Search

Introduction The use of implantable cardiac electronic devices (ICEDs) is rising in part due to expanding indications and an aging population [1]. ICED infection occurs in 0.5-2.2% of all implants [1] and implicated in around 10% of cases of infective endocarditis (IE) [2]. ICED infection is associated with prolonged antibiotic therapy and hospital stays, with a mortality rate as high as 35% [1,3,4]. Echocardiography has an established role in the diagnosis of ICED infection and is advocated by guidelines [1], however whether imaging can guide duration of antimicrobial therapy is unclear. Methods We aimed to determine whether echocardiography following ICED extraction was associated with length of antibiotic therapy. Consecutive patients undergoing ICED extraction at Leeds Teaching Hospitals between 01/01/2006 and 31/12/2017 were identified from a prospectively maintained IE database. Patient demographics, microbiology and echocardiographic data were recorded. Results During the study period 64 patients underwent ICED extraction, had suspicion of IE and had available imaging data; of whom median age was 67.5(17) years and 52(81.3%) were male. IE was caused by Staphylococcus species in 44(68.8%), 51(79.7%) episodes were definite according to the Duke criteria and 13 Duke possible. Device removal was usually performed percutaneously in 48(75%), with surgical extraction in 16. Reported complications included lead fracture (1) and superior vena cava dissection (1). Median episode duration (diagnosis to completion of antibiotics) was 40.5(23.75) days, diagnosis to extraction was 17(23) days and extraction to completion of antibiotic therapy was 14(17.75) days. All included patients had an echocardiogram prior to extraction, 41(64.1%) had evidence of vegetation and of these, 32 were confined to the ICED lead whilst nine had endocardial involvement. Following extraction 43(67.2%) underwent echocardiography after a median of 8(7) days, demonstrating residual vegetation on native structures in 11 (figure 1). Compared to those without evidence of persistent vegetation, the duration of antibiotic therapy following extraction (20(10) vs 7.5(15.3) days) and imaging (28(19) vs 14(13.8) days) was longer. Patients without evidence of vegetations on post extraction echocardiogram had similar episode duration to those who did not have further imaging (38.5(23) vs 36(29) days). Twenty (31.3%) were pacing dependent and either received temporary pacing or immediate re-implantation, and a further 37(57.8%) were re-implanted during the index presentation (table 1). Conclusions In this series, in the absence of persistent vegetations on echocardiogram, after removal of an infected ICED, duration of antibiotic therapy was shorter. This suggests imaging can help guide course length, potentially reduce the length of hospital stays, antibiotic risks and selection for resistant bacteria. Conflict of Interest No