The diagnostic yield of implantable loop recorders stratified by indication: "real-world" use in a large academic hospital.

Abstract

To report on the clinical utility of implantable loop recorders (ILRs) in a large academic hospital setting over a 4-year period. Retrospective study (2013-2016) of patients receiving ILRs for any indication including syncope, cryptogenic stroke (CrS), atrial fibrillation (AF) burden, palpitations, ventricular arrhythmias (VA), and other. Remote checks, symptomatic transmissions, and in-person checks were reviewed. Time to diagnosis was documented. A total of 263 patients (54% male, mean age 63 ± 15years, mean follow-up 601 (range 9-1714) days) received ILRs for 324 indications; multiple indications were noted in 53/263 (20.2%) patients. ILR indications were 126 (39%) syncope, 81 (25%) CrS, 46 (14%) AF, 37 (11%) palpitations, 10 (3%) VA, and 24 (7%) other. Diagnostic yield for each indication was compared to the overall yield for all other indications. Three indications showed a significantly higher yield: AF (65% vs. 22%, p<0.002), palpitations (60% vs. 24%, p<0.001), and VA (70% vs. 28%, p<0.004). For all other indications, there were no significant differences. Syncope had nearly half the diagnostic yield of previously published trials (28% vs. 43-56%). We observed a fourfold increase in ILR implant rate over the study duration. In a "real-world" academic hospital setting, the diagnostic rate of ILRs was highest for AF, palpitations, and VA; however, these high yield indications comprised only 29% of all indications. The diagnostic yield for the commonest indication (syncope) was approximately half that reported in the previously published trials. With increasing implantation rates, additional studies are required to refine guideline-based indications for ILR implantation to improve diagnostic yield.