- Design
- Target-trial emulation in US registry and Medicare data, overlap weighted
- Population
- 3,200 effective patients with diabetes on maintenance haemodialysis
- Primary outcome
- Death on dialysis and all-cause hospitalisation
- Effect
- Mortality sHR 0.82 (0.70–0.97); hospitalisation sHR 0.91 (0.83–0.99)
Investigators used US Renal Data System and Medicare records from 2011 to 2022 to emulate a trial comparing semaglutide with dulaglutide or liraglutide in people with diabetes on maintenance haemodialysis. After overlap weighting the effective sample was 3,200, mean age 61, median BMI 34, followed for a median of about a year.
Semaglutide was associated with lower death on dialysis (sHR 0.82, 95% CI 0.70–0.97; one-year incidence 11% vs 13%) and fewer all-cause admissions (sHR 0.91, 0.83–0.99). Weight loss at a year was a little greater, about 4.8% against 3.4%. The advantage held against dulaglutide alone but was not statistically significant against liraglutide.
Dialysis patients were largely excluded from the GLP-1 outcome trials, so observational data like these are most of what exists. They do not prove semaglutide is better; residual confounding by who gets prescribed the newer agent is likely. Where a GLP-1 agent is already indicated in a dialysis patient, this is reasonable support for choosing semaglutide, with attention to nutrition and lean mass.
- In people on haemodialysis already suited to a GLP-1 agent, semaglutide is a reasonable first choice over dulaglutide on current evidence.
- Weigh weight loss carefully on dialysis, where protein-energy wasting is common; check albumin and dietetic input.
- Review gastrointestinal tolerance early, as nausea and poor intake can worsen interdialytic problems.
- Do not expect a difference against liraglutide; that comparison showed no significant benefit.
Why it matters
Dialysis patients were missing from the GLP-1 outcome trials, so the choice between agents has been made without data.
Don't overread it
This was observational; it cannot show that switching to semaglutide saves lives.
The statistics, in plain English
An sHR of 0.82 means about 18% lower rate of death on dialysis, accounting for transplant as a competing event. The interval 0.70–0.97 stays below 1, but the upper end is close to no effect. In absolute terms that was 11% against 13% at a year.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for diabetes & endocrinology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free