Diagnosis for Sale: How Clinics Hoard Diseases for Cash and Wreck Global Health Statistics

Analysis of insurance data revealed a curious pattern. Almost every patient, whatever they came in with, had gastritis added to their chart. That diagnosis got the insurer to approve pricier tests and gastroprotective drugs. Then the insurer tightened its rules and started rejecting claims under code K29 (gastritis) – and a statistical miracle occurred. In 2025 the clinic had reliably reported around 1,500 cases of gastritis a month. In 2026 – exactly zero. But suddenly there appeared precisely the same number of patients with gastroesophageal reflux disease (GERD). Obviously, people could not have "recovered" from gastritis overnight and come down with GERD in unison – these are different diseases, with different causes and different treatments. The clinic had simply switched to the code that still unlocked the insurer's wallet.

Another Emirati scenario was the "dehydration epidemic." Respiratory visits (cough, cold) started getting code E86 – dehydration – tacked on en masse. That added 30% to the bill: the patient got hooked up to a saline drip. By medical standards, intravenous hydration is warranted only in cases of serious fluid loss – vomiting, bleeding, an inability to drink. In ordinary life, an adult with a mild ailment restores fluids perfectly well by simply drinking water. It is hard to believe that crowds of unconscious people suddenly descended on the GP's office demanding emergency drips. But for the paperwork, three digits were enough – E86.

In South Africa, the scheme turned out to be craftier. There they didn't churn out diagnoses in bulk; they swapped one condition for another, exploiting a subtlety of local regulation. There is a so-called PMB list – Prescribed Minimum Benefits, a catalogue of conditions the insurer is obliged to cover in full, with no co-payment from the patient. The list includes emergency care and arterial hypertension.

In practice it looked like this. A patient is admitted to hospital under an "emergency" code with a diagnosis of "gastroenteritis" (an intestinal infection). Yet he receives no treatment for gastroenteritis whatsoever: no antiemetics, no antibiotics, no drips. Instead, on day two he is put through a full examination – heart, comprehensive blood chemistry, spirometry, even a treadmill test. And on day four he is discharged with a diagnosis of "arterial hypertension." The logic is simple: hypertension is on the PMB list, so the insurer is required to cover every service rendered, with no co-payment from the insured. On paper, the patient received "treatment" for a reimbursable diagnosis; in reality, he got a check-up he would otherwise have had to pay for out of pocket.

The most obvious consequence is the unfair distribution of insurance money. Clinics that have gotten better at "drawing" diagnoses are rewarded disproportionately more than those who honestly record the real picture. But financial losses are only the tip of the iceberg. Far more destructive is the corruption of the informational foundation on which every healthcare system rests.

The chain works like this: diagnoses entered into insurance claims and invoices are aggregated at the clinic level, then passed on to national regulators, and from there into government and international statistical bulletins. Critical policy decisions rest on these numbers – national programs are launched, budgets allocated, targets set for cutting disease burden. When every tenth record in the system, or even every second one, is a fiction, decisions are inevitably built on faulty premises.

Falsified statistics do not stay within a single country's borders. They flow into international databases, including the repositories of the World Health Organization. Researchers the world over rely on these figures to build models, hunt for cause-and-effect relationships, and publish studies that then underpin global health rankings and burden-of-disease estimates.

The UAE example is especially telling here. WHO's international databases recorded an unusually high prevalence of chronic gastritis and GERD in the United Arab Emirates – an anomaly that sent epidemiologists searching for explanations in the local diet, water quality, or the population's genetic predisposition. In reality, the cause was purely an accounting artifact: waves of upcoding for K29 and K21, shifting in lockstep in response to insurers' policies, had manufactured a phantom morbidity profile. The international scientific community spent real resources studying a phenomenon that did not exist.

So garbage data in produces garbage policy out. Money is thrown to the wind while real diseases go ignored. Once distorted data enters the global rankings, it shapes the allocation of international aid, pharmaceutical companies' investment, the direction of research grants – even the tourist appeal of entire regions. A false causal link drawn from fabricated numbers can wander through academic papers and textbooks for years, accumulating authoritative citations along the way.

Breaking this vicious cycle takes a combination of measures: independent cross-audits of insurance claims, and the deployment of algorithms that detect anomalous patterns – such as the synchronized drift of codes. Such algorithms already exist, not in theory but in practice: solutions from Mains Lab make it possible to automatically track suspicious shifts in coding, compare morbidity profiles across clinics, and flag anomalies before they harden into the "new normal" of the statistics. It is precisely this kind of technological audit – continuous and data-driven – that can choke off the schemes that have been parasitizing insurance systems for years.