Predictors of short-term mortality in patients undergoing a successful uncomplicated extraction procedure.

BACKGROUND: The prognosis of patients with untreated cardiac implantable electronic device (CIED) infection is poor. Whether removal of all leads by a successful transvenous lead extraction (TLE) procedure changes the prognosis is unclear. OBJECTIVE: To identify predictors of mortality in patients with CIED infection despite successful TLE. METHODS: Retrospective single-center analysis of prospectively collected database from consecutive patients undergoing TLE at our center. Predictors for mortality were identified and a score predicting high mortality rate was calculated. RESULTS: A total of 371 consecutive patients underwent TLE, of whom 337 (90.8%) had complete hardware removal. Most were extracted due to infectious causes (81.3%). Approximately one-third (35%) died during a mean follow-up of 1056 ± 868 days. There was significantly higher mortality observed in the infectious group. Multivariate logistic regression models for infectious group only identified creatinine and albumin measurements as risk markers for 30 days mortality (odds ratio [OR], 1.68; 95% confidence interval [CI], 1.19-2.38; P = .003 and OR, 0.4; 95% CI, 0.16-0.97; P = .039, respectively). A risk score was created based on cutoff values of creatinine ≥2md/dL (1 point) and albumin ≤3.5 g/dL (1 point). A value of 2 points predicted a 50% chance of 30-day mortality and a 75% chance of 1-year mortality (P < .0001 for both). CONCLUSIONS: Creatinine and albumin can be used as a combined risk score to successfully identify patients at risk of death despite undergoing a successful TLE procedure for infectious reasons. This score could help decision making when contemplating on conservative antibiotic treatment vs TLE.