Patients included those with native and prosthetic valves.
Survival was for the 5 years after discharge for the initial hospitalization.
The authors used propensity scores for matching patients treated with surgery and patients treated medically to compensate for treatment bias in a retrospective study.
Patients who were more likely to undergo surgery:
(1) younger age (an elderly patient might not be considered a good surgical risk)
(2) transfer from an outside hospital
(3) more severe congestive heart failure
(4) intracardiac abscess
(5) evidence of infective endocarditis on physical examination (embolization, murmur, etc)
(6) Staphylococcus species as pathogen
(7) undergoing hemodialysis without a chronic catheter (as opposed to having a chronic catheter)
The presence of contraindications to surgery is important in determining who does not have surgery.
Risk factors for mortality:
(1) medical therapy (surgical therapy associated with reduced risk)
(2) diabetes mellitus
(3) presence of a chronic intravenous central catheter (Hickman, port-a-cath, etc.)
(4) paravalvular complications (fistula, abscess or perforation) seen on echocardiography
