonin etc.) to determine duration of treatment. A strategy including these data in treatment algorithms ensures an individualized treatment, targeting the patient's course and response to treatment. Thus, the purpose of this open-label, prospective, non-inferiority, RCT study is to investigate the safety and effectiveness of shortening treatment of endocarditis based on the individual patient's initial treatment response, sampling 750 patients, approx. 200 patients with each type of bacteria (Streptococci; Enterococcus faecalis; Staphylococcus aureus). Interim analysis will be conducted when 150 patients have been included, to assess the frequency of the event rate and inclusion rate in order to adjust the intended size of the study population.

Inclusion Criteria:

1. Patients hospitalized and diagnosed with bacterial, left-sided endocarditis as determined by the department clinician responsible based on the revised Duke criteria.
2. The patient may be included <14 days after beginning of relevant antibiotic treatment.
3. Left-sided endocarditis with one of the following microorganisms: Streptococci; Enterococcus faecalis; Staphylococcus aureus.
4. Patients ≥ 18 years.

Exclusion Criteria:

1. Known / suspected immunocompetence (HIV, chemotherapy, prednisolone treatment (> 20 mg / day)).
2. Incapability to give informed consent for participation.
3. Relapse Endocarditis (Endocarditis with the same bacteria within six months).