Long-term outcome of infective endocarditis in non-intravenous drug users

To describe postdischarge survival rates and late complications in non-intravenous drug users (non-IVDUs) after treatment of infective endocarditis (IE).

This prospective study consists of consecutive cases of IE in non-IVDUs seen between January 1, 1994, and August 31, 2005. Patient treatment (ie, pharmaceutical and/or surgical) and cardiac valve involved in infection (ie, aortic and/or mitral; whether valve was native or prosthetic) were recorded. Patient follow-up, to March 31, 2007, occurred at 1, 2, 3, and 4 years. Complications, survival, and mortality were statistically analyzed.

During the study period, 230 episodes of IE in 222 non-IVDUs were attended. A total of 143 patients (64%) were discharged from the hospital. Mean +/- SD age of discharged patients was 61+/-17 years. Survival at 1-, 2-, 3-, and 4-year follow-up was 88%, 82%, 76%, and 67%, respectively. Survival was similar for patients with native-valve IE and those with prosthetic-valve IE. The only independent predictors of long-term mortality after discharge were age (hazard ratio, 1.04; 95% confidence interval, 1.01-1.06+/- P=.002) and comorbidity (Charlson index HR, 1.33; 95% confidence interval, 1.18-1.49; P<.001). Surgery during hospitalization showed no clear association with long-term survival. Six patients (4%) had 8 recurrent episodes of IE (1.3% per patient-year). All recurrent episodes happened at 3 months or later after discharge and involved either microorganisms that were of different strains than those of the initial episodes (3 cases) or patients who had suboptimal pharmaceutical or surgical therapy. Only 5 patients (3%) underwent valvular surgery after discharge.

Among non-IVDUs discharged after treatment for IE, 4-year mortality was 33%, and mortality increased with age and comorbidity. Recurrent endocarditis was uncommon in properly treated patients. Survival was similar for patients with native-valve IE and those with prosthetic-valve IE. Survival was also similar for patients who underwent surgery during hospitalization and those who did not.