Beyond the Pocket: Clinical, Routine Biomarker, and Microbiological Associations with Systemic Versus Localized Cardiovascular Implantable Electronic Device Infection.

Background/

Objectives

Cardiac implantable electronic device (CIED) infections range from localized pocket infection to systemic disease. This study evaluated clinical, laboratory, microbiological, and imaging findings associated with the final classification of systemic CIED infection and described device management and in-hospital outcomes.

Methods

This retrospective single-center study included 75 adults with definite CIED infection after case-by-case diagnostic re-adjudication, comprising 61 patients with localized pocket infection and 14 with systemic infection. Exploratory univariable and Firth penalized logistic regression analyses were used to examine associations with systemic infection.

Results

Of 75 patients with definite CIED infection, 14 (18.7%) had systemic infection and 61 (81.3%) had localized pocket infection. Fever was observed in 5/14 (35.7%) patients with systemic infection and 3/61 (4.9%) patients with localized infection (p = 0.005). Procalcitonin was available for 70 patients and was higher in systemic than in localized infection (0.23 [0.06-1.57] vs. 0.05 [0.03-0.12] ng/mL; p = 0.025). Chronic kidney disease stage G3a-G5 was more frequent in systemic infection (3/14 [21.4%] vs. 2/61 [3.3%]; p = 0.042), and estimated glomerular filtration rate was lower (64 [45-83] vs. 89 [73-99] mL/min/1.73 m2; p = 0.014). Among 33 culture-positive episodes, coagulase-negative staphylococci were recovered in 12/22 localized episodes, whereas Staphylococcusaureus was recovered in 6/11 systemic episodes. Complete system removal was documented in 12/14 (85.7%) systemic and 46/61 (75.4%) localized infections; the median interval from admission to first removal attempt was 10 [6-17] and 5 [3-8] days, respectively. In the exploratory Firth model including age, fever, and log2-transformed procalcitonin, fever (aOR 6.17, 95% CI 1.02-37.52) and procalcitonin per doubling (aOR 1.35, 95% CI 1.02-1.81) were associated with systemic infection. In the kidney-function sensitivity model, the procalcitonin estimate was attenuated (aOR 1.31, 95% CI 0.99-1.74). Two in-hospital deaths occurred, both among patients with systemic infection.

Conclusions

In this small retrospective cohort, fever and higher procalcitonin concentrations were exploratory correlates of the systemic-infection classification. The study did not evaluate diagnostic accuracy, clinical prediction, calibration, or external validity. Larger prospective studies with standardized diagnostic work-up and external validation are needed before biomarker-based clinical decisions can be considered.