Risk factors for the development of prosthetic valve endocarditis.

Abstract

Risk factors for the development of prosthetic valve endocarditis (PVE) were analyzed in 2642 patients undergoing initial valve replacement at the Massachusetts General Hospital from 1975 to 1982. Follow-up was available for 2608 patients (98.7%); the mean length of follow-up was 39.8 months. PVE developed in 116 patients (4.4%). The actuarial risk of PVE was 3.1% at 12 months and 5.7% at 60 months. A Cox model was used to identify risk factors for PVE. Recipients of multiple valves had a higher risk of PVE than single valves (p = .01). There was no difference in the risk of PVE for patients receiving aortic valves vs those receiving mitral valves. Recipients of mechanical valves had a higher risk of PVE than recipients of porcine valves in the first 3 months after surgery (p = .02), but the risk of PVE was higher for porcine valve recipients 12 months or more after surgery (p = .004). Despite this difference in the time course of development of PVE, there was no significant difference in the cumulative risk of PVE by 5 years of follow-up between mechanical and porcine valve recipients. Male sex was a risk factor for PVE within 12 months of aortic valve replacement (p = .008) but not thereafter; sex did not influence the risk of PVE after mitral valve replacement. Older patients had a higher risk of late PVE after multiple (p = .04) or mitral valve replacement (p = .08), but not after aortic valve replacement.