Claims Handling: Why Cost Containment for Medical Treatment Abroad Is Not Price Haggling
Between the invoice submitted and a reasonable tariff sit review, care pathway steering and network management.
Short answer
Cost containment in international claims is routinely confused with after-the-fact discount negotiation. That is not only the weakest lever — it is the legally weakest one. The insurer is not a party to the treatment contract. Without a network agreement or a prior guarantee of payment, it has no claim against the provider; it can only reduce payment to its own policyholder, which pushes the difference onto the patient. Effective cost steering therefore happens before the invoice exists: through medical triage at first contact, care pathway steering, network contracts, and a guarantee of payment with a defined scope.
Key points
- The insurer is not the physician's counterparty. That single fact explains why post-hoc negotiation is structurally weak.
- The biggest single lever sits in the first 30 minutes — in where and at what level of care the patient is treated, not in the unit price of a line item.
- Abroad, the yardstick is missing. Germany's GOÄ and Switzerland's TARDOC have no extraterritorial effect. Without a reference, you are negotiating over a number with no basis.
- Three questions must never be merged: Is it covered? Was it necessary? Is the price reasonable? The money sits in question two.
- Fast payment lowers prices. Providers price in payment risk and delay.
- The savings rate is a vanity metric. It measures how inflated the incoming invoice was, not how well the case was steered.
- Every unilaterally deducted euro that lands on the patient is not a saving but a transfer — with complaint, ombudsman and lapse consequences.
1. The conceptual error
In many organisations, "cost containment" is understood as a function that reviews incoming foreign invoices and pushes the amount down. It is organised accordingly: downstream of claims handling, with access to the invoice and to nothing else.
The problem is not the review itself, which is necessary. The problem is the timing. By the time the invoice arrives, everything that actually determined the cost has already been decided: the patient went to an emergency department rather than a physician's office. They were admitted as an inpatient rather than treated ambulantly. A CT was performed where a clinical examination would have sufficed. They stayed six days instead of three. Treatment happened at a private clinic in a tourist district rather than at the contracted hospital twenty minutes away.
The sum of those decisions is the cost of the case. The unit price per line item is the smaller part — and the only one still open to challenge afterwards.
The point: Starting at the invoice is not cost steering. It is negotiating over what is left.
2. Why after-the-fact negotiation is legally weak
The decisive point is rarely stated plainly: the insurer is not a party to the treatment contract.
In a cross-border claim, the treatment contract exists between patient and provider under the law of the place of treatment. The insurer sits outside that relationship. It has no claim to a price reduction against the clinic — unless it has created one: through a network agreement, through a guarantee of payment with an agreed scope, or through membership of a network with contractually established tariffs.
Absent that prior legal relationship, only one option remains: reducing payment to its own policyholder. That is legally possible — in Germany, for example, via medical necessity under § 192(1) VVG or, where there is a conspicuous disproportion between charge and service rendered, under § 192(2) VVG. But it has a consequence that never appears in the savings statistics: the difference lands on the patient. The provider pursues it with them. The customer complains, goes to the ombudsman, lapses.
That is not a saving but a transfer. It shows up as success in claims expenditure and as a loss in complaints, lapse and reputation — only in a different department.
Sustainable cost steering is therefore provider-side and contractual, not insured-side and retrospective.
3. The missing price tag
Domestically, an objective yardstick exists. Germany has the GOÄ, a fee schedule with defined items and multipliers; § 12 GOÄ requires a formally proper invoice, without which the claim does not fall due. Switzerland has had TARDOC and outpatient flat rates as a new nationally uniform tariff structure since 1 January 2026, replacing TARMED. Austria, France, the Netherlands — each with its own reference system, but each with one.
Outside those systems, the yardstick does not exist. In Miami, Bangkok, Dubai or Istanbul there is no fee schedule a European insurer could invoke. Prices emerge from market position, willingness to pay and — in tourist regions — from openly practised dual pricing: one tariff for self-payers, another for the insured.
Hence the second core proposition of this article: if you want to steer cost abroad, you have to create the yardstick yourself. There are exactly three ways to do it:
- Contractually — network agreements with established tariffs or discount structures.
- Case by case — a guarantee of payment that defines scope, expected length of stay and cost frame before treatment.
- Empirically — a reference database of own and third-party claims data showing a market level by diagnosis group, country and level of care.
Without at least one of these three anchors, any price discussion is an argument about a number with no reference point. And that argument is regularly won by the side that has already treated and discharged the patient.
4. Three questions that must never be merged
Much of the friction in claims handling arises because three distinct reviews collapse into one step:
| Question | Subject of review | Owner | Where the money is |
|---|---|---|---|
| 1. Coverage | Policy status, territorial scope, waiting periods, exclusions, pre-existing conditions, deductible | Claims handling | low value, wide spread |
| 2. Medical necessity and appropriateness of care | Was treatment indicated? At the right level of care? For the right duration? | Medical team / case management | the bulk sits here |
| 3. Price reasonableness | Coding, duplicate billing, unbundling, tariff level against reference | Bill review / repricing | relevant but secondary |
Asking question 3 without having answered question 2 means arguing over the unit price of a service that may not have needed to be rendered at all. That is the most expensive mistake in claims handling — and the most common.
Conversely, question 2 can only be answered if medical competence sits inside the process. A purely administrative review can find formal errors. It cannot judge whether an inpatient admission was indicated.
5. The steering chain
Cost containment is not a review step but a chain of intervention points. Thinking about it in this order is the actual difference between processing and steering.
| Stage | Timing | Lever | Effect |
|---|---|---|---|
| 1. Product and wording design | before inception | Network obligation, pre-authorisation requirement, deductibles, exclusions | creates the steering rights in the first place |
| 2. First contact / medical triage | minute 0–30 | Level of care: telemedicine, office, urgent care, emergency department | largest single lever |
| 3. Provider steering | before treatment starts | Direction into the contracted network | high |
| 4. Guarantee of payment | before treatment starts | Scope, length of stay, cost frame fixed | high; converts open into defined liability |
| 5. Medical case management | during treatment | Length of stay, ambulatory potential, step-down, repatriation decision | high in severe cases |
| 6. Bill review and repricing | after treatment | Coding, unbundling, tariff comparison | medium |
| 7. Recovery and coordination of benefits | after payment | Third-party liability, statutory first payer, double insurance | medium, often unused |
Stages 1 to 5 act on the cost of the case. Stages 6 and 7 act on the residual amount. Both are necessary. But in most organisations the order of attention is exactly reversed.
A note on classification: the weighting of levers reflects experience from international claims portfolios and is not a published industry benchmark. Reliable magnitudes only emerge from analysing your own book.
6. The first call
The largest lever is also the cheapest — and the least developed.
An insured with abdominal pain on holiday has four realistic paths: a telemedicine consultation, a physician's office, an urgent care facility, or the emergency department of a private hospital. For uncomplicated cases all four produce the same clinical outcome. The cost spread between the first and the fourth is substantial — in many destination markets a double-digit multiple.
Which path the patient takes is decided in the first few minutes and depends on three things:
- Reachability. An assistance line staffed around the clock in the relevant language gets used. One that is not, does not.
- Medical competence on the phone. Triage is a clinical activity. A call centre without physician backup can provide addresses; it cannot steer.
- Incentive in the policy wording. If the contract treats network and free choice identically, there is no reason to call. Differentiated deductibles, direct settlement only inside the network, or a pre-authorisation threshold change behaviour far more effectively than any communications campaign.
The decisive point is that the incentive has to come from the product, not from the claim. Steering rights that are not in the wording do not exist when the claim occurs.
7. Network and guarantee of payment
The network is where the missing yardstick is created contractually. What matters is not network size or country count — the usual vendor selling points — but:
- Depth where your claims actually occur. A network of 900,000 providers worldwide is of little use if there is no contracted hospital in the five destinations accounting for 60 per cent of your claims cost.
- Type of agreement. A fixed tariff, a discount off list price, and a plain direct-settlement arrangement are three very different things. Only the first is price steering.
- Balance billing. A network agreement that does not exclude additional charges to the patient shifts the residual risk onto the insured.
- Payment terms. More on this shortly.
The guarantee of payment is the sharpest single instrument because it moves the moment of price determination forward. Before treatment, the insurer has a negotiating position: the provider wants the undertaking. After treatment, it does not.
A well-constructed guarantee of payment therefore contains more than a payment undertaking: scope of treatment, expected length of stay, cost frame, an explicit exclusion of additional charges to the patient, and a notification obligation on deviation. A guarantee of payment without those elements is a blank cheque on letterhead.
8. Bill review: what is actually examined
Even though it is not the largest lever, it is indispensable — and it involves considerably more than a price comparison:
- Formal review. Policy validity, treatment date, patient identity, completeness and traceability of the invoice. In Germany § 12 GOÄ provides a clear standard; abroad the requirement must be set contractually.
- Coding review. Do the billed items match the documented diagnosis? Classic findings: upcoding, unbundling of items that belong together, duplicate billing, standard consumables billed separately, inflated facility fees.
- Clinical plausibility. Do diagnosis, therapy and length of stay cohere? This requires physician judgement, not a rule set.
- Tariff comparison. Against the network agreement, against a local reference, or against your own data for the same diagnosis group in the same country.
- Abuse patterns. Clusters around individual providers, systematic referral chains between hotels, agents and clinics, identical treatment patterns across different patients.
Point 5 stays invisible without data analysis. Case-by-case review finds suspicious invoices. Only portfolio analysis finds suspicious providers.
9. Travel medical and iPMI: two different toolkits
The two models are frequently conflated. Their steering options differ fundamentally:
| Travel medical insurance | International private medical insurance (iPMI) | |
|---|---|---|
| Nature of the case | acute, unforeseen, short | plannable and potentially chronic, long |
| Pre-treatment steering | minimal; the event arrives without notice | feasible; pre-authorisation can be established |
| Network obligation | hard to enforce; the patient is a visitor | enforceable; the insured is a local resident |
| Dominant lever | first-contact triage, repatriation decision | network steering, case management, pre-authorisation |
| Claims profile | many small claims, occasional large losses | higher average claim, more predictable |
| Coordination of benefits | highly relevant: statutory first payer in EU/EEA/CH | relevant where local compulsory cover exists |
| Repatriation | a genuine economic option | rare; the insured lives at the place of treatment |
The coordination of benefits point deserves attention because it is routinely left on the table. For statutorily insured German residents, coordination under Regulation (EC) No 883/2004 applies within the EU, the EEA and Switzerland; necessary care is covered via the European Health Insurance Card. Private travel medical cover is then economically subordinate or supplementary. An insurer that does not pursue that first-payer entitlement systematically pays amounts it does not owe — and it never shows up in a savings statistic.
On repatriation: the choice between continued local treatment and return transport is first a clinical decision. But it carries an economic dimension that is significant in high-cost markets — in the United States an organised repatriation can cost less than a few additional inpatient days. That trade-off belongs with a physician who has cost visibility, not with an administrator holding a repatriation budget.
10. Operating model — and an incentive problem that is rarely discussed
"In-house, TPA or hybrid" is usually treated as a cost question. It is a control question.
In-house gives full control over data and process but requires 24/7 operations, medical capacity and network access — hard to justify economically on a small book. TPA delivers network, scale and round-the-clock operations but costs data ownership and steering depth unless the contract explicitly secures them. Hybrid — core steering internal, processing and network external — is in practice often the most viable solution for mid-sized books, but demands clean interfaces and genuine data return.
Regardless of model, there is one design flaw as widespread as it is rarely discussed: performance fees based on savings achieved.
Where a service provider is paid a percentage of the difference between the invoice submitted and the amount paid, it has an economic interest in high incoming invoices. Precisely the steering that prevents an inflated invoice from arising — triage, network direction, guarantee of payment — reduces its own fee. The model rewards cost correction, not cost avoidance.
More robust are blended models: a base fee per case or per insured member, plus performance components tied to cost per case against benchmark rather than to the difference from the submitted invoice. That is harder to agree and harder to measure. But it measures the right thing.
And one point that beats most negotiation strategies: providers price in payment risk and delay. A payer that reliably settles within 14 to 21 days systematically obtains better terms than one that negotiates hard and pays in 120 days. Payment discipline is a pricing instrument — and one of the few that improves the provider relationship rather than straining it.
11. Metrics: the savings rate measures the wrong thing
The most common cost containment metric is the savings rate against the submitted invoice. It has a fundamental design flaw: it rises when incoming invoices are inflated. A perfectly steered portfolio produces accurate invoices — and therefore a low savings rate. The metric penalises good steering.
A small alternative set is more informative:
| Metric | What it shows |
|---|---|
| Cost per case by diagnosis group against country benchmark | whether the level is right, independent of the invoiced amount |
| First-contact rate | share of cases with contact before care is accessed — the steering access itself |
| Network penetration | share of claims cost inside the contracted network |
| Guarantee-of-payment rate | share of inpatient cases with prior undertaking |
| Ambulatory conversion and length of stay | effectiveness of case management |
| Turnaround time to payment | price driver and early indicator of network quality |
| Balance-billing and complaint rate | whether the saving is real or merely transferred |
| Coordination rate | share of cases where third-party first payment was pursued |
The last two rows are the most honest. They show whether a reported saving relieved the book — or only the insurer, at the customer's expense.
12. Law and data protection
Three points that need to be robustly settled in international operations:
Medical necessity. In Germany it is the trigger for the reimbursement obligation (§ 192(1) VVG); reduction for conspicuous disproportion between charge and service (§ 192(2) VVG) is subject to a high threshold and is not a general right of appropriateness review. Building a cost strategy on it overestimates the instrument.
Collecting data from third parties. Obtaining health data from physicians, clinics or other insurers requires consent (§ 213 VVG). Queries to a treating physician abroad presuppose valid, sufficiently specific consent — the wording used at application stage determines what can be reviewed at claim stage.
Data protection. Health data are special categories of personal data under Art. 9 GDPR. Any transfer to a TPA, network operator or reinsurer outside the EU requires a processing agreement, appropriate safeguards and a documented transfer impact assessment. Equally important and more often overlooked: separating medical from administrative review. The claims handler needs the review outcome, not the diagnosis.
13. Seven mistakes in practice
- Cost containment as a downstream review function rather than a steering chain across the whole case.
- Steering rights missing from the wording. What is not agreed cannot be enforced when the claim occurs.
- No medical access at first contact. Triage without physician backup is address provision.
- Network selected by size rather than by depth in your own claims concentrations.
- Guarantee of payment without scope, length of stay and balance-billing exclusion.
- Performance fees based on invoice differential — an incentive against upstream steering.
- No coordination of benefits. Third-party first-payer entitlements go unpursued because they are nobody's KPI.
14. Outlook: the reference systems are shifting
Two developments are changing the benchmarks against which claims are reviewed:
Switzerland. Since 1 January 2026, TARDOC and outpatient flat rates have been in force, replacing TARMED after more than twenty years. The system comprises roughly 1,370 TARDOC items and 315 flat rates; mixed billing between the two is not permitted. For review routines and billing systems this means a complete rebuild of the Swiss reference logic.
Germany. Reform of the GOÄ is finally moving after decades. The Federal Medical Association and the association of private health insurers submitted an updated version of their joint draft to the Federal Health Ministry in July 2026; the ministry is preparing a departmental draft. The draft contains more than 5,500 billable items. Both sides consider entry into force on 1 January 2028 achievable, though no binding date has been set. Until then the existing GOÄ applies unchanged.
The practical implication for claims operations and TPA contracts: review rule sets, billing systems and reference databases must be versionable. Hard-wiring review logic to one fee schedule builds yourself a migration project.
Frequently asked questions
About the author
Frank Reuter advises health and travel insurers, MGAs and investors on claims handling, cost containment and TPA operating models in international business. Focus areas: international private medical insurance (iPMI), Severe Illness, network and claims architecture, 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 · Severe Illness programmes: the three load-bearing pillars
This article reflects the position as at September 2026 and is provided for general information. It does not replace legal or medical advice in individual cases.