Declined international medical insurance claims, and how to appeal them

A decline is a decision about a clause

When a claim is refused, something specific has happened: an assessor has matched the facts of your treatment against a provision in the wording and concluded that the provision applies. That is a narrower event than it feels like, and the narrowness is what makes it possible to challenge.

The useful question is therefore never why they will not pay. It is which clause, and on what evidence. Almost every successful appeal consists of new material addressed at one identified provision; almost every unsuccessful one is a general account of how unfair the outcome is, sent to a claims team with no authority to depart from the wording.

The grounds fall into three groups: what you disclosed, where your cover's boundaries sit, and whether the treatment itself qualifies. The third is the least understood and increasingly the most important.

Disclosure grounds: what was said at application

Non-disclosure is the most serious category, because the remedy is not always limited to the claim in front of the assessor.

AXA's published wording is unusually direct: where a question was not answered fully or accurately, even if by accident, treatment for that condition is not covered. There is no requirement that you intended to mislead. The William Russell plan agreement goes further on remedy — where an application omits facts or contains materially incorrect or incomplete information, the insurer reserves the right to declare the policy void, or alternatively to impose special terms which apply from the original date of entry. The second option is the one people miss: cover is not cancelled, it is retrospectively rewritten, so a claim you assumed was covered was never covered on the amended terms. Wordings separately carry fraud provisions covering members who mislead intentionally or carelessly, at application or when claiming, with consequences extending to cancellation without refund and recovery of sums already paid.

This ground is settled long before anyone is ill: if you are unsure whether something belonged on your application form, raise it now, in writing. The mechanics are set out in disclosure, non-disclosure and voided cover.

Cover-boundary grounds: conditions, clocks, limits and geography

The second group is where most declines sit, and none implies you did anything wrong.

Pre-existing and linked conditions under a moratorium. A moratorium excludes conditions from before you joined until you have completed a defined period of membership and a defined trouble-free period. The trap is linkage: insurers publish tables connecting conditions to others they will also decline. A diabetes diagnosis can carry ischaemic heart disease with it; treated hypertension can carry cataract, diabetic retinopathy, arterial disease and stroke. The claim being declined may be for a condition you have never been treated for, on the basis of one you were — see how the moratorium clock actually works and linked exclusions.

Benefit exhausted or an inner limit reached. The annual maximum is rarely the constraint. Inner limits are — a per-year cap on a specific benefit, a lifetime cap on mental health treatment, a per-transplant limit, a night-count on accommodation. A claim can be refused while millions remain on the headline limit, as explained in annual benefit limits and inner limits.

Treatment outside the area of cover. Zone structures are more granular than most members realise: some zones exclude a territory entirely, others cover it only on a restricted emergency basis with a monetary cap and a trip-length limit. Treatment sought during a trip that ran past the permitted duration is outside cover even though the territory appears on the certificate. See area of cover explained.

An unexpired waiting period. Maternity, well-being benefits, preventive cancer treatment, HIV treatment and dental cover commonly carry their own clocks, running from your date of entry rather than from renewal.

No pre-authorisation. Usually a reduction rather than a refusal — one wording reimburses 80% of eligible costs, another may cover 75% where the treatment is later proven medically necessary. But some categories, notably treatment in the United States under a plan with restricted US cover, are not paid at all without prior authorisation. See how an IPMI claim works.

A reasonable and customary shortfall. Where a provider charges materially above the local going rate, the insurer pays the customary amount and the balance is yours. That is a partial decline, and it is often mistaken for an administrative error.

Treatment-eligibility grounds, and the conventional treatment test

The category that catches sophisticated readers off guard is not about you or your policy at all. It is about whether the medicine qualifies.

AXA publishes the clearest version of this test, and it is worth understanding as the market's benchmark. To be eligible, treatment must be established as best medical practice in the country where it is being given — not somewhere else, and not in the country where the drug was developed. It must be clinically appropriate. And it must be proven effective and safe through high-quality clinical trial evidence.

Applied to drugs, a medicine must be licensed by the relevant authority — the MHRA in the United Kingdom, the EMA in Europe outside the UK, or the FDA elsewhere — and used within the terms of that licence. A licensed drug prescribed off-label for a different indication can fail the test even though the prescription is lawful and clinically reasonable. Applied to surgery, the procedure must appear in AXA's published schedule of procedures and fees, with an exception route where an authoritative medical body recognises it.

AXA also closes the obvious gap: complications arising from unproven or experimental treatment are not covered either. A member who self-funds a novel therapy is therefore not merely unfunded for the therapy — they are unfunded for what follows it.

This matters for anyone considering treatment in a permissive regulatory environment, anyone offered a promising oncology agent outside its licensed indication, and anyone drawn to a clinic marketing a procedure that has not reached a mainstream schedule. Insurers reinforce the test with a right to review treatment and decline what they consider inappropriate against established clinical practice.

Getting it in writing, then building the appeal

Ask for the decline in writing, with the clause relied on and the evidence used — not a benefit code, and not a verbal summary. You cannot appeal what you cannot see, and a decision that cannot be tied to a provision is sometimes one that has not been properly made, so the request itself occasionally resolves the matter.

Ask at the same time what further evidence would cause the decision to be reconsidered. It saves commissioning the wrong report. Wordings generally place the cost of additional medical information on the member — including any independent medical examination the insurer requires — and provide that refusing to supply reasonably requested information means the claim cannot be assessed at all.

An internal appeal is then a request to reconsider on new material aimed at the identified ground.

If the ground is that a condition is pre-existing or linked to one, the evidence needed is clinical: a consultant's letter addressing date of onset, the absence of prior symptoms, investigation or medication, and, where relevant, why the current condition is not pathologically linked to the earlier one. If the ground is medical necessity, the letter must explain why this treatment, at this time, in this setting, was clinically indicated, and what the alternatives were. If the ground is that the treatment is unproven, the response has to engage with the test on its own terms — the marketing authorisation and the indication, the trial evidence, or recognition by an authoritative medical body in the country of treatment.

If the ground is a limit, geography or a waiting period, new clinical evidence will not help. Those are arithmetic and calendar questions, and the appeal is either a factual correction — a date, a zone, a benefit misallocated to the wrong heading — or it has no prospect. Recognising which kind of ground you are facing saves months.

The formal complaint, and where it goes next

If the appeal fails, the next stage is not a further appeal. It is the insurer's formal complaints procedure, a regulated process with defined clocks.

The William Russell wording states that a complaint will be acknowledged within two working days, that the insurer will investigate and respond within four weeks, and that it will endeavour to issue a final response within eight weeks, writing to explain any delay. The NIMBL wording, handled primarily under FCA rules by an FCA-authorised UK entity, allows up to eight weeks for a final decision. The eight-week mark recurs because it is the point at which the external route generally opens, whether or not a final response has been issued.

Put the complaint in writing to the address the wording specifies rather than to the claims team — complaint handling is a separate function with separate authority — and state the outcome you are seeking, since a complaint that identifies a remedy is easier to concede than one that expresses dissatisfaction.

External escalation depends on who carries the risk

Here is the part that catches people out, and it is structural rather than procedural. The external route is determined by who carries the risk and where that entity is regulated, not by the brand on your card.

International medical cover is routinely arranged by one company, administered and claims-handled by another, and underwritten by a third in a different country. Two current client wordings show how far apart the answers can be.

Under the William Russell arrangement, the product is arranged by a Belgian-registered entity whose UK branch is FCA-authorised, William Russell acts as mandated underwriter for policy administration, claims handling and complaint handling, and the risk sits with a French insurer regulated by the ACPR. The wording accordingly directs UK-based customers to the UK Financial Ombudsman Service and non-UK customers to La Médiation de l'Assurance in Paris.

Under the NIMBL arrangement, an FCA-authorised UK company acts as managing general agent on behalf of a French mutual insurer, with a Portuguese affiliate as insurer for some members. Complaints are handled primarily under FCA rules, with the wording noting that for certain types the insurer is ultimately responsible and telling the member which applies. Dissatisfied complainants may be entitled to refer the matter to the UK Financial Ombudsman Service — the conditional phrasing reflects that eligibility depends on complainant status, not merely on holding the policy. Separate mediation routes exist for group schemes and for members insured through the Portuguese entity.

Two points generalise from that, and no more. Access to a national ombudsman depends on where the regulated entity sits, where you are resident, and whether you count as an eligible complainant under that scheme's rules — there is no single global ombudsman for international medical insurance, and any source implying otherwise is describing one country's arrangements. And identifying the risk carrier is the prerequisite for everything else, which is the subject of who carries the risk and a check worth running at purchase, alongside insurer solvency and due diligence.

Where no ombudsman has jurisdiction, the residual routes are the insurer's home regulator and the courts — which is what wordings mean when they note that these rights are without prejudice to legal action.

What to do, and in what order

Get the decision in writing with the clause. Establish which of the three families it belongs to, because that determines whether evidence can help. If it can, commission one targeted consultant's letter rather than several general ones, and ask the insurer in advance what would change its mind. If the eight-week point passes without a final response, escalate. And read the complaints section and regulatory footnote of your own wording first — because the identity of the insurer carrying your risk determines your options, and it is rarely the name printed largest on the card.