Accuracy of Swabs, Tissue Specimens, and Lead Samples in Diagnosis of Cardiac Rhythm Management Device Infections

AIMS: Pacemaker and implantable-cardioverter defibrillator lead infections widely increased with consequent need to accurately recognize responsible bacteria. METHODS: Between May 2003 and December 2007, we extracted 118 leads, 104 (87.3%) due to infections (sepsis, lead-associated endocarditis, pocket infection) or chronic draining sinus (with negative local bacteriological analyses). Swabs and tissue specimens from pocket and fragments of pin and tip of each extracted lead were obtained during extraction and sent for bacteriological examination. RESULTS: Cultures from explanted lead pins returned positive results in 100% of the cases presenting with local infections and in 92.5% of those with chronic draining sinus. In cases of sepsis, positive results of blood samples are less common than lead samples (58.3 vs 86.7, P = 0.02), the latter being more sensitive for infection diagnosis. Concordance between bacterial isolates from pocket and lead is quite low, approaching 45%, seemingly due to contamination effect. Concordance between isolates within the lead (pin and tip) is quite high, close to 70%, reflecting a more accurate expression of the real infection. In cases of sepsis, concordance between lead and blood samples, and mainly from tip and blood, is very high, resembling 80-85%; consequently bacterial isolates from the lead, particularly from lead tip, are clearly associated with clinical infections. CONCLUSIONS: Our results strongly support the hypothesis that chronic draining sinus is often sustained by infection. Moreover, diagnostic accuracy of lead samples is higher not only than swabs and tissue samples, but also than blood samples to confirm an infection and to guide effective therapy.