The GLP-1 Exclusion Needs an Escape Hatch

The Readout · Issue 02

On July 1, 2026, The Cigna Group ended employee-plan coverage of weight-loss GLP-1s, including Wegovy and Zepbound. Type 2 diabetes coverage stayed. Reuters reported that employees were directed to cash-pay options, with that spending excluded from their deductible. [1]

This was an employee-plan decision, not a Cigna-wide change. Cigna's public Benefit Exclusion Overrides policies show a middle lane for plans adopting them. Adult criteria require a baseline BMI of at least 32, or at least 27 with two specified weight-related conditions, documentation, and diet and behavioral requirements. Continuation requires evidence of benefit. Separate pathways address cardiovascular risk, MASH, and obstructive sleep apnea. They do not establish Cigna employee coverage; the applicable plan document controls. [2]

The headline says no.

The machinery says maybe.

Open coverage is easy to explain and hard to budget. Peterson-KFF's analysis of 2023 survey data suggests roughly one-third of adults under 65 with employer-sponsored insurance could qualify through obesity alone. A small change in uptake can turn a specialty-drug line into the pharmacy budget. [3]

A blanket drug exclusion is easy to budget and hard to defend. It treats a relatively low-risk member seeking weight loss like one with cardiovascular disease, moderate-to-severe obstructive sleep apnea, or progressive liver disease. It may not stop treatment. It may move the prescription to cash, the cost to the employee, and the data outside the plan.

Self-funded employers have more than two buttons.

The clinical case is real. In separate trials of adults without diabetes, average weight loss was 14.9% with weekly semaglutide 2.4 mg versus 2.4% with placebo at 68 weeks, and 20.9% with weekly tirzepatide 15 mg versus 3.1% with placebo at 72 weeks, all with lifestyle intervention. Those are averages, not individual promises. [4]

FDA approvals now reach beyond weight: Zepbound for moderate-to-severe obstructive sleep apnea in adults with obesity; Wegovy for cardiovascular risk reduction in adults with established cardiovascular disease and overweight or obesity; and Wegovy injection, under accelerated approval, for adults with noncirrhotic MASH and moderate-to-advanced fibrosis, stages F2–F3. In the cardiovascular trial, major events occurred in 6.5% with Wegovy versus 8.0% with placebo over a median 41.8 months: a 20% relative reduction in the event hazard and a 1.5-percentage-point difference in observed event rates. [4][5]

Wegovy, Zepbound, and Foundayo carry boxed warnings about thyroid C-cell tumors based on animal or drug-class findings; the relevance to humans remains uncertain. Product-specific labels address certain thyroid-cancer histories and MEN 2, hypersensitivity, pancreatitis, gallbladder disease, dehydration-related kidney injury, hypoglycemia, severe gastrointestinal effects, delayed gastric emptying, and pregnancy. Individual risk belongs with the clinician, not the benefits committee. [6]

The cost problem is duration. In the STEP 1 extension, participants regained about two-thirds of their lost weight within a year after semaglutide and the trial's lifestyle intervention were stopped. In SURMOUNT-4, those who continued tirzepatide lost additional weight, while those switched to placebo regained substantial weight despite continued lifestyle counseling. [7]

A benefit designed as a one-year experiment is at war with the evidence.

Clinical value is not employer savings. A prevented heart attack matters beyond a plan-year spreadsheet. Employer return depends on net price, uptake, persistence, turnover, and when avoided events appear. No trial can answer that for one workforce.

If a vendor credits every possible future saving against today's pharmacy claim, take the calculator back.

Calling obstructive sleep apnea a workaround would be wrong. These FDA-approved indications have evidence behind them, not creative diagnosis coding. [5]

An exclusion of “drugs for weight loss” can coexist with coverage of the same drug and person under another indication. The molecule did not change. The covered problem and its evidence did.

Manufacturers are building a road beside the formulary. Novo and Lilly offer oral weight-loss GLP-1s through self-pay channels, with advertised starting-dose prices from $149 a month for eligible patients. Higher doses and program terms can change the price. Through Lilly Employer Connect, Zepbound KwikPen is available from Lilly to network pharmacies at $449 across doses. That is not the employer's final cost: pharmacy, administration, dispensing, and support arrangements affect the total. [8]

If the plan's net cost is higher than the manufacturer route, what exactly is the PBM adding?

Consider an illustrative self-funded plan with 1,000 enrolled employees and 1,800 covered members. No client data. If 10% qualify for the clinical override and 25% start, the plan has 45 users.

Assume all 45 users receive 12 months of treatment and support: net plan cost of $400 per user monthly after discounts and member cost sharing, plus $50 for clinical support.

45 users × 12 months × ($400 drug + $50 care) = $243,000 per year

Across 1,000 enrolled employees, that is $20.25 per employee per month.

At a $700 drug cost, the pathway costs $405,000, or $33.75 per employee per month. Keep the lower price but double uptake, and it becomes $486,000, or $40.50 per employee per month.

Care costs $27,000 a year. To earn the word savings, it must justify that cost through lower prices, less waste, better outcomes, or a guarantee. Better persistence may improve care while raising near-term pharmacy spending.

This is the decision rule:

eligible members × expected uptake × expected months × (net drug cost + care cost) ÷ enrolled employee months

The care fee is not the main lever. Eligibility, uptake, treatment duration, and net drug price are.

Build a narrower access lane without treating the drug as free money or cosmetic care.

Start with an indication matrix. Diabetes, obesity, cardiovascular risk, sleep apnea, and MASH need separate evidence, prescriber requirements, approval periods, and appeal paths.

Cigna calls its higher adult BMI threshold a product choice, not a clinical conclusion. A finite budget may justify prioritizing higher-risk members. Call that resource allocation, not medical truth. [2]

Place GLP-1 therapy inside an obesity-care benefit alongside nutrition, physical activity, other medications, and bariatric treatment when indicated. Support should begin with treatment, not become an obstacle course designed to make people quit. [6]

Continuation belongs in clinical review after titration: adherence, tolerability, safety, benefit, supply interruptions, and prescriber judgment. A response threshold must follow the evidence, allow exceptions, and remain a clinical rule. It cannot become an employer wellness contest.

Buy the pathway like stop-loss. Require net prices, rebates, fees, dispensing economics, data rights, direct-cash benchmarks, and guarantee math. Hold vendors financially accountable for promised savings, not only members.

Measure starts, persistence, discontinuation, adverse-event exits, net pharmacy cost, and medical use. Give business leaders an aggregate scorecard. Keep identifiable clinical information within care delivery and authorized plan administration, with safeguards and separation from employment decisions. [9]

The exclusion is easy. These questions test whether the policy is clinically credible and financially real.

1. What exactly is excluded?

Is the plan excluding a drug, the obesity indication, or every indication? What happens to people already in treatment?

2. What drives the budget?

How many members qualify at each threshold? What uptake and persistence are assumed? What is the net cost after rebates and fees, compared with direct channels?

3. Who owns the clinical and data risk?

Is support treatment, a voluntary service, or a coverage condition? What exceptions apply? Who receives identifiable data, and who audits the guarantee?

34%. Among firms with at least 200 workers that covered GLP-1s for weight loss in 2025, about one-third required a dietitian, case manager, therapist, or lifestyle program as a condition of coverage. A year earlier, it was 10%. [3]

The market is already testing the idea.

That does not mean the legal or clinical design is settled.

Cigna's employee decision is not a template. Its public override policies are the useful signal: covering everyone at any price is not a strategy. Neither is sending everyone to cash.

The better answer is a clinical pathway with a budget: defined indications, narrow access, clinician-led care, transparent net cost, member protections, and a vendor with something at risk.

Support the behavior. Do not turn HR into the behavior police.

Got a GLP-1 coverage design that looks like a blank check or a brick wall? Describe the decision through Submit a case.

Do not send PHI, diagnoses, claimant details, identifying information, or confidential employer files. Submissions are considered for possible analysis and publication. Nothing is published without permission.

The cases that teach something are worth breaking down in the open. The names and private details are not.

Know a benefits stakeholder facing this decision? Forward this issue. This argument gets expensive when nobody has it in the open.

See you in two weeks.

Scott

P.S. Next issue: contribution strategy. The premium increase is one number. How much the employer and employee each absorb is a separate decision.

General information, not insurance, medical, legal, or tax advice. Plan sponsors should obtain clinical and ERISA counsel review before changing coverage or conditioning access on participation.

The Explainable Broker · The strategy behind smarter benefit decisions · Unsubscribe anytime.