Medicare GLP-1 Bridge Launches July 1: What Part D Plan Actuaries Need to Know

Part D plan actuaries carry zero GLP-1 drug cost risk through December 2027 because the Medicare GLP-1 Bridge opening July 1 runs entirely outside Part D, with a $245 negotiated monthly net price and a Humana-operated central processor no plan touches (CMS, June 2026). The harder work is building 2028 bids around the still-unpublished BALANCE model.

The Bridge pairs that $245 net price with a $50 beneficiary copay that sits outside both the deductible and the true out-of-pocket cap, routed through enhanced LI NET infrastructure. That design simplifies the 2027 bid considerably, but the 2028 problem starts immediately: the BALANCE model’s pricing, utilization, and sponsor-obligation parameters remain unpublished, on a bid calendar that will require those assumptions before the parameters arrive. CMS published full operational guidance in June, closing the informational gap that existed when earlier analyses addressed the $50 copay announcement and early structural unknowns in May. The picture is cleaner in one dimension and more uncertain in another, and the place to start is who actually qualifies.

$50/mo

Fixed Beneficiary Copay (Outside Deductible and TrOOP)

$245

Negotiated 30-Day Manufacturer Net Price (WAC Difference Goes to CMS)

Dec 2027

Bridge Program End Date, When BALANCE or Legislative Action Must Follow

How the Bridge Pays, and Why No Plan Is in the Chain

The Bridge operates under Section 402 demonstration authority, which lets CMS build a payment channel parallel to, rather than within, the standard Part D benefit. Humana serves as the central processor on an enhanced version of the LI NET (Low-Income Net) infrastructure that already handles Part D gap coverage for low-income beneficiaries awaiting enrollment. The choice was deliberate: LI NET connects to essentially every Medicare network pharmacy and routes claims without plan-level integration.

The payment chain has four steps. A beneficiary with a qualifying prescription presents at any network pharmacy; the pharmacy collects $50; the claim routes to the Humana processor, which reimburses the pharmacy at no less than the drug’s wholesale acquisition cost minus the $50 copay plus a standard dispensing fee; the manufacturer then remits to CMS the difference between WAC and the $245 negotiated net price, with the federal government absorbing the net program cost. No Part D plan appears anywhere in that flow, which is why CMS’s June guidance confirms that sponsors do not opt in and do not count Bridge drugs toward Part D financial reporting, gross covered prescription drug costs, direct subsidy calculations, or risk corridor exposure. Within the Part D framework, Bridge claims effectively do not exist.

Three drug families qualify. Wegovy (semaglutide), made by Novo Nordisk and covering the injectable and newer tablet forms, carried a pre-deal list price around $1,350 per month and prices at $245 under the Bridge. Zepbound (tirzepatide), made by Eli Lilly and restricted to the KwikPen formulation, carried a pre-deal list price above $1,080 per month and also prices at $245. Foundayo (orforglipron), Eli Lilly’s oral GLP-1 receptor agonist approved by the FDA on April 1, 2026, enters with self-pay pricing from $149 per month for the lowest dose. Bridge pricing for Foundayo has not been specified under the same $245 framework as the injectables, and that ambiguity matters for utilization projections, because oral formulations have shown historically higher adherence. The program runs July 1, 2026 through December 31, 2027, an 18-month window with no announced extension mechanism and no standing legislative authority to continue without further administrative action.

Eligibility Is Narrower Than the Headline BMI Threshold

The Bridge’s criteria are meaningfully tighter than the BMI 27+ figure that drove early estimates. Roughly 14 million Medicare beneficiaries carried overweight or obesity diagnoses in 2020 claims data (KFF), and that number has circulated as a proxy for the eligible pool. It is the wrong denominator, because eligibility is gated on comorbidity, not BMI alone. The first tier covers a BMI of 30 or higher plus one of three conditions: heart failure with preserved ejection fraction, uncontrolled hypertension, or chronic kidney disease at stage 3a or above. The second tier covers a BMI of 27 or higher plus one of four cardiovascular or metabolic diagnoses: pre-diabetes, prior myocardial infarction, prior stroke, or symptomatic peripheral artery disease. A beneficiary must clear both a BMI threshold and a qualifying diagnosis.

Applying those screens to the 14 million top-line narrows the clinically eligible pool substantially. Drawing on published Medicare claims prevalence for cardiovascular disease, CKD, and pre-diabetes in the BMI 27+ population, a first-order estimate puts the eligible pool at 6 to 9 million beneficiaries. The range carries real uncertainty: claims-based prevalence understates undiagnosed conditions, particularly pre-diabetes, and qualifying diagnoses are documented at different rates across geography, specialty access, and socioeconomic status. Plans should model the eligible pool as a distribution, not a point estimate.

The Comorbidity Gate Concentrates the Highest-Cost Members

The eligibility architecture produces a risk-selection pattern that is actuarially significant for MA-PD sponsors. The conditions that qualify a beneficiary are also the ones that place Medicare beneficiaries in the highest-cost tiers under HCC coding. A beneficiary qualifying under the tier-one heart failure or CKD pathway typically carries a risk score above 1.8 and annual total cost of care well above $20,000. The Bridge does not neutralize that cost profile; it only lifts the GLP-1 drug cost out of the plan’s exposure, while medical costs, hospitalizations, and Part A utilization for these members stay fully on the plan’s books.

That splits the actuarial picture in two. Near term, the federal processor absorbs GLP-1 obesity drug costs for the plan’s sickest members. Medium term, if GLP-1 therapy delivers the documented cardiovascular benefit, a 20% relative reduction in major adverse cardiovascular events for semaglutide versus placebo in a BMI 27+ population with prior cardiovascular disease (SELECT trial), the Bridge could produce downstream medical cost offsets through lower utilization. Those offsets only land after the 18-month window closes, and only for members who stay in the same plan and maintain adherence, timing that standard Part D bid frameworks capture poorly because they do not model cross-bucket interactions between Part D pharmacy and Part A/B medical costs within a single planning cycle.

Formulary and Appeal Obligations Survive the Bridge

While Bridge claims sit outside Part D, the June guidance is explicit that the carve-out does not let plans contract their existing GLP-1 obligations. Standard formulary coverage for non-weight-loss indications must continue uninterrupted: Wegovy and Zepbound for weight management and cardiovascular risk reduction; Ozempic, Victoza, and Rybelsus for type 2 diabetes; Trulicity in wide diabetes use. Plans may not cut formulary access to GLP-1 receptor agonists approved for diabetes or cardiovascular indications because the obesity indications now run through the Bridge, and prior authorization, step therapy, and quantity limits must track standard clinical and federal formulary criteria. Beneficiary appeal rights are likewise preserved: a dispute over a coverage determination for a GLP-1 prescribed for a formulary-covered indication still runs through the standard Part D appeals process, not a Bridge channel. AMCP’s March 2026 regulatory guidance on the CMS FAQ confirmed that plans may communicate the Bridge to beneficiaries as an obesity-access pathway but are not required to, and cannot be penalized for not doing so.

The TrOOP and LIS Exclusion Reshapes Copay Accounting

Excluding the $50 copay from true out-of-pocket spending sends non-obvious effects through catastrophic-phase projections and low-income subsidy calculations. Under the IRA-redesigned Part D benefit, beneficiary out-of-pocket spending accumulates toward the $2,100 TrOOP cap (2026 figure); crossing it triggers the catastrophic phase, where cost sharing falls to zero and risk sits heavily on plan sponsors after the redesign cut the federal reinsurance subsidy. Year-one redesign data showed catastrophic phase utilization running roughly 22% above actuarial base cases (CMS Part D financial data, 2026), driven partly by beneficiaries hitting the TrOOP cap faster than models predicted. Because Bridge copays do not count toward TrOOP, a bridged beneficiary accrues TrOOP only on formulary drugs. For the subset whose other Part D spend would have pushed them past $2,100, that exclusion keeps them in the initial coverage or coverage gap phase longer, trimming plan liability at the catastrophic threshold.

The reduction in plan-level catastrophic exposure is directional but member-specific in magnitude, so actuaries should isolate the Bridge-eligible members previously forecast to reach the TrOOP cap, exclude their GLP-1 spend from TrOOP accumulation, and recompute the catastrophic-phase enrollment rate. In aggregate the effect is likely modest, since Bridge-eligible members are high-cost patients who often carry enough other specialty pharmacy use to reach the cap regardless. The LIS interaction runs the other direction: low-income subsidy beneficiaries cannot apply LIS cost-sharing reductions to the $50 Bridge copay, so for the dual-eligible population, where LIS typically eliminates cost sharing entirely, the copay is a net new expense the subsidy does not cover. Actuaries building D-SNP and dual-eligible bids should flag that the $50 copay may become an access barrier for the lowest-income, highest-comorbidity members, suppressing uptake in the segment most likely to benefit clinically.

BALANCE: Paused, Not Cancelled, and Silent on the Parameters That Matter

CMS introduced the BALANCE (Beneficiary Access to Lifesaving Anti-obesity Necessary Coverage Enhancement) Model through the Innovation Center in December 2025 to pull Part D sponsors into the GLP-1 obesity coverage chain the Bridge deliberately avoids. CMS issued the Request for Applications in March 2026, set an April 20 application deadline, and announced a pause on April 21 after the program failed to reach the 80% NAMBA-weighted enrollment threshold specified in Section 2.3.1 of the RFA. The Bridge extension through December 31, 2027 followed as the immediate substitute. BALANCE is paused, not cancelled, which is why its March RFA still governs 2028 planning.

The RFA gives a partial map. Under it, Enhanced Alternative plans and employer group waiver plans (EGWPs) would pay a $50 copay per covered GLP-1 fill; standard plans would pay $125. Catastrophic-phase cost sharing would be eliminated for BALANCE-covered drugs, and plans would have to place all qualifying model drugs on formulary without tier discrimination. Risk corridor protection would apply at 2.5% thresholds rather than the standard 5%, and only for plans exceeding one standard deviation above mean national utilization. What the RFA left open is precisely what 2028 bids hinge on: the manufacturer pricing structure (whether the $245 Bridge net price carries over); the utilization assumptions CMS would use to set the benchmark rate; the eligibility criteria, which may differ from the Bridge’s comorbidity gate; the formulary scope, including whether oral agents like Foundayo are covered separately; and whether LIS beneficiaries keep cost sharing or have it eliminated.

Plans filing 2028 Part D bids in mid-2027 will do so before CMS is likely to finalize any of that. BALANCE therefore has to be modeled as a scenario variable, not a known input, across at least three states: it launches on terms close to the March 2026 RFA; it launches with modified terms that expand sponsor risk; or it is abandoned and the Bridge extends through 2028, preserving the current zero-plan-risk structure. Each state drives materially different catastrophic-phase projections, formulary utilization assumptions, and net plan liability for GLP-1s as a class.

Three Cost Projections Bracket the Utilization Range

Three published projections set the range plan actuaries should bracket in sensitivity testing, and the spread between them is mostly about uptake and adherence, not price. The Congressional Budget Office published its estimate before the Bridge existed: assuming standard Part D cost sharing, it projected 12.5 million newly eligible beneficiaries, modeled only 0.3 million (2%) actually using anti-obesity medications in year one, and put the ten-year net federal cost near $35 billion from 2026 through 2034 (CBO, October 2024), with per-user cost starting around $5,600 in 2026 and declining to $4,300 by 2034. Because CBO’s assumptions predate both the $245 price and the $50 copay, both far below the cost sharing it modeled, actual Bridge utilization should run above its base case.

The JAMA Health Forum microsimulation modeled broader coverage and produced a ten-year net Medicare cost of $47.7 billion (JAMA Health Forum / Hwang et al., April 2025), against $65.9 billion in gross drug spending and $18.2 billion in medical cost offsets, a 27.6% offset ratio. It assumed 10% uptake with 40% long-term adherence and an annual net semaglutide cost near $8,412 per beneficiary. The $245 monthly Bridge price annualizes to $2,940 net before the $600 annual copay offset, roughly 65% below that drug cost assumption, so lower drug costs and lower copays both point toward higher utilization than the simulation projected. The $47.7 billion figure is the widely cited baseline in congressional discussions of legislative options (STAT News, April 2026). A $1,743 annual federal cost per user follows at the $245 price after the $50 monthly copay offset (American Action Forum, November 2025); its mid-range scenario of 2.2 million users implies roughly $3.8 billion in annual steady-state cost, with the partial-year fiscal 2026 figure near half that given the July 1 launch.

Source	Eligible Population	Projected Year-1 Users	10-Year Net Federal Cost	Drug Cost Assumption
CBO (October 2024)	12.5 million	0.3 million (2%)	~$35 billion	Standard Part D cost sharing
JAMA/Hwang et al. (April 2025)	30 million cumulative	3 million (base case)	$47.7 billion net	~$8,412/year semaglutide
American Action Forum (November 2025)	~54.8 million Part D enrollees	2.2 million (mid-range)	~$3.8 billion/year at steady state	$245/month manufacturer net price

The $50 copay is a 96% reduction from Wegovy’s pre-Bridge list price and an 80% reduction from the $245 net price manufacturers are absorbing, and roughly half of current GLP-1 users describe the drugs as difficult to afford (KFF survey). Removing that affordability barrier for a Medicare population with higher obesity prevalence and lower average income than the commercially insured creates a demand environment no pre-Bridge model fully captures, so the gap between the CBO and JAMA figures is best read as a range to weight, not a single number to adopt.

Building 2028 Bid Assumptions Before the Parameters Arrive

The core modeling problem for 2028 bids is a timing mismatch, not a research gap: the BALANCE scenarios require materially different formulary assumptions, drug cost trend factors, and catastrophic-phase enrollment projections, yet plans must file roughly six months before CMS is expected to finalize BALANCE, and no amount of research closes a gap that sits between the Bridge end date, the BALANCE launch timeline, and the bid calendar. The workable response is to treat GLP-1 obesity coverage as a scenario-weighted expected-value problem, assigning explicit probability weights to the three states and sensitivity-testing the parameters that move the answer: eligible population size, uptake rate, adherence persistence, drug cost per member per month, and medical cost offset timing. The scenario structure and weights belong in the supporting memorandum, not buried inside an aggregate trend factor.

Each scenario carries a distinct build. The March 2026 RFA is detailed enough to stand up a preliminary model on its terms; the open question is what fraction of the current Bridge-eligible population transitions into the plan’s formulary risk at launch, and at what utilization and adherence profile. Plans with large Bridge-eligible blocks should assume a meaningful share continues GLP-1 therapy through BALANCE, producing an identifiable step-change in formulary GLP-1 costs at the 2028 transition. Under the expanded-risk scenario, CMS recalibrates eligibility broader than the Bridge’s comorbidity gate, potentially to BMI 27+ without the specific cardiovascular or renal qualifications. That enlarges both the eligible population and plan-level drug cost exposure, because the lower-acuity members added under broader criteria have not been pre-screened by the high-HCC filter that defines today’s Bridge pool. Under the extension scenario, 2028 reads as a continuation of 2027: zero plan-level GLP-1 obesity cost, the same comorbidity gate, the same $50 copay, the same TrOOP and LIS exclusions. It is the simplest to model but not the most likely, since federal budget pressure and the politics of GLP-1 access make a perpetual zero-plan-risk Bridge an improbable permanent design.

The 2027 MA bid environment already reflects compressed margins across the NAMBA-weighted enrollment base, so plans that leaned on aggressive utilization assumptions in their 2027 filings have little pricing room to absorb BALANCE-driven cost increases in 2028 without rate increases that could accelerate the disenrollment trend that pushed roughly 3 million beneficiaries to switch plans in 2026. The 18-month window will not generate adherence, offset, or uptake data credible enough for 2028 bids, and the analytical infrastructure that absorbs the BALANCE transition, member-level utilization tracking, eligible-population monitoring, adherence modeling, HCC interaction analysis, and catastrophic-phase projections across the three scenarios, is best built before the parameters that will populate it are finalized. Plans that wait will price under the same uncertainty that sank the original BALANCE model’s 80% enrollment threshold in April 2026. The Bridge is also generating the first real-world Medicare GLP-1 utilization data the profession has had, and the plans capturing it now will file the better-supported bids whichever way BALANCE breaks.