What this document covers.

This white paper documents the technical and operational basis for the Waybright 30-day readmission warranty. It is written for underwriters, actuaries, benefits consultants, and CFOs evaluating Waybright as a maternity navigation partner. The reader is assumed to be familiar with episode-based payment design, stop-loss reinsurance structures, and the actuarial conventions of new product-line reserving.

Six questions are answered, in order:

• What is the warranty, contractually, and what does it cover?
• How is exposure modeled, and what is the empirical basis for the model?
• How are reserves set, and against what industry precedent?
• Why is the warranty priced above expected loss, and what does the loading fund?
• How are claims adjudicated, paid, and disputed?
• What is the reinsurance posture at launch, and what triggers a structural shift?

A note on what this document is not.

This paper does not replace a formal actuarial opinion. The reserve framework documented here is policy at launch — calibrated against published industry precedent and the empirical distribution of postpartum readmission events in commercial claims data — and will be refined as twelve months of Waybright-specific claims experience accumulate. Where Waybright has not yet produced its own data, this paper says so explicitly.

The warranty, in plain language.

If a Waybright-navigated member is admitted to an inpatient facility within thirty days following discharge from a delivery admission, and the admitting diagnosis is determined to be pregnancy-related per the inclusion framework in Section 5, Waybright pays the cost of that second episode at the full plan-allowed amount as adjudicated by the employer's TPA or carrier. Payment flows directly against the claim. This is the defining commitment of the product, and it is worth stating precisely: Waybright pays the cost of the readmission out of its own pocket, against its own reserves. It is not a refund of Waybright fees. It is not a credit toward future PEPM. Waybright pays the claim.

Fee-at-risk models refund what you paid the vendor. Waybright pays what the readmission costs. Those are different liabilities, and the difference is the point.

Coverage period.

The warranty applies from midnight on the date of discharge from the index delivery admission through midnight on the thirtieth day thereafter. A readmission on day thirty is covered; a readmission on day thirty-one is not. Coverage attaches per-pregnancy, not per-member; a subsequent pregnancy within the contract term begins a new warranty period.

Enrollment trigger.

A member becomes eligible upon enrollment in Waybright navigation — completion of intake risk stratification and assignment of a named navigator. Enrollment must occur prior to the third trimester (28 weeks gestation) for coverage to attach. Members enrolled after 28 weeks receive navigation but are not warranted; their status is documented in the quarterly outcome report alongside the warranted cohort.

Per-episode cap.

The warranty pays the actual plan-allowed amount of the readmission, with no per-episode cap. Waybright explicitly accepts unlimited per-episode exposure within the included diagnosis categories. Capping per-episode payment would create a perverse incentive for the warrantor to discourage adequate readmission care; an uncapped warranty aligns Waybright's incentive with the member's clinical interest.

Aggregate stop.

Per contract year, total warranty payouts to a single employer are capped at 200% of that employer's annual Waybright PEPM revenue. This protects Waybright's balance sheet against single-employer concentration risk during the launch reserve period. At the modeled net exposure described in Section 2 — a small fraction of annual PEPM revenue — the 200% aggregate is reached only in extreme tail scenarios. The cap converts at the reinsurance threshold in Section 6.

What the warranty costs us, in expectation.

Exposure proceeds from three empirical inputs: the baseline 30-day postpartum readmission rate, the expected cost of an obstetric readmission, and expected enrollment volume per contracted employer. Each is grounded in published literature, cited in the references. We model exposure as a four-step cascade and show every step, because the honest exposure number depends on which step you are standing on.

Per illustrative 10,000-employee book per year. Modeled at 220 enrolled pregnancies per 10,000 covered employees (commercial pregnancy rate ~10 per 1,000 covered lives⁷), a baseline all-cause 30-day readmission rate of 1.5% (the upper bound of typical commercial experience; published median is 1.01%¹), and a modeled obstetric readmission cost of $18,000.⁸

Step one — gross all-cause exposure.

$59,400 is the gross all-cause figure: every postpartum readmission, related to pregnancy or not.

Step two — in-scope warranted exposure.

Not every postpartum readmission falls within the inclusion framework. In the largest multi-state analysis of postpartum readmissions, psychiatric illness accounts for roughly 7.7% of 30-day events.⁸ Because mental-health and postpartum-depression admissions are excluded from warranty scope at launch, alongside unrelated injury and routine non-obstetric stays, Waybright models the in-scope warranted share at approximately 65% of gross all-cause exposure — a modeled assumption, not a measured rate. The navigator continues to screen for postpartum depression and coordinate care throughout; the exclusion is from the financial guarantee, not from the clinical service. The result, $38,610, is what Waybright is actually on the hook for before any navigation effect.

Step three — net expected payout.

Navigation effect — the readmission reduction produced by claims-integrated risk stratification, scheduled postpartum protocol, and facility steering — is modeled conservatively at 30%, against Pomelo Care's published 37% preterm-birth reduction as a category upper bound.⁹ Applying that to in-scope exposure yields $27,027, the modeled expected payout per 10,000-employee book per year.

How we reserve, and against what precedent.

Reserves are held on Waybright's balance sheet against net expected loss from Section 2, multiplied by a reserve factor. The reserve factor is the key actuarial choice; it sets the buffer between modeled expected loss and the working capital supporting the warranty.

Industry precedent.

Episode-level performance warranties and downside-risk arrangements in healthcare typically reserve at 1.3x to 2.0x expected loss, with new product lines at the upper end until claims credibility builds. Medical stop-loss carriers reserving against single-condition exposure in their first two to three underwriting years hold 1.5x to 2.0x, narrowing toward 1.2x–1.3x as credibility establishes. ACO downside-risk arrangements reserve at roughly 1.3x–1.6x of modeled downside. Surgical-bundle warranties operate on a different basis (provider-held risk) and are not directly comparable.¹⁰

Waybright's launch reserve.

Waybright reserves at 1.75x net expected exposure — toward the conservative end of new-product-line precedent, deliberately above the midpoint. The factor reflects three pre-launch realities: the absence of Waybright-specific historical claims, conservative modeling of the navigation effect, and the principle that a new warrantor benefits more from over-reserve discipline than from working-capital efficiency.

Illustrative · pending actuarial certification. Reserve held per 10,000-employee employer per year. Reviewed quarterly; recalibrated toward 1.3x–1.4x after twelve months of Waybright-specific experience as credibility develops. Reserve methodology is policy at launch, refined as claims data accumulates.

Why the embedded warranty cost is above expected loss.

This section exists because a careful reader will compute it anyway, and we would rather state it than be caught having buried it.

Waybright charges a single blended PEPM. We do not sell the warranty as a separate priced line of insurance, and we do not describe it as risk transfer, because it is not priced as risk transfer. The portion of the blended fee attributable to the warranty commitment is set materially above modeled expected loss. This is deliberate, and it is standard for a warranty as opposed to an insurance contract: extended warranties across industries — appliance, electronics, automotive, home — routinely price well above expected claim cost, because the price funds the operation that makes claims rare, not merely the claims themselves.

The loading is not a margin we hope you don't notice. It is the forcing function. It funds the navigator, the claims integration, and the screening protocol — the work that makes the warranty rarely fire.

Concretely: net expected warranty payout is roughly $27,027 per illustrative 10,000-employee book per year. The warranty commitment is embedded in a blended fee that also funds named navigation, claims-integrated risk stratification, facility steering, and the postpartum screening protocol. Underpricing the commitment to bare expected loss would remove the capital that funds that upstream work, and the upstream work is precisely what drives the modeled 30% navigation effect that lowers expected loss in the first place. The price and the prevention are the same mechanism. An actuary evaluating Waybright should read the warranty as a funded accountability commitment, not as a stop-loss policy, and should evaluate the blended PEPM against modeled total savings, not the warranty component against expected payout in isolation.

What the warranty pays. What it doesn't.

Inclusion and exclusion are determined by the primary admitting diagnosis on the readmission claim, mapped against ICD-10 categories. Below is the working framework; the exhaustive code-level list is appended to the contract and updated annually.

How a claim becomes a payment.

Mechanics are designed to minimize documentation burden on the employer and TPA. Waybright initiates the payment process; the employer or TPA confirms claim validity and receives payment. The member is never involved.

• Monthly claims feed via 837 file or SFTP equivalent, per the data exchange agreement executed at contract.
• Within five business days of receipt, Waybright flags any inpatient admission within thirty days of a delivery discharge for an enrolled member.
• The flagged admission is reviewed against the Section 5 inclusion framework; determinations are documented in the case log.
• The employer and TPA are notified within fifteen business days of flag, with the determination and proposed payment amount.
• Within thirty business days of confirmation, Waybright issues payment at the plan-allowed amount. No deductible. No coinsurance.

Dispute resolution.

When the inclusion determination is disputed — typically when the admitting diagnosis is ambiguous between pregnancy-related and unrelated cause — the case goes to joint clinical review with the employer's medical director and Waybright's chief medical officer, completed within forty-five days. Unresolved disputes escalate to a mutually selected neutral external reviewer with maternal-fetal medicine board certification, whose determination is binding. Disputes are uncommon; pregnancy-related diagnoses are usually unambiguous on admission coding.

Self-reserved at launch. Reinsured at scale.

Waybright self-reserves at launch; no reinsurance treaty is in place at first contract. At modeled exposure for a one-to-ten employer book in year one, expected aggregate exposure remains well below the threshold at which reinsurance becomes operationally efficient, and the working-capital cost of carrying reserves on Waybright's balance sheet is materially lower than a treaty over the same band.

The structural shift is triggered by whichever occurs first: aggregate annual warranty exposure across the book approaching $2,000,000 in expected loss, or covered membership exceeding 250,000 lives. At that scale a non-proportional stop-loss treaty becomes more cost-efficient than self-reservation and removes single-event concentration risk. The employer-facing warranty terms are unchanged by reinsurance; the warranty remains a Waybright obligation, with reinsurance backstopping Waybright rather than the employer.

Methodology limitations.

• Waybright-specific claims experience does not yet exist. The 30% navigation effect is calibrated against Pomelo's published outcomes, adjusted conservatively. The actual Waybright effect on 30-day readmission will not be known until twelve months of operational data accumulate.
• The 1.75x reserve factor is calibrated to industry precedent, not Waybright's own data. Recalibration is planned at month twelve.
• The exposure cascade is illustrative and pending actuarial certification; the certified figures will supersede this model.
• The exclusion framework will refine with case experience, particularly at the boundary between pregnancy-induced and pre-existing condition flares. The dispute procedure handles these without operational disruption.