Objective To assess long-term outcomes and predictors of mortality in patients treated according to current recommendations for cardiac implantable electronic device (CIED) infection.

Design Two-group matched cohort study.

Setting Tertiary-care institution.

Patients Consecutive patients admitted for CIED infection between 2004 and 2008 were prospectively enrolled. Study subjects were matched to a cohort of uninfected CIED patients by age, sex and type of device.

Interventions In all infected patients, the therapeutic approach consisted of complete hardware removal whenever possible, antimicrobial therapy, and implantation of a new device, if indicated. Patients were systematically followed, with standardised outcomes assessment.

Main outcome measures All-cause mortality and predictors of long-term mortality.

Results 197 patients were included and matched 1:1 to controls. Pocket infections were present in 41.1% and definite or suspected infective endocarditis in 58.9%. Total or subtotal hardware removal was achieved in 98.5% of cases. Median follow up was 25 months (12–70). Mortality rates in the study group and controls were 14.3% vs 11.0% (NS) at 1 year and 35.4% vs 27.0% (p=NS) at 5 years. Independent predictors of long-term mortality were older age (HR=1.09, p<0.001), cardiac resynchronisation therapy (HR=3.70, p=0.001), thrombocytopenia (HR=5.10, p=0.003) and renal insufficiency (HR=2.66, p=0.006). In patients with reimplanted devices, epicardial right ventricular pacemakers were associated with higher mortality (HR=2.85, p=0.034).

Conclusion In patients with CIED infection managed by recommended therapy, long-term mortality rates are similar to comparable controls. Independent predictors include patient and disease-related factors, in addition to implantation of right ventricular epicardial pacemakers.