Most studies measure whether a drug works. This one measured that and, in the same seventy-eight people, how many would buy it — which is the number this site exists for and is almost never reported next to an effect size, any more than the real cost of a course appears next to a trial result.
The setup
174 adults with obesity entered a twelve-week program of diet, activity and education. [1] At week six, anybody who had lost less than 2.5% of their body weight was classed an early non-responder and offered liraglutide for the remaining six weeks.
Seventy-eight people met that definition. Twenty-one accepted. Fifty-seven declined, and the paper names the reason: out-of-pocket cost.
It worked
Between weeks six and twelve, the treated group lost 3.5 kg, or 3.2% of body weight. The group that declined lost 1.3 kg, or 1.4%. p < 0.001.
With liraglutide use, baseline body mass index, sex, age and week-six weight change all entered as variables, only the drug was significantly associated with what happened next. Tolerance was unremarkable: four people reported nausea, two upper abdominal pain.
What the 73% means
Fifty-seven of seventy-eight is nearly three quarters of the people for whom a clinician judged the drug appropriate, declining it on price, inside a study where somebody was explicitly offering it.
That is the demand curve nobody publishes. It sits next to what happened when coverage arrived for one indication and points the same way: this market is rationed by who pays, not by who benefits.
Two limits before anyone quotes the number
The drug is liraglutide — the oldest of this class, a daily injection, and weaker than what most people are choosing between now. A 3.2% figure over six weeks does not read across to a newer molecule.
And twenty-one people is a small treated arm. The direction is clear, the size is not firmly established, and the honest comparison for anyone shopping today is between the molecules actually on sale at what six months of each comes to.