Transvenous Lead Extraction in a Real-World Cohort: Clinical Characteristics, Outcomes, and Performance of the SAFeTY Score.

Background/Objectives: The increasing use of cardiovascular implantable electronic devices has increased the need for transvenous lead extraction (TLE), which carries a relevant complication risk. The SAFeTY score was developed to predict procedure-related major complications. We evaluated TLE outcomes, explored associations between the SAFeTY score and retrospective endpoints, and compared conventional extraction (CE) with mechanical rotational sheath-assisted extraction (MRSE). Methods: This retrospective study included consecutive patients undergoing TLE at a tertiary center between 2015 and 2025. The SAFeTY score was calculated in patients undergoing MRSE and examined in relation to procedure-related and all-cause in-hospital mortality, significant post-procedural hemoglobin decrease, and a composite endpoint. Results: Among 314 patients, 193 (61.5%) underwent MRSE. These patients had longer cumulative lead dwell time and more previous procedures, indicating greater procedural complexity. Overall procedural success was 95.9%, procedure-related mortality was 0.3%, and all-cause in-hospital mortality was 4.8%. Exploratory unadjusted comparisons showed no significant differences between MRSE and CE in procedural success or mortality. In the MRSE group, 17.1% had a post-procedural hemoglobin decrease > 30 g/L. The SAFeTY score was not associated with procedure-related or all-cause mortality but was associated with the composite endpoint (odds ratio 1.12 per point, p = 0.025), with modest discrimination (area under the receiver operating characteristic curve [AUC] 0.59). Conclusions: TLE achieved high procedural success and very low procedure-related mortality. Because MRSE was non-randomly selected for more complex procedures, the study cannot establish equivalence or an independent effect of extraction strategy. Associations between the SAFeTY score and alternative retrospective endpoints should be considered exploratory and do not constitute validation for its original purpose.