Infective Endocarditis: Predictive Factors for Diagnosis and Mortality in Surgically Treated Patients

Figure 1. Contingency analysis of qualitative diagnostic tools. Sensitivity (a) and specificity (b) for IE diagnosis, 30-day and 1-year mortality of imaging, microbiologic, nuclear and histopathological diagnostics. HP = histopathology; other abbreviations as Table 2.

Figure 2. Simple logistic regression of quantitative diagnostic tools. Receiver-operating characteristic curves showing the predictive power for diagnosis of either IE (a–e) of white blood cell count (WBC), C-reactive protein (CRP), procalcitonin (PCT), vegetation size and N-terminal prohormone of B-type natriuretic peptide (NT-proBNP) or 1-year mortality of NT-proBNP (f).

Figure 3. Intergroup differences for quantitative parameters. Scatter plots of serum markers for inflammation (WBC, CRP and PCT, a–c), of a marker for cardiac stress (NT-proBNP, (d) and of the vegetation size (measured by transesophageal echocardiography, (e) in patients who underwent operative therapy for IE with confirmed (IE+) or rejected (IE-) diagnoses for IE. (f): Patients with positive (Mort+) vs. negative mortality (Mort-) after 1 year of follow-up. Red dotted lines represent the optimal cut-off values for vegetation size (d) and log(NT-proBNP) (f), comparison of Table 3 and Table 4b. Bars represent the mean ± SD. * p < 0.01, ** p <0.0001, Student’s t-test. Abbreviations as in Table 1.

Figure 4. Presence of systemic embolism reduces the probability of survival. Kaplan–Meier estimates showing the probability of survival of patients with IE and surgical treatment and who had systemic embolism (Emb+) compared to patients without embolism (Emb-). The p value for the intergroup difference is < 0.001; see text for details.