There are very few data on the prognosis of possible versus definite infective endocarditis (IE). We studied data from 365 consecutive patients with IE involving native heart valve seen in an academic institution from 1990 to 2012. Patients were classified according to the modified Duke criteria for IE: patients with possible IE (n = 101, 28%) and those with definite IE (n = 264, 72%). Patients with possible IE were older than those with definite IE (66 ± 15 vs 62 ± 16, p = 0.05). A causative microorganism was identified in 66% of patients with possible IE versus all patients with definite IE (p <0.0001) and only 41% had major echocardiographic criteria (vs 100%; p <0.0001). Overall, 139 patients died over a mean ± SD follow-up of 3.9 ± 4.5 years (median 2.2, interquartile range 5.9 years). Patients with possible and definite IE had a similar risk of death. Independent predictors of long-term mortality were increasing age (hazard ratio [HR] 1.02, 95% confidence interval [CI] 1.01 to 1.04; p = 0.0009), vegetation length >15 mm (HR 1.87, 95% CI 1.14 to 3.06; p = 0.01), and stroke (HR 4.10, 95% CI 1.84 to 9.17; p = 0.0006), whereas infection of mitral valve (HR 0.57, 95% CI 0.34 to 0.94; p = 0.03) and surgery (HR 0.43, 95% CI 0.19 to 0.99; p = 0.05) were associated with a better prognosis. Patients with definite IE and those with possible IE who did not undergo surgery had a worse prognosis than their counterparts with surgery. In conclusion, unselected patients with possible IE (Duke criteria) had a similar prognosis than those with definite IE.
Copyright © 2017 Elsevier Inc. All rights reserved.