AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
Cardiovascular Outcomes and the Role of Treatment Persistence in a Nationwide SELECT-Like Population Initiating Semaglutide
By Myhre PL, Brant SB, Rundblad A et al.
Background: Treatment with semaglutide reduced the incidence of major adverse cardiovascular (CV) events (MACE) compared with placebo in the SELECT trial. In routine clinical practice, however, suboptimal treatment persistence may limit the extent to which trial findings translate to real-world outcomes. Research Questions: Are CV event rates in real-world SELECT-like semaglutide initiators comparable to those observed in the semaglutide arm of SELECT? Does greater treatment persistence indicate improved CV outcomes among real-world SELECT-like semaglutide initiators? Methods: We conducted a nationwide observational cohort study of all adults filling a semaglutide prescription in Norway between 2021 and 2025, using linked individual-level data from the Prescription Drug Registry, the Patient Registry, and the Cause of Death Registry. New users of semaglutide fulfilling SELECT-like eligibility criteria—age ≥45 years, established atherosclerotic cardiovascular disease, and no type 2 diabetes—were included. Outcomes were described for all SELECT-like patients, for varying definitions of persistence (<30, <60, and <90 days between end of 28-day supply period of one prescription to the next prescription filling), and were indirectly compared with the semaglutide arm of the SELECT trial. The primary outcome was MACE-3, defined as myocardial infarction, stroke, or CV death. Secondary outcomes included heart failure hospitalization or CV death, and total hospitalizations. Time-to-event was described using cumulative incidence curves. Results: Of the 198,658 individuals initiating semaglutide during the study period, 5,376 met SELECT-like eligibility criteria. Their age distribution (mean (SD) 62 (9) years) was comparable with the SELECT trial (62 (9) years), but there were fewer men (54 %) compared with SELECT (72 %, Panel A). The real-world estimated 12-month cumulative incidence of MACE-3 was numerically the same as the semaglutide arm of SELECT (1.7 %, Panel B). A strict definition of persistence (<30 days) showed numerically lower 12-month incidence rate of heart failure hospitalization or CV death than for all SELECT-like individuals (0.4 % vs 0.6 %, Panel C). The 12-month incidence rate for total hospitalizations were 14.0 % for SELECT-like while it was 14.6 % in the semaglutide arm in SELECT (Panel D). Conclusions: In real-world patients initiating semaglutide and classified as SELECT-like, CV event rates were comparable with those observed in the SELECT trial, suggesting generalizability to routine clinical practice. While persistence may be challenging in a real-world out-of-pocket setting, higher persistence indicated more favourable CV outcomes, suggesting a potential importance of sustained treatment in this population.