Thinking About Dropping GLP-1 Coverage?

The Measurement Question Most Plans Skip

If you run benefits for a self-funded company, GLP-1 drugs are probably the single loudest line in your pharmacy report right now. Ozempic, Wegovy, Zepbound, and the rest have gone from a rounding error to the top of the spend chart in about three years. You are not imagining the pressure. Across employers, 51% now name GLP-1s their number one prescription cost driver, ahead of oncology. When a category climbs that fast, the instinct to cut is completely reasonable.

So the question shows up at renewal, usually from finance. Should we just drop GLP-1 coverage?

I want to make the case that dropping is a real option, but it's the blunt one. And the honest, harder question underneath it is the one most plans skip. Not "should we drop it," but "if we drop it, do we actually save money, and can we prove it?"

Dropping is a lever. It just isn't a free one.

Cutting coverage does one thing cleanly. It removes the drug cost from your plan. On next year's budget, that shows up as a number going down, which is exactly what everyone in the renewal meeting wants to see.

The trouble is that the drug cost was never the whole story. GLP-1s are expensive, but a lot of them are being prescribed for conditions that are also expensive when they go unmanaged. Type 2 diabetes. Cardiovascular risk. Obesity-related conditions that drive claims across the rest of your plan. When you drop the drug, you don't drop those conditions. You just stop paying for one of the tools aimed at them, and you keep paying for everything downstream.

That's the part the budget line doesn't show. The savings are visible and immediate. The costs you keep are diffuse, delayed, and scattered across medical claims that never get traced back to the decision.

The number that actually matters is net, not list

Here's where most of the conversation goes wrong. People argue about the list price of these drugs, which is genuinely high. But your plan doesn't pay list.

Net of rebates, the real employer cost of a GLP-1 lands closer to $569 to $664 per treated member per month (ICER and Blue Health Intelligence, net of rebates). That's still a serious number. It is not the four-figure sticker price the headlines use, and the gap matters, because you're about to weigh it against the costs you keep if you drop.

So the first honest step is boring and important. Know your net GLP-1 cost, after rebates, not the list price on the PBM report. If you're making a drop-or-keep call on the sticker number, you're deciding against a cost you don't actually carry.

The comorbidity question you can't skip

Once you have the net cost, the real comparison begins. Dropping saves you that net drug spend. What does it cost you on the other side?

This is where I have to be honest about what's known and what isn't. The clinical research on whether GLP-1s reduce total medical cost is still maturing, and the answer clearly depends on who's taking them and why. For some populations the offset is meaningful. For others it takes years to show up, longer than the average employee stays on your plan. Anyone who tells you the drugs always pay for themselves is selling something, and so is anyone who tells you they never do.

The point isn't that dropping is wrong. The point is that dropping without measuring the other side of the ledger isn't a savings decision. It's a guess dressed as one.

Can you even measure it? Usually, not yet.

Here's the question I'd actually sit with. If you dropped GLP-1 coverage this year, would you be able to tell, twelve months from now, whether it saved you money net of everything?

For most plans, the honest answer is no. To measure it, you'd need to watch what happens to the medical claims of the people who lost coverage. Did diabetes control slip? Did related admissions tick up? Did they move to a cheaper drug, or stop treatment entirely? That's a claims-analysis question, and it requires reading your own data across pharmacy and medical, over time, on purpose.

Most teams don't have that set up. So the decision gets made, the drug line goes down, and nobody ever closes the loop on whether the total actually fell. The plan gets credit for savings it never verified.

The better version: a measured decision you can defend

None of this is an argument to keep GLP-1 coverage no matter what. Plenty of plans have good reasons to restrict or reshape it. It's an argument to decide it as something you can defend, instead of a reflex you'll be asked about later.

There's usually more than two options, too. Between "cover everything" and "cover nothing" sit the levers a self-funded plan controls. Indication carve-outs that cover the drugs for diabetes while setting different terms for weight management. Prior authorization tied to real clinical criteria. Formulary tier placement. Step therapy where it fits. These let you manage the cost without taking the blunt, all-or-nothing swing, and they're easier to defend to both your CFO and your people.

Whatever you choose, the discipline is the same. Know your net cost, not the list price. Name the downstream costs you'd keep if you dropped. Decide the lever. Then set the one or two numbers you'll watch, net pharmacy spend and the relevant medical trend, so a year from now you can actually say whether the call worked.

That last step is the one that turns a decision into an accountable one. Not just what you decided about GLP-1s. Whether the thing you decided actually saved what you thought it would.

If you're heading into a renewal with this question on the table and no clean way to measure either side of it, that's a fair problem to raise. The window to set up the measurement is open before you make the call, not after.