Clinical Predictors of Cardiovascular Implantable Electronic Device-Related Infective Endocarditis

Methods: We retrospectively reviewed all cases of CIED infection seen at Mayo Clinic Rochester between 1991 and 2008. CIED-IE was classified using pathologic and clinical criteria. Clinical predictors of CIED-IE were identified using logistic regression, and quantified using a summary score and plotted against the distribution of CIED-IE. Results: Ninety-three (22.4%) of the 416 patients with CIED infection had CIED-IE. Host factors including chronic immunomodulator therapy exclusive of corticosteroid (odds ratio [OR], 3.79 [confidence interval (CI) 1.10, 13.04]), chronic corticosteroid therapy (OR, 2.15 [CI 0.93, 5.00]), hemodialysis (OR, 3.24 [CI 1.39, 7.55]), or remote infection (OR, 1.77 [CI 0.99, 3.14]) were associated with increased odds of CIED-IE. Patients with CIED-IE were at increased odds of presenting with fever (OR, 3.78 [CI 1.93, 7.40]), or malaise (OR, 1.87 [CI 1.02, 3.41]), and have findings of leukocytosis (OR, 3.61 [CI 1.51, 8.62]). In marked contrast, they were at decreased odds of exhibiting signs/symptoms of infection at the generator pocket site (OR, 0.19 [CI 0.10, 0.36]). Summary scores of 6 and 11 predicted CIED-IE in approximately 50% and 90% of cases, respectively. Conclusions: Certain comorbid conditions and systemic manifestations of infection were associated with CIED-IE. In contrast, pocket site infection was negatively associated with CIED-IE. These findings should assist clinicians in identifying patients who would more likely benefit from further investigation of CIED-IE with transesophageal echocardiography. (PACE 2011; 34:450-459).