Cardiac Electrophysiologist University of Ottawa Heart Institute, Canada
Background: Cardiac implantable electronic device (CIED) infections are associated with significant morbidity, mortality, and healthcare costs. The PADIT risk score may help identify high-risk for infection for targeted prevention strategies. Recent studies have also distinguished between pocket and systemic infections, with preliminary evidence suggesting differing risk profiles. However, these subtypes remain underexplored in existing risk models.
Objective: (i) Validate the PADIT risk score in a contemporary cohort. (ii) Compare risk factors for pocket vs. systemic infections. (iii) Assess PADIT’s performance in predicting each infection subtype.
METHODS AND RESULTS:
Methods: Analyses used a prospective registry of all CIED implant procedures initiated January 2007 in collaboration with our hospital infection prevention team to prospectively identify all potential CIED infections and PADIT risk score components for each procedure. All potential CIED infections were independently adjudicated by two physicians and subclassified into pocket alone or systemic infection, with blinding for PADIT score and other pre-operative variables. Frequency statistics were generated to characterize the registry cohort. A logistic regression prognostic model was built to predict the outcome of adjudicated infection, with subgroup analyses for pocket vs. systemic infection. Predictive performance was evaluated using the Akaike information criterion (AIC), Bayesian information criterion (BIC), c-statistics, and calibration plots.
Results: Among 14,225 patients undergoing CIED procedures, 103 (0.7%) developed infections. These patients were younger (66.0 ± 15.0 vs. 72.2 ± 13.8 years, p< 0.01), more likely to receive ICD or CRT devices, and had ≥2 prior procedures (21.4% vs. 8.8%, p< 0.01). The PADIT risk score showed a moderate discrimination for overall CIED infection (c-statistic 0.690, AIC 1178.3) with excellent calibration (ratio 1.000). Subgroup analyses showed similar performance for pocket infections (c-statistic 0.691) and higher discrimination for systemic infections (c-statistic 0.730) (Table 1).
Conclusion: We independently confirm that the PADIT score demonstrates a reasonable predictive capacity in identifying the risk of CIED infection, and suggest it performs well across infection subtypes, with highest discrimination for systemic infections.
