ABSTRACT
AIMS: Device-related infections (DRI) are not well understood in patients implanted with a cardiac resynchronization therapy (CRT) device. The aims of this study were: (i) to evaluate the prevalence of CRT DRI; (ii) to establish the factors predictive of CRT DRI. METHODS AND RESULTS: Between January 2001 and May 2007, CRT implantation was performed in 303 patients (247 men, 82%). The mean follow-up was 31 +/- 19 months. Population characteristics were a mean age of 70 +/- 10 years old; 56 female; aetiology includes (202 dilated and 101 ischaemic cardiomyopathy); NYHA class 3.2 +/- 0.3; LVEF (26 +/- 6%), and a QRS width of 171 +/- 31 ms. Thirteen patients developed a DRI: endocarditis in four, pocket erosion in three, pocket abscess in five, and septicaemia in one. The prevalence of DRI was 4.3%. By univariate analysis, predictive factors of DRI were: procedure time (skin to skin: median of 85 vs. 57.5 min; P = 0.03), re-intervention (54 vs. 6.5%; P < 0.0001), haematoma (31 vs. 8.6% P = 0.01), lead dislodgement (23 vs. 6.2%; P = 0.03), dialysis (23.1 vs. 1.72%; P = 0.003), and procedure type [CRT-ICD (8.6%) vs. CRT PM (1.6%) or system up-grade (1.5%); P = 0.03]. Significant correlations were found between re-intervention and lead dislodgement (r = 0.8; P < 0.001), haematoma (r = 0.2; P < 0.001). Four independent predictive factors of DRI were identified as procedure time (P = 0.002); dialysis (P = 0.0001); re-intervention (P = 0.006), and procedure type (CRT-ICD vs. other procedures; P = 0.01). CONCLUSION: This study found that the prevalence of CRT DRI is close to 4.3% at 2.6 years (1.7% per year incidence). Four independent predictive factors of infections were identified including re-intervention, procedure time, dialysis, and primo CRT-ICD implantation. These parameters should be part of the risk-benefit evaluation in patients selected for CRT implantation.