A clean kidney loses about three-quarters of a millilitre of filtration a year after the age of forty. You will never feel it. A kidney being ground down by both diabetes and hypertension loses about 3.2 millilitres a year, three to four times faster, and you will not feel that either, not for years, because the kidney has no nerve endings for the loss and the body compensates until it cannot. By the time anything is felt, the number is already low. The number is eGFR, the estimated glomerular filtration rate, and it is the one figure in this piece that runs on a clock you cannot reset.
This is the diagnosis the relocation brochures never cost, because it does not happen at arrival and it does not photograph. It happens late, on a schedule set by the metabolic profile the move quietly worsens, and when it arrives it does something almost nothing else in medicine does: it converts, on a fixed date, into an indefinite self-pay bill that no expat policy will renew into. Most acute events are a bill and then they are over. End-stage renal disease is a bill that is never over. It is metered three times a week, in a chair, until a transplant or a death stops it.
This piece is the model the brochures will not draw. It takes the slope of the decline, converts it into the year the kidney crosses the dialysis line, overlays the annual self-pay cost on a drawdown pot, and produces a single number: the eGFR at which the money has a defined expiry date.
The pipeline the move feeds
Diabetes and hypertension cause up to 80 percent of end-stage renal disease worldwide. In the United States they account for about 76 percent of newly diagnosed cases: diabetes 46 percent, hypertension 29 percent. Globally, 4.59 million people were on kidney-failure replacement therapy in 2023. These two conditions are the funnel, and they are the two conditions the cheap-living expat life is built to produce.
State the mechanism honestly, because the data does not let you state it as a measured expat rate. No published study has measured diabetes or hypertension prevalence in the Western retiree-expat cohort in Southeast Asia. What is measured is the input. Excessive alcohol is associated with every component of metabolic syndrome: raised blood pressure, dyslipidaemia, type-2 diabetes, central obesity. The retiree life that the survivorship-curated channels sell is a near-perfect generator of exactly the two upstream conditions that feed the renal funnel: cheap restaurant food eaten three times a day, beer that costs less than water, the imposed routine of a working life removed, the long unstructured afternoon. The slope is a transferred mechanism, not an expat measurement. But the inputs all push the same way, and they push harder abroad than at home.
So treat the diabetic-plus-hypertensive slope as the working assumption for this cohort, and flag the proxy every time it carries weight. It carries weight in the next section, where it becomes a clock.
The slope
Here is the decline, by cause, from a Qatar CKD cohort of 1,020 patients followed for a mean of 9.6 years. It is the cleanest stratified set of annual decline rates in the literature, and the population is a general CKD cohort in the Gulf, not Western expats — read it as the shape and the magnitude of the risk, not a personal forecast.
| Primary driver | Annual eGFR loss (mL/min/1.73m²/yr) |
|---|---|
| Primary driver Healthy aging baseline (after 40) | Annual eGFR loss (mL/min/1.73m²/yr) −0.75 to −1.0 |
| Primary driver Hypertension alone | Annual eGFR loss (mL/min/1.73m²/yr) −2.2 ± 1.8 |
| Primary driver Diabetes alone | Annual eGFR loss (mL/min/1.73m²/yr) −2.4 ± 2.0 |
| Primary driver Diabetes AND hypertension | Annual eGFR loss (mL/min/1.73m²/yr) −3.2 ± 4.3 |
| Primary driver Highest albuminuria grade (A3) | Annual eGFR loss (mL/min/1.73m²/yr) −3.2 ± 2.4 |
| Primary driver Not on ACEI/ARB therapy | Annual eGFR loss (mL/min/1.73m²/yr) −3.3 ± 3.7 |
| Primary driver HbA1c over 10% | Annual eGFR loss (mL/min/1.73m²/yr) −4.1 ± 6.2 |
Source: Hamdi et al., long-term CKD progression analysis, Qatar (PMC9713692); healthy-aging baseline from Baltimore Longitudinal Study via PMC4291282 · checked 2026-06-13
The diabetic-plus-hypertensive slope of −3.2 sits at the upper-central band, and it is corroborated from other directions. A type-2 diabetic cohort split decline by albuminuria: 1.5 with normal albumin, 2.9 with microalbuminuria, 7.1 with macroalbuminuria. A Palestinian three-year prospective cohort of type-2 diabetics put the mean at 4.2. An Italian cohort of 105,163 newly-diagnosed diabetics found 12.9 percent losing function faster than 5 a year, the threshold the literature calls rapid decline. The honest band for the cohort this site is about is −2.4 to −4.2 a year, central −3.2, with a long upper tail for the macroalbuminuric and the poorly controlled.
A slope is not a diagnosis. It is a clock. Set the dialysis line at an eGFR of 10 (stage-5 ESKD begins at under 15, and symptomatic initiation typically falls near 10) and the arithmetic is one division. Years to dialysis equals starting eGFR minus 10, divided by the annual loss. That is the whole of stage one of the model. It is also the calculation no relocation channel has ever shown a single viewer.
The annual bill
Once the line is crossed, the bill begins, and it does not stop. Price it at two levels, because the gap between them is where the foreigner lives.
In the public systems, dialysis is modelled at roughly nine thousand dollars a year. The Thai economic evaluation by Assanatham and colleagues put annual direct medical cost without complications at US$9,079 for hemodialysis and US$7,955 for peritoneal dialysis, rising to US$12,121 and US$9,221 once chronic complications set in; lifetime societal cost of an HD-first pathway reached US$73,311 from a mean start age of 55.7. The Philippine cost-of-illness analysis put chronic hemodialysis at PHP 560,383 (US$9,852) a year and peritoneal dialysis at PHP 401,261 (US$7,055). These are system costs. They are not the foreigner’s bill.
The foreigner’s bill is the private one, because the uninsured Western expat is not admitted to the public ward where those costs are realised. He goes to the private hospital. What the private hospital charges him is the part this page cannot yet answer honestly, and the reason is worth more than the number would have been.
The premium Bangkok hospitals — Bumrungrad, Bangkok Hospital, Samitivej — publish no dialysis price at all, and the medical-tourism aggregators that quote figures on their behalf disagree with each other by a factor of four. What can be verified is one tariff on a hospital’s own site. Ramkhamhaeng Hospital 2, a Bangkok private hospital, sells an ordinary session for THB 2,500 — doctor’s fee, nursing and hospital service included — plus THB 900 for the dialyser when it is changed. At 156 sessions that is THB 390,000, about US$11,800 a year. The same aggregators put Thailand at US$300 to US$500 a session, THB 10,200 to 17,000, four times what the hospital itself charges.
The gap does not track hospital tier and it does not track province. It tracks which door the patient comes through: a Thai-language retail page priced for domestic self-pay, or a foreigner-facing package priced for someone who flew in. One verified tariff is not enough to publish that as a finding. So the figures below are held at the higher, aggregator-derived level rather than revised down onto a single price — which means the model that follows may overstate the bill by something near a factor of two. That is said here rather than left for a reader to find. In the Philippines the private session runs PHP 2,500 to 5,000 cash.
The Filipino citizen has a backstop the foreigner does not. PhilHealth’s hemodialysis package climbed from PHP 2,600 a session in 2023 to PHP 4,000 in July 2024 to PHP 6,350 by October 2024, covering up to 156 sessions, close to a million pesos a year. A foreigner without PhilHealth membership watches that backstop apply to the patient in the next chair and pays the private rate himself.
Why no policy renews
This is the property that makes ESRD different from every other line item on this site, and it is worth stating without hedging.
The condition that creates the lifelong bill is the same condition that makes you permanently uninsurable. That is not a coincidence; it is the mechanism of the trap.
Individual expat health plans are medically underwritten. They will not cover a pre-existing condition, or they cover it only after a treatment-free moratorium of commonly two years. A declared history of chronic kidney disease, dialysis, or transplant gives the underwriter three options (a premium loading, a permanent exclusion of anything renal, or a decline of the application), and for established ESRD the answer is the exclusion or the decline. Routine maintenance dialysis is classed as ongoing care, which travel and many expat policies exclude unless it is an acute, unexpected onset. A man already in the chair three times a week is not an acute, unexpected onset. He is a known, indefinite liability, and the market for known indefinite liabilities priced at the individual level does not exist.
Even the lucky path is capped. Develop ESRD while already insured and the policy may pay, up to its chronic-illness annual or lifetime maximum. Dialysis is indefinite. The maximum is finite. Once the limit is reached, the policyholder is responsible for all future costs. So the insured patient is not safe; he is on a timer. The uninsured patient cannot buy in. There is no door. This is the same renewal cliff this site documents at the insurance age-out and the pre-existing-condition cliff at renewal, in its most absolute form: with ESRD there is no insurer of last resort, no high-deductible fallback, no specialist loading high enough to clear the underwriting desk. The condition self-selects you out of the only market that could have hedged it.
The time-to-ruin model
Put the two stages together. Stage one turned the eGFR slope into a year. Stage two turns a drawdown pot into a runway against the annual bill: pot divided by annual cost equals the years the money lasts once dialysis starts. The output is the combined expiry — the year the kidney crosses the line, plus the years the pot then buys, equals the year both the kidney and the capital are gone.
Run it for three entry points, all at the central diabetic-plus-hypertensive slope, all against an illustrative US$200,000 standing pot at a central US$23,000-a-year private self-pay bill. The slope is the Qatar/diabetic-cohort transfer, not a measured expat rate; the cost is the bottom of the 2026 private in-center band; the pot is illustrative. Change any input and the cell moves — that is the point of showing the work.
| Entry stage → | Years to dialysis | Self-pay years the pot buys | Total runway |
|---|---|---|---|
| Stage 3a (eGFR 58) | Years to dialysis ~15 yrs | Self-pay years the pot buys ~9 yrs | Total runway ~24 yrs |
| Stage 3b (eGFR 38) | Years to dialysis ~9 yrs | Self-pay years the pot buys ~9 yrs | Total runway ~17 yrs |
| Stage 4 (eGFR 25) | Years to dialysis ~5 yrs | Self-pay years the pot buys ~9 yrs | Total runway ~13 yrs |
Source: Model: (eGFR − 10) ÷ 3.2 for years-to-dialysis; US$200k ÷ US$23k/yr for self-pay runway. Slope from PMC9713692 (proxy cohort); cost band from Assanatham 2022 / Journal of Medical Economics 2025 / private-hospital self-pay re-verified 2026-08-21 · checked 2026-08-21
The totals look almost survivable on paper, which is exactly the deception the model exists to remove. Three things collapse them.
The first is the slope itself. Move from the central −3.2 to the macroalbuminuric −5, and stage-4 entry reaches dialysis in three years instead of five. The poorly-controlled diabetic is not on the same clock as the average; he is on a faster one, and he is over-represented in precisely the cohort that drank its way into the funnel.
The second is the cost, and it is the input this model is least sure of. The US$23,000 central figure is derived from the foreigner-facing quotes, not from the one tariff a Thai hospital publishes itself; read against that tariff it is roughly double. It is held high deliberately, because a single verified price is not a band. Push it the other way instead — a premium-hospital course at the top of the aggregator range, US$78,000 a year — and the runway falls from about nine years to under three. The honest statement is that the runway is somewhere between three years and twenty, that nobody selling the move will tell you which, and that the bill rises every year with medical inflation while the pot, drawn down, does not.
The third is the single transaction that ends it all at once. A dialysis patient who needs to get home is rarely stable enough to fly without dialysis arranged at both ends and a medical escort. The ICU-equipped air ambulance from Bangkok or Manila runs US$120,000 to 180,000, paid by no standard cover. One medevac removes five to eight years of self-pay runway in a day.
Source: Model scenarios s4 (benign) / s2 (central) / s5 (adverse), sidecar the-dialysis-trap-end-stage-renal-self-pay; slope PMC9713692 (proxy cohort); cost band Assanatham 2022 / Journal of Medical Economics 2025 / private self-pay; medevac Travel Care Air · checked 2026-08-21
Read the adverse path. The pot is intact for the first five years, because at stage 3b the patient feels completely well, and then it falls off a cliff: dialysis by year six on the rapid slope, premium private billing, and a single medevac that empties what is left by year eight. The benign path is the same eGFR on the slow end of the sourced band, −2.4 a year, billed at the bottom of the private range: it delays the chair by three years and the zero by three, and it still reaches zero. Every line does. A finite pot against an indefinite bill has only one shape. A man whose eGFR reads 38 at sixty-eight has no symptom to warn him which of these three lines he is on. The number on the lab report is the only warning the body provides, and the relocation channels have trained their audience to read every other number except that one.
The go-home fallacy fails twice
The reflex answer to all of this is the one this site has costed before: I will just go home if it goes wrong. For ESRD it fails on both ends of the flight.
It fails on the way out, because the flight itself is the medevac line above — a six-figure transaction the dying-kidney patient is least able to fund and least able to physically tolerate. And it fails on arrival, because the home system the plan assumes is waiting may no longer be his. A Briton who deregistered to chase the lower cost of living loses ordinary-residence NHS access and does not simply walk back into free dialysis; a returning American faces Medicare enrolment timing and the structure around it. The flight home buys a system he may have to re-qualify for from a dialysis chair. The full anatomy of why the fallback is not free sits in the go-home fallacy; ESRD is the case where it is most expensive and least optional.
There is one genuine exit, and it is a hard one. A kidney transplant ends the recurring dialysis bill: an acute cost of about US$34,500 in the Philippines, then about US$8,100 a year for immunosuppression and monitoring. Over a long horizon the transplant is better arithmetic than US$23,000 a year of dialysis forever. But it requires a matched organ, a surgical candidacy a frail elderly diabetic may not have, and a lifetime of drugs the same uninsurable patient self-funds. It trades a recurring bill for a lump and a smaller recurring bill. It is an exit from the dialysis chair, not from the self-pay condition.
The arithmetic no one runs
The kidney gives no warning the brochures teach you to read. It declines on a number most people never check, at a rate the expat life accelerates, toward a line that begins a bill no policy will renew into. The arithmetic is not hidden. It is one division for the years to the line and one for the years the money lasts after it, and it is never run, because the people selling the move are paid at arrival and the bill arrives at the end.
Find the eGFR number on your last blood panel. Subtract ten. Divide by three. That is roughly how many years you have before the chair, if the slope is the cohort’s. Then divide your standing capital by twenty-three thousand — and know that the one Thai hospital price this page could verify at source would halve that divisor and double your answer. That is how many years the money lasts after you sit down in it. The sum of those two numbers is the year, and it is closer than the brochure ever implied, and it gets closer every year you do not look at the first number.