Blood stream infection in patients with implantable-cardioverter defibrillator admitted to non-cardiology units: incidence, diagnostic work-up and outcome

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Background There is limited knowledge on the incidence of blood stream infections (BSI) in patients with implantable cardioverter defibrillators (ICDs) or cardiac resynchronization therapy devices (CRTs), as well as about the diagnostic work-up and outcome when these patients are admitted to non-cardiology units. Purpose We assessed the incidence of BSI in patients with an ICD or CRT, evaluated the diagnostic work-up performed, and investigated the outcome. Methods This is a single-center, retrospective cohort analysis of all 515 patients implanted with a de-novo ICD or CRT between 2012 and 2021 at our centre, and subsequently admitted to a non-cardiology unit of any public hospital in Canton Ticino with a BSI (≥ 2 positive blood cultures). Device infection was defined according to the 2019 International Cardiac Implantable Electronic Device (CIED) Infection Criteria. Results A total of 43 patients were diagnosed with BSI during a median follow-up of 48 months (IQR 24-78 months). Incidence rate for BSI was 19 (95%-CI 14 – 26) and 23 (95%-CI 14 – 35) per 1000 patient-years for ICDs and CRTs, respectively. The most common source for BSI was urinary tract infection (21%), which was mostly caused by gram positive pathogens (68%) with coagulase-negative staphylococci being the most frequent one (39%). Definite CIED infection rate in patients presenting with BSI was 11.6% (see table). The majority of patients (56%) had no cardiac imaging including transthoracic or transesophageal echocardiography; a PET-CT was performed in only 4 patients (9%). Nearly all patients (86%) were treated with short-term antibiotics; in contrast, system extraction (2.3%) and chronic antibiotic therapy (4.7%) were rarely performed. Throughout the entire follow-up period, patients with BSI had a 8-fold higher probability (HR 7.8, 95% CI 4.8-12.7; p<0.001) of all-cause mortality compared to those without BSI. Conclusions Awareness in non-cardiology units about possible device infection in patients with BSI is rather low leading to limited diagnostic work-up and ultimately to a high mortality rate of these patients.