Countries building national health insurance are in an unusual position right now. They can watch a much older, much richer system make an expensive mistake in public, in real time, and simply design around it.
The mistake is worth understanding precisely, because it is not obvious, it is not about corruption, and it will absolutely repeat wherever health funding follows recorded data.
The design that looks sensible
Start with the problem every insurance scheme faces. If you pay providers or insurers a flat amount per enrolled person, the rational strategy becomes enrolling healthy people and avoiding sick ones. Nobody wants a system that punishes whoever cares for the frail.
So mature systems adjust for health status. The United States does this at an enormous scale through Medicare Advantage, where private insurers cover more than thirty million older Americans and receive monthly payments that rise according to how ill each member’s medical records show them to be. Someone documented with diabetes, heart failure, and kidney disease generates a higher payment than someone in good health. The intent is fairness, and the intent is correct.
Where it goes wrong
Notice what the design creates: every recorded diagnosis carries a price. Not every treatment, not every outcome. The record itself.
Over fifteen years, that quietly reshaped behaviour. American insurers built substantial operations dedicated to re-reading patients’ old medical files, looking for conditions that could be added to the record. Each addition raised the risk score, and the risk score raised the payment. Nobody built an equivalent operation to find conditions that should be taken off, because removals reduce revenue.
The consequences surfaced this spring. Federal auditors examined three insurance plans and found that between 81 and 91 percent of the high-risk diagnosis codes they sampled were not adequately supported by the medical records behind them. The commonest fault was not fabrication. It was history recorded as present tense: a stroke from years earlier coded as though current, an illness that resolved but never left the file. In March, one major insurer agreed to pay 117.7 million dollars to settle federal claims about how its records had been assembled, with investigators focused specifically on the fact that the review programme almost never worked in the other direction.
Congressional advisers estimate the accumulated overpayments across the industry in the tens of billions of dollars annually.
The design principle that fixes it
The correction now being enforced is conceptually simple and worth writing into any scheme from day one: reviews of medical documentation must run in both directions.
A credible programme adds conditions that are genuinely present but were never properly recorded, because rushed clinicians miss documentation constantly and patients deserve funding for what they actually have. And the risk adjustment solutions for health plans that regulators now treat as credible make removal an equal obligation, even though every removal costs the organisation money in the short term.
That second half is the whole test. Any review process that only ever finds errors benefiting whoever commissioned it is not measuring reality; it is manufacturing a preferred version of it. American investigators have made the one-directional pattern the central evidence in enforcement actions, which is a useful precedent for regulators anywhere.
What a new scheme should build on day one
Three things follow directly, and all of them are cheaper to build than to retrofit.
Verification arrives with the incentive, not fifteen years later. America audited late, and the gap between incentive and enforcement is exactly where the tens of billions went. A scheme that funds its audit capacity in year one never accumulates that gap.
Diagnoses need provenance. The most defensible clinical fact is one tied to a real, dated encounter between a clinician and a patient, with the provider identified and the evidence in the note. Records assembled retrospectively, however efficiently, drift toward whoever benefits from the assembling.
The direction of corrections gets measured. A regulator can ask a single devastating question of any insurer or provider: across all the coding corrections you made this year, how many reduced your payment? Nobody needs to prove intent when the ratio speaks for itself.
Health financing reform is difficult everywhere, and no country gets it entirely right. But this particular failure has now been documented at a scale, and with a price tag, that makes it genuinely avoidable for everyone building next. That is a rare gift in public policy, and it costs nothing to accept.
READ ALSO: Where Seniors Discover New Friendships and New Opportunities
