Severe Illness Programmes: What the Reinsurer, the Treatment Network and the Reporting Must Deliver

The three load-bearing pillars decide whether a programme holds — not the policy wording.

---

Short answer

Severe Illness programmes rarely fail on product design. They fail on three elements routinely treated as procurement items: the reinsurer, the treatment network and the reporting. In this line the reinsurer is effectively a co-designer and the price-setter, because no own experience exists. The network is not an added service but the product itself — without centres that can actually treat the conditions, all that remains is a promise to pay. And reporting must produce leading indicators, because a Severe Illness book is statistically almost invisible in its first two years. If any one pillar is missing, the programme does not hold, however well the wording is drafted.

---

Key points

• Severe Illness is an access product, not only a financing product. Its value arises where local treatment capability is insufficient.
• Severe Illness is not Critical Illness. Treatment costs versus lump sum — two lines of business, two pricing logics, two distribution routes.
• The reinsurer delivers more than capacity: the pricing model, the condition definitions, the anti-selection framework and claims expertise. Selecting on rate alone is a mistake in this line.
• Back-to-back wording is not negotiable. Any gap between policy and treaty is carried by the cedant alone.
• Network depth beats network size. What matters is a few centres with demonstrated case volume in the covered condition groups.
• Episode pricing rather than fee-for-service is the single most effective cost instrument — and in this line it is negotiable.
• The metric is incidence, not loss ratio. The loss ratio stabilises too late to steer by.
• A Severe Illness programme needs five years to credible experience. Contracts and expectations must be built for that horizon.

---

1. What a Severe Illness programme is — and is not

Severe Illness covers the treatment costs of defined serious conditions: typically oncological disease, major cardiovascular events and procedures, complex neurosurgery, organ and stem-cell transplantation, and sometimes advanced or rare disease. Cover is trigger-based — it responds on diagnosis of a listed condition — with high limits and, as a rule, tied to a treatment network.

Critical Illness, by contrast, pays a lump sum on diagnosis. The amount bears no relationship to actual treatment cost. Systematically that is a benefit product; Severe Illness is medical expense insurance.

This distinction is not semantics. It determines the line of business and therefore the authorisation or notification requirement, the pricing logic, the tax treatment and the distribution route. A programme that blends the two terms in its concept notifies the wrong line and discovers it at the first claim.

The economic signature of Severe Illness is unusual and explains everything that follows:

• very low frequency — a book of 20,000 lives produces a two-digit number of claims per year, depending on target group and condition list;
• very high severity — limits of one to two million euros or dollars per condition are market standard, and individual treatment courses can exhaust them;
• long treatment courses — an oncology pathway runs over quarters, not days;
• high exposure to anti-selection — people buy when they suspect something.

It follows that a single year's loss ratio says almost nothing. It is a function of whether two or seven large claims happened to occur in the observation period. A programme steered on annual loss ratios is steered wrongly.

And the decisive conceptual point: Severe Illness is typically sold where access to advanced care is constrained — in growth markets, to diaspora and expatriate populations, to corporate books in regions with limited oncology, cardiac or transplant capacity. The customer is not primarily buying money. They are buying access to treatment. That is why the network is not a service feature but the substance of the promise.

---

2. The programme architecture at a glance

A viable Severe Illness programme involves six roles that rarely sit with one party:

The three pillars of this article cut across these roles. They are therefore exactly the places where responsibility routinely falls between parties.

---

3. Pillar 1: What the reinsurer must deliver

In Severe Illness the reinsurer is not the capacity supplier at the end of the chain. It is effectively a co-designer of the product. Treating it as an interchangeable vendor and selecting on rate loses precisely the contribution that matters.

3.1 Structure, not just capacity

The usual base is a quota share. It splits frequency and severity proportionally, aligns interests and supports the cedant across the whole book. Above it belongs an excess of loss per risk for the individual-case tail.

What is routinely forgotten in programme negotiations and makes the difference in the early years: an aggregate stop-loss for the ramp-up phase. As long as the book has no own experience, annual volatility cannot be calculated, only capped. A programme that comes under discussion in year two because of two unfavourable large claims regularly loses its distribution partner — not because the product was wrong, but because nobody priced the volatility in advance.

3.2 Pricing in a data-poor line

At launch there is no own experience. The price must therefore be constructed from incidence data (cancer registries, national and international health statistics), care pathway costs by country and centre, and utilisation assumptions — because not every diagnosed member travels for treatment.

The question to put to the reinsurer: is the price built from incidence × pathway cost × utilisation, or is it a loaded Critical Illness rate? The latter occurs more often than one would expect and is systematically wrong. CI incidence is not SI utilisation: some of those diagnosed never take up the treatment benefit, while others generate a multiple of what a lump sum would have cost.

3.3 Condition definitions and back-to-back wording

The reinsurer should co-own the wording — in particular the definitions of covered conditions and trigger points. That is where later disputes live: staging thresholds in oncology, treatment of in-situ findings, the definition of a transplant event, treatment start versus date of diagnosis.

And then the sentence that decides the balance sheet in this line: policy and reinsurance treaty must be congruent. Any benefit owed under the policy but excluded under the treaty is carried by the cedant unprotected. With individual cases in the high six or seven figures, a single such gap is existential. Back-to-back wording is not a negotiating position; it is a precondition.

3.4 Anti-selection

Severe Illness is as exposed to anti-selection as any cover in the market. The available instruments:

• waiting periods — regularly longer in oncology than in other condition groups;
• exclusion of pre-existing conditions and their sequelae, with a clearly defined knowledge standard;
• medical underwriting in individual business versus guaranteed acceptance with a participation threshold in group business;
• minimum group size and minimum participation rate — the most effective instrument in corporate business;
• automatic inclusion of whole workforces rather than opt-in.

The critical point is that these requirements must be settled before the distribution agreement. Programmes are regularly signed with a distribution partner whose model structurally cannot achieve the participation rate the reinsurer requires. That is not a sales problem to be solved later; it is a design fault.

The fact that market entry sensibly begins with employee benefits in corporate business, with retail and affinity following later, has exactly this reason: in a group, participation rate and risk mix are controllable; in individual business they are not.

3.5 Continuation of treatment and run-off

The most underestimated item: who carries an ongoing course of treatment when the contract ends, the member leaves, or the programme is discontinued?

An eighteen-month oncology pathway outlives any annual policy period. Where the policy promises continuation of treatment already begun but the treaty ends on a clean-cut basis, an unfunded liability arises. The treaty's run-off provision must match the policy's promise — both must cover the same period.

3.6 What a good reinsurer brings in addition

Medical underwriting support, adjudication expertise on complex individual cases, own or partner networks, benchmark data from comparable books. In the ramp-up phase that contribution is worth more than a few points of commission.

And a selection criterion rarely tested: will this reinsurer still be on risk in five years? A programme needs that period to reach credible experience. Appetite shifts in health lines are real. The question is not only the rating but continuity in the segment.

---

4. Pillar 2: What the treatment network must deliver

4.1 The network is the product

For a member in a market with limited oncology capacity, the promise reads: "you will be treated at a centre that can treat this." Without named, contracted and genuinely capable centres, what remains is a promise to pay — substantially less valuable, substantially harder to sell, and substantially harder to honour at the moment of claim.

4.2 Depth over breadth

Here Severe Illness differs fundamentally from travel or iPMI networks. What is needed is not coverage across countries but depth in a few centres for a few condition groups. The tests:

• Condition-specific capability. Does this centre actually perform allogeneic transplants? Does it have the specific therapeutic option the condition list promises?
• Case volume in the indication. The relationship between volume and outcome in complex surgery and oncology is well established. Volume is therefore a harder criterion than any accreditation.
• Accreditation as a floor, not as proof — international accreditation says something about process, little about outcomes in a given indication.
• Admission capacity and timelines for international patients. A centre with a six-week wait for foreign patients is not a network partner for a new oncology diagnosis.
• International patient department. Language, visa support, records transfer, support for accompanying family.

4.3 Geographic tiering as a pricing instrument

The standard architecture uses three tiers: the home market, where local capability exists; regional hubs with internationally recognised centres and a substantially lower cost base; and a global tier for what is not available regionally, usually sub-limited or subject to specific approval.

The cost difference between tiers is not a percentage but an order of magnitude. Tiering is therefore not a service feature but the product's central pricing instrument. It belongs in the wording, not the brochure.

4.4 Episode pricing rather than fee-for-service

The single most effective cost instrument in this line: agreeing package or episode prices per treatment episode instead of fee-for-service billing. An episode price caps the tail and makes the case calculable before it occurs.

And it is negotiable — precisely here. Severe Illness patients are, for a centre, high-value, plannable, internationally funded admissions with a secured payment undertaking. That is the position from which centres agree package prices. Not having that conversation forfeits the largest available lever.

4.5 The second opinion as a gate

A structured medical second opinion before treatment approval does three things at once. It confirms the diagnosis — and Severe Illness triggers are diagnosis-based, while misdiagnosis and overdiagnosis are real. It tests the proposed treatment plan against guidelines. And it directs the patient to the right centre.

It is therefore simultaneously the strongest quality feature in distribution and the sharpest cost control. Few instruments in insurance are both sellable and effective. This one is.

4.6 Logistics, payment and provider trust

Cross-border treatment of serious illness means visas, travel with a companion, accommodation, transfers, translation, continuity of records and return travel. If that chain is not clearly owned, it lands in the insurer's call centre — at the worst moment in the customer relationship.

And centres do not admit international patients on a promise. A robust guarantee of payment is needed, issued quickly and honoured quickly. A payer with reliable settlement obtains better package prices and faster admission slots than one that negotiates hard and pays late. Payment discipline is a pricing instrument here too.

---

5. Pillar 3: What the reporting must deliver

5.1 The core problem: statistical invisibility

A Severe Illness book produces too few cases in its early years to be meaningful. Loss ratios in months 1 to 24 are noise. A conventional quarterly bordereau of premium, paid and outstanding delivers nothing steerable in that phase — and by the time it does become meaningful, the opportunity to correct has passed.

Reporting must therefore be designed around leading indicators, not results.

5.2 The reporting layers

5.3 Three points that make the difference

First: case reserves must be medically informed. An IBNR factor applied to a book with forty claims is a guess without a basis. Every open case should carry an expected-total-cost estimate from medical case management, updated as the treatment plan evolves. The case manager knows whether six or eighteen months of therapy remain. No reserving model knows that. This is the single largest improvement most programmes can make to their financial reporting.

Second: measure incidence against the pricing assumption, not the loss ratio against budget. The loss ratio tells you what happened. Incidence per 1,000 lives by condition group tells you whether the pricing model is right — and it stabilises considerably earlier. An insurer that finds in year two that cancer incidence in the book is half again above assumption can act. One that finds a 130 per cent loss ratio in year four can only negotiate.

Third: the mix of new entrants is the anti-selection alarm. Anti-selection shows up in the age, gender and geographic structure of new entrants long before it shows up in claims. This report costs almost nothing and is almost never produced.

5.4 Reporting is a contractual deliverable

If the contracts with network operator, TPA and distribution partner do not specify field by field what data are delivered, in what format and at what cadence, the risk carrier will receive PDF reports. To be fixed before launch:

• data definitions — diagnosis coding to ICD-10 as a minimum, plus a pre-agreed mapping to the condition groups of the policy list; unambiguous centre IDs; episode boundaries;
• format and transmission, machine-readable, not as documents;
• cadence and deadlines, with consequences for non-delivery;
• data ownership — who owns the claims data at contract end? Without that clause the carrier loses its entire experience base when it changes provider.

Retrofitting a data model onto a live programme is expensive and usually stays incomplete.

And a governance point: carrier, reinsurer and distribution partner must read from one reporting set. Divergent figures are the most reliable way to turn a renewal discussion into a dispute.

---

6. How the three pillars interlock

They are not parallel but mutually dependent:

• The reinsurer defines conditions and triggers — which determines what the network must be able to treat.
• The network tiers and episode prices determine cost per case — which underpins the pricing model and the reinsurance structure.
• The reporting produces the experience — which determines whether the reinsurer renews and at what price, and therefore whether the programme survives.

The loop closes. The governance consequence is uncomfortable but clear: one party must own the programme as a whole. If carrier, reinsurer, MGA, network and TPA each optimise their own segment, nobody optimises the programme. In practice the programme manager takes that role — provided the mandate extends beyond the binding authority.

---

7. Timeline: when a programme counts as viable

The practical consequence: every commercial agreement — reinsurance treaty, distribution contract, partner expectations, investor objectives — must be built for that horizon. A programme whose participants expect a stable result in year two will come apart in year two, regardless of its quality.

---

8. Eight mistakes that sink programmes

1. Critical Illness thinking applied to Severe Illness. Lump-sum logic and treatment-cost logic are not interchangeable.
2. Policy and treaty not congruent. Every gap is carried by the cedant alone.
3. Network bought by size rather than by case volume in the covered condition groups.
4. Fee-for-service instead of episode pricing. The tail stays open.
5. No second opinion as an approval gate. Diagnosis, treatment plan and centre go unchecked.
6. Continuation of treatment without reinsurance cover. The policy promises longer than the treaty carries.
7. Reporting designed as a finance deliverable rather than as a steering instrument with leading indicators.
8. Reacting to the first year's loss ratio. It contains no usable information.

---

9. Regulatory placement for DACH programmes

For programmes with a German or Austrian nexus, four points need settling upfront:

• Line of business. Severe Illness is medical expense insurance, not life or accident. Authorisation or notification must cover the correct line. Where a Critical Illness component sits in the same programme, that means two lines — and depending on structure, two legal entities.
• Non-substitutive. Severe Illness does not replace statutory health cover and therefore, as a rule, falls outside the substitutive health insurance regime under § 146 VAG. That classification should be documented, not assumed, particularly for target groups without statutory cover.
• Product oversight and governance. The process under § 23(1a) VAG is substantive for a trigger-based product with a condition list: target market, negative target market and distribution strategy must reflect who a cover bound by waiting periods and pre-existing exclusions is actually suitable for.
• Data protection. Diagnosis data are special categories under Art. 9 GDPR. A programme with centres and second-opinion providers outside the EU requires processing agreements, appropriate safeguards and a documented transfer impact assessment — plus a clean separation of medical and administrative processing.

---

10. Outlook: the cost pressure comes from therapy

One structural development will affect every Severe Illness programme in the coming years: cost per case is rising not through hospital price inflation but through new therapeutic options. Cell and gene therapies and targeted oncology agents push the ceiling of what a single treatment episode can cost substantially higher.

In parallel, a common European assessment framework is emerging. The EU HTA Regulation (EU) 2021/2282 has applied since 12 January 2025, initially covering new oncology active substances and advanced therapy medicinal products; selected high-risk medical devices follow from 2026, orphan medicines from 2028 and all newly authorised medicines from 2030. Joint clinical assessment does not replace national reimbursement decisions, but it changes the evidence base on which payers decide about new therapies.

Three implications for programme design: limits and sub-limits must reflect the new cost dimension rather than absorb it without bound; the condition list needs a defined rule for handling new therapeutic modalities; and network agreements should provide for episode pricing on high-cost therapies too, rather than leaving them open as an exception.

---

Frequently asked questions

About the author

Frank Reuter designs and implements Severe Illness programmes across several markets — from product architecture through reinsurance and network structure to administration. Further focus areas: international private medical insurance (iPMI), claims handling and cost containment, DACH market entry, and M&A in healthcare and insurance. He is a broker in charge under § 34d GewO, holds Prokurist and managing director roles in several international insurance and assistance structures, and takes on interim mandates at executive level.

Reuter Consulting, Munich · frank.reuter@reuterconsulting.com

From this series: Severe Illness topic page · DACH market entry for foreign personal lines insurers · Cost containment in international medical claims

---

This article reflects the position as at September 2026 and is provided for general information. It does not replace legal, actuarial or medical advice in individual cases.