Epidemiology, diagnosis, and treatment of endocardi- tis have evolved 24 and changed the long-term outcome, disclosed both in clinical 6,7 and in surgical series. 16,17,19,22,23,25 Although these studies added important contributions to the subject, we hypothe- sized that controversial or elusive prognostic factors might be further evaluated in larger series of patients and additional risk factors identified to help assess the long-term prognosis. Thus we studied the frequency and risk factors for relapses and recurrences of infective endocarditis, valve replacements, and deaths in the long-term follow-up of a large series of patients with different cardiac conditions after successful treatment of infective endocarditis. Methods Patients We studied an inception cohort of 420 patients who sur- vived infective endocarditis from October 1978 to August 1994 and were discharged from a tertiary 423-bed university hospital dedicated to cardiology. Patients’ ages ranged from 2 months to 83 years (mean 34.2 ± SD 17.2 years); 270 (64.3%) See related Editorial on page 6. Morbidity associated with infective endocarditis may extend beyond the successful treatment of the active disease and hospital discharge and includes relapse or recurrence, 1-6 valve replacement, 2,6,7 and death 1,2,7-10 in the long-term follow-up. In different studies recurrences were reported in 2.4% to 10.9% of the patients. 1,4,6-8,11-13 Valve replacement during the follow-up was performed in 17.8% to 46.2% of the patients. 6,7,10,14-17 The 10-year survival rate after endocarditis ranged between 47% and 81% 2,6,7,14,18 and between 50% and 60% in surgical series. 16,19-23 From the Heart Institute-InCor, São Paulo University Medical School, São Paulo, Brazil. Submitted January 13, 2000; accepted August 14, 2000. Reprint requests: Alfredo José Mansur, MD, Heart Institute-InCor, São Paulo Univer- sity Medical School, Av Dr Eneas de Carvalho Aguiar, 44, São Paulo, Brazil 05403-000. E-mail: ajmansur@incor.usp.br Copyright © 2001 by Mosby, Inc. 0002-8703/2001/$35.00 + 0 4/1/111952 doi:10.1067/mhj.2001.111952 Valvular and Congenital Heart Disease Relapses, recurrences, valve replacements, and mortality during the long-term follow-up after infective endocarditis Alfredo José Mansur, MD, Creusa M. R. Dal Bó, MSc, Júlia T. Fukushima, MSc, Vitor S. Issa, MD, Max Grinberg, MD, and Pablo M. A. Pomerantzeff, MD São Paulo, Brazil Background Late prognosis after infective endocarditis has not been systematically studied in large series of patients with different underlying heart conditions in recent years. Methods We studied an inception cohort study of 420 patients discharged after treatment of endocarditis from a uni- versity tertiary care hospital. The patients were aged 34.2 ± 17.2 years (mean ± SD), ranging from 2 months to 83 years; 270 (64.3%) were men and 150 (35.7%) were women. Mean follow-up was 6.1 ± 4.3 years for survivors and 3.7 ± 3.7 years for the patients who died during the follow-up. We studied the frequency and risk factors for relapses and recurrences of endocarditis, cardiac valve replacements, and deaths during the follow-up. Statistical analysis was performed through comparison of groups, of event-free survival, and risk ratios. Results Relapses were observed in 14 (3.3%) patients. There was one recurrence of endocarditis in 48 (11.4%) patients, two (0.5%) in 2 patients, three in 1 patient (0.2%), and five (0.2%) in 1 patient. Valve replacement was performed in 83 (19.7%) patients. Ninety-eight (12.3%) patients died. Risk factors for recurrent endocarditis were increasing age (risk ratio 1.02) and male sex (risk ratio 1.61). Risk factors for valve replacement were recurrent endocarditis (risk ratio 1.62) and prosthetic valve endo- carditis (risk ratio 1.61). Risk factors for death were increasing age (risk ratio 1.03) and recurrent endocarditis (risk ratio 2.06). Conclusions The long-term event-free survival for patients who survived their first episode of endocarditis was low. Recurrent endocarditis, prosthetic valve endocarditis, and increasing age contributed to the high rate of events during the fol- low-up. (Am Heart J 2001;141:78-86.)