Skip to content
Buy This GLP
← Research
Evidence

Does a GLP-1 Save Money on Other Healthcare? Yes, Somebody Else's

Healthcare costs excluding the drug fell $319 a month on tirzepatide and 22% over a year on semaglutide. Every one of those savings accrues to an insurer, not to you.

Carla Medina9 min read
Monthly healthcare costs avoided, per person6–12 months−$14512–18 months−$319These exclude what the drug itself costs

Some of it, to somebody else. Measured in claims, healthcare costs excluding the drug fell by $319 per person per month at 12 to 18 months on tirzepatide. A separate analysis of branded semaglutide found medical costs 22% lower over a year, or $4,639 per patient [2]. Both figures leave out what the drug costs, and neither lands in the pocket of a cash buyer.

The one study that counts both sides is the comparison against surgery. Across 6,748 insured patients, three-year out-of-pocket cost was highest for semaglutide at $7,752, against $5,980 for sleeve gastrectomy[3]. Total three-year spending did not differ significantly between the groups, so the drug shifts who pays more than it shifts what is spent.

Cost-offset analyses are how expensive drugs get justified to payers, and they are usually modeled. This one is measured — real claims, real people, over three years — which makes it worth reading carefully rather than dismissing, and unlike the lifetime simulations it does not project.

What was measured

Adults over 55 with obesity, or overweight plus a weight-related condition, and without type 2 diabetes, who started tirzepatide for weight management between November 2022 and September 2025. [1] They were propensity-matched against people who took no incretin at all, leaving 15,843 in each group, and all-cause healthcare costs were compared using a difference-in-differences framework at three intervals.

At 6 to 12 months, costs in the treated group were $145 per person per month lower, 95% CI −266 to −25, a 12.3% reduction. At 12 to 18 months, $319 lower, 95% CI −544 to −94, a 25.4% reduction. Using a stricter censoring method the figures were $181 and $607. Fewer inpatient and emergency department visits accounted for part of it, with an incidence rate ratio of 0.69, 95% CI 0.50 to 0.94.

The comparison is with people who took nothing

And the conclusion describes persistent users. That is the authors’ word. People who keep taking an expensive drug for eighteen months differ from people who never started one, in insurance, engagement and health. Every claims analysis this desk has read runs into the same wall. Adherence selection inflates whatever the treated group is measured on.

One of the components is also not significant. The medical-cost driver at 12 to 18 months comes with p = 0.065, and the abstract presents it as a driver anyway. The emergency and inpatient reduction is significant; that component is not.

Who wrote it

The author affiliations include Eli Lilly and Company, which manufactures tirzepatide, along with two contract research organizations. The indexed record carries no separate conflict-of-interest statement, but the affiliations are printed and checkable, which is more informative than most disclosures.

That does not make the claims data wrong. Claims are claims. It does mean the analytic choices — which intervals, which censoring method, which comparison group — were made by people with an interest in the answer. The paper also ends by relating its findings to a specific Medicare net cost figure for a named program. That is a policy argument as well as a result.

What it means for someone paying cash

Less than it looks. These offsets accrue to an insurer, not to a patient — fewer emergency visits mean fewer claims paid, and somebody buying from this roster has no insurer absorbing those costs in the first place.

The transferable finding is the shape: the healthcare savings are real, they grow with time on treatment, and they are smaller than the drug. That is the honest version of the cost-effectiveness case, and it is a different statement from the one a headline makes. For a cash buyer the arithmetic is unchanged. It is what six months costs against how many people have to be treated for one event to be avoided, and neither of those improves because an insurer spent less.

Where an offset is large, specific and still not yours

The sharpest offset in the literature is surgical. A claims analysis matched 425 semaglutide users against 2,514 controls having anterior lumbar interbody fusion. It found roughly $10,400 lower same-day surgical costs and $9,700 lower 90-day costs, on 0.6 fewer days in hospital and no significant difference in complications [4]. Half a day of hospital stay does not account for the same-day figure, which is worth reading as a finding still looking for its mechanism.

Every number on this page accrues to an insurer, a hospital or a health system. For a reader paying cash, the offset is real and belongs to somebody else, while total spending with the drug included goes up. That asymmetry is the whole point of putting these studies on a price desk rather than in a payer’s deck.

Frequently asked

Does tirzepatide pay for itself?
Not on this analysis. The $319 monthly figure at 12 to 18 months excludes the drug's own cost, so it is money back against a larger bill rather than a net saving.
Where do the savings come from?
Partly from fewer inpatient and emergency department visits, with an incidence rate ratio of 0.69. The medical-cost component at 12 to 18 months carries a p-value of 0.065 and is not significant.
Would I see any of this money?
No. These are healthcare costs paid by an insurer. Someone buying from this roster has no insurer absorbing those costs to begin with.
Who conducted the analysis?
Author affiliations include Eli Lilly and Company, which manufactures tirzepatide, and two contract research organizations. The indexed record carries no separate conflict-of-interest statement.

Sources

  1. [1] Upadhyay N, et al. (2026). Trends in Cost of Care With Tirzepatide in Adults Aged Over 55 Years With Obesity or Overweight Without Diabetes: A Matched Cohort Analysis Diabetes, Obesity and Metabolism. PMID 42638166
  2. [2] Michalak W, et al. (2025). Impact of semaglutide 2.4 mg on healthcare resource utilization and medical costs in patients with atherosclerotic cardiovascular disease in the United States (SHINE-ASCVD) Journal of Medical Economics. PMID 40579810
  3. [3] Chhabra KR, Gencerliler N, Orandi BJ, Wang VH, et al. (2026). Semaglutide vs. Bariatric Surgery: Comparing Costs and Clinical Outcomes in Patients With Diabetes and Obesity Obesity (Silver Spring, Md.). PMID 42089543
  4. [4] Ng MK, Mastrokostas PG, Tabbaa A, Razi A, Johnson M, Said M (2026). Semaglutide Use Is Associated With Decreased Length of Stay and Hospital Costs in Patients Undergoing Anterior Lumbar Interbody Fusion: A Retrospective Cohort Study Orthopedics. PMID 41636429

Where to get it

Price the injectable sellers

The desk lists every seller that publishes an injectable figure, with the advertised price struck against the one a buyer is billed.

Open the price desk

More in Evidence