This desk has compared these two paths on cost, in the drug costs you more and the system the same, and on weight in procedure beat tablet at six months. This study compares them on cardiovascular outcomes, and the result is not close[1].
Among people without type 2 diabetes, the composite of coronary events, cerebrovascular events, heart failure and cardiac arrest occurred in 4.4% of surgical patients against 6.6% of those on semaglutide, a hazard ratio of 0.402. Among people with diabetes the rates were 9.0% against 14.6%, a hazard ratio of 0.421. Heart failure showed the largest relative differences, at 0.293 and 0.346.
What makes this worth publishing is that it closes a hole in an earlier comparison. Previous work matching surgery against “GLP-1 drugs” was weakened because the drug arm was largely older, weaker agents. Here the drug arm is semaglutide specifically, and surgery still wins by roughly the same margin.
A second asymmetry is specific to how the two treatments work. An operation happens once and cannot be skipped afterwards; a prescription requires filling every month, and roughly two-thirds of people stop within a year, as two thirds stop within a year sets out. Refill requirements appeared only in sensitivity analyses here, so a substantial part of what is being measured may be that the surgical intervention was actually received and the drug one often was not.
Read the absolute numbers before deciding what a hazard ratio of 0.402 means. Without diabetes it represents 2.2 fewer events per hundred people over five years; with diabetes, 5.6 fewer. Those are substantial figures and they are considerably less dramatic than “60% lower” sounds — the arithmetic set out in the falls and fractures cohort. And none of this touches the question of what each path costs, which is where most readers of this desk actually decide, covered in how long the drug takes to cost more.