Predicting the difficulty of a transvenous lead extraction procedure: Validation of the LED index

A lead extraction difficulty (LED) score was proposed to predict the difficult transvenous lead extraction (TLE) procedures, defined by means of the fluoroscopy time. The aim of this study was to validate the estimation model based on the LED index above 10 on an independent data set of TLE cases.

Consecutive patients undergoing TLE between January 2014 and January 2016 were included in this analysis. The fluoroscopy time related to the leads removal was dichotomized as above or below its 90th percentile (PCTL).

In total, 446 permanent leads were removed during 233 TLE procedures. Complete procedural success was achieved in 232 (99.1%) patients. The LED index resulted >10 in 83(35.6%) procedures. Among these cases, 20 had fluoroscopy time above the 90th PCTL (23.3 minutes) and were classified as true positive. Over the 150 procedures with LED score <10, 147 were classified as true negative, and only 3 resulted false negative. The sensitivity, the specificity, and the negative predictive value of the LED index in predicting complex cases resulted 86.9 (confidence interval [CI] 66.4-97.2)%, 70.0 (CI 63.3-76.1)%, and 98.0 (CI 94.3-99.6)%, respectively. The multivariate logistic regression analysis confirmed a 12% increased risk of high fluoroscopy for each additional point of the LED score (OR 1.12, CI 1.05-1.21, P = 0.001).

The validation of the estimation model based on the LED index <10 confirmed its high efficacy in predicting simple TLE procedures.