Introduction:
Typically, cardiac implantable electronic device (CIED) related infections are caused by Gram positive bacteria. We sought to describe the clinical course and outcomes of patients with Gram negative infections leading to CIED extraction.
Methods:
A retrospective analysis of all CIED extraction procedures at the Sheba Medical Center from January 2013 to September 2016 was performed. Patients with confirmed Gram - bacteria cultures were identified, clinical course and outcomes were described and compared with findings of patients with staphylococcus aureus (SA) infections.
Results:
Of 141 patients who underwent extraction for pocket or systemic infections, 29 had infections with Gram- (20.57%) and 90 had infections with Gram+ bacteria (63.83%), of which 39 had SA. In the Gram- group, 17 patients had Pseudomonas (58.62%), 9 had Enterobacteriaceae (31.03%), one patient had Acinetobacter, one had Haemophilus parainfluenzae, and one had unspecified Gram- rods (3.45% each). Systemic infection was present in 15 patients (51.72%), eight of them had concomitant pocket infection. Isolated pocket infection was present in 14 patients (48.27%). When compared with SA cases, the groups did not vary in age or gender. Appropriate antimicrobial coverage was less likely to be administered empirically in Gram- infections (70.97% vs. 97.44%). Median time from last CIED surgery to symptoms onset was shorter (91 [11.9-678.3] vs. 623.98 days [272.44-1,337.49]) for SA cases (P<0.01). Median time from symptoms onset to CIED extraction was similar (23 days [14-47.5] vs. 17.5 [9.75-55.5] for SA, P=NS). There were no re-infections during follow-up. One-year all-cause mortality rate was 17.24% vs. 20.51% (P=NS).
Conclusion:
Gram- infections account for a significant portion of infection-related extractions, with approximately half of all Gram- patients presenting with systemic infections. Gram- infections are more frequently treated with inadequate empiric ABx. These factors may lead to a more virulent course, resulting in shorter time to symptom onset.