Operator learning curve for transradial percutaneous coronary interventions: implications for the initiation of a transradial access program in contemporary US practice

Abstract

Our study aimed to assess the characteristics and outcomes of transfemoral approach (TFA) versus the initial steps of a transradial approach (TRA) program and to assess the learning curve of TRA in contemporary, US practice. When compared to TFA, TRA has been shown to lower bleeding and vascular complications during percutaneous coronary intervention (PCI). However, use of TRA is still low. There are limited data regarding the characteristics of TRA learning curve, especially in an era with designated TRA equipment. Consecutive patients undergoing PCI in a single US center were divided into two cohorts according to vascular access approach: the last 250 TFA patients prior to the establishment of a TRA program and the initial 239 TRA patients following the establishment of a TRA program. Subgroup analysis of the TRA group, which was divided into five sequential case groups of 50 cases per group, was performed in order to assess TRA learning curve. Overall, the baseline characteristics of TFA vs. TRA groups were comparable. Fluoroscopy time was significantly longer during TRA procedures (18±11 vs. 15±8min, respectively, p=0.002); however, contrast use was lower during TRA procedures (161±72 vs. 180±63ml, respectively, p=0.002). In-hospital outcomes were similar between the two groups, with low frequencies of mortality, myocardial infarction, and stent thrombosis. Subanalysis of TRA group for learning curve assessment showed no major differences in patient demographics among the five subgroups. In the initial cases, more PCI was performed among non-acute cases (62% in patients 1-50 vs. 8-27% in patients 51-239, p<0.001). Despite these differences, characteristics of the treated vessels were similar between groups. There was no significant change in fluoroscopy time or in the amount of iodinated contrast volume delivered. Similarly, no differences in procedural, in-hospital, and long-term outcomes were documented. Adopting TRA as a default is feasible for high-volume operators without significant learning curve effects.