Thin corneas, high myopia — the options map

Procedure routing at partner refractive centers · Researched July 2026 · Published August 2026

The tissue budget, and who spends what

The candidacy guide walks through the residual-bed arithmetic in detail; here is the part that matters for this page. Every laser correction is a withdrawal from a fixed account — your central corneal thickness, typically somewhere around 540–560 microns. What makes the procedures different is their overhead: the depth they spend before correcting a single diopter.

Orientation, not surgical planning. Flap and ablation depths vary by platform and treatment zone; surgeons run these numbers on your actual scans, not on averages.

This is why "your corneas are too thin for LASIK" so often coexists with "TransPRK is fine" (that rerouting conversation — and TransPRK's cost — has its own page) — on a 500-micron cornea, not spending the ~110-micron flap is the whole difference between failing and passing the residual-bed rule. And it is why a cornea can be too thin for any laser at −9 D yet comfortably qualify at −3 D: the withdrawal scales with the prescription, roughly 12–15 microns per diopter over a standard zone.

Thin is a finding, not a diagnosis

Two people can share a 495-micron pachymetry reading and receive opposite answers, because thickness never travels alone:

• Thin and regular. Topography is symmetric, the back surface is unremarkable, the prescription is stable. Surface ablation is frequently still on the table for modest corrections, and ICL is available for larger ones. This is the common case.
• Thin with a suspicious pattern. Asymmetry, inferior steepening, or back-surface elevation — the tomographer's warning signs for keratoconus-spectrum disease. This ends the laser conversation at any thickness, because removing tissue from a cornea that may be biomechanically weak is the one outcome screening exists to prevent. ICL remains discussable in stable cases, at the surgeon's judgment, because it removes nothing.

Surgeons also increasingly think in proportions rather than raw microns — one published framing caps the percentage of corneal depth altered (flap plus ablation against total thickness) at around 40%, which catches thick corneas with huge corrections just as it catches thin corneas with average ones. The lesson for a reader is the same either way: no number you can look up about yourself settles this. The measurements that decide it — front and back topography, true pachymetry rather than an optometrist's estimate, prescription under cycloplegia — come from the workup's instruments.

That published framing has a name — percent tissue altered — and one detail that matters here: it was derived for flap-based LASIK, so applying the same figure to SMILE or a surface procedure is a category error, which is exactly what happens in most online replies. The workup decoder works the arithmetic through per procedure and lists the three ways the estimate reliably goes wrong.

Before you trust the number: two ways a thin reading is not your cornea

Everything above assumes the pachymetry figure in front of you is a property of your eye. Twice over it may not be — and both failure modes push in the direction that matters here, which is towards thin and towards irregular. Neither is exotic. Both are ordinary, documented, and undone by waiting rather than by treatment.

A cornea inside a contact lens is not the cornea a surgeon will measure

Long-term lens wear moulds the cornea, and the moulding outlasts the lens. In one published series of refractive-surgery candidates, 12% showed significant contact lens-induced warpage, and in the affected patients the mean duration of prior wear was around 21 years — a long-wearer's problem, not a beginner's. Warpage produces exactly the two findings this page is about: it pulls the topography map towards asymmetry, and it moves the curvature and thickness values a routing decision is read from.

The published resolution times are the part that rarely reaches the patient, because they are far longer than the instruction they are attached to. Time to stabilisation of refraction, keratometry and topography pattern averaged 7.8 ± 6.7 weeks, range 1 to 20 weeks, and it separates sharply by what you wore:

Published means and standard deviations from refractive-surgery screening literature, researched August 2026. Group means, not a schedule for an individual eye — your own cessation period is set by the examining clinician.

The familiar clinic rule — two weeks out of soft lenses, five out of rigid ones — comes from the same literature, but it travels with a precondition that gets dropped in transmission: those intervals were found adequate for patients whose refraction and topography maps were already within 0.5 D of earlier values. The short cessation period is validated for eyes that have been shown to be stable, not asserted for eyes that have not been looked at before. If no one holds an earlier map of your cornea to compare against, two weeks is an assumption rather than a measurement.

What that changes for someone flying in. The cessation clock runs at home, before the screening you are about to buy a plane ticket on — and for a rigid-lens wearer the published mean is longer than most people's window between deciding and travelling. So the questions worth asking about your own scan are which lenses you wore, for how many years, and how many days you had been out of them when the measurement was taken. If the answer is three days, the scan is not evidence in either direction. A cornea out of lenses too briefly can be told it is too thin and told it is too irregular — and a repeat scan some weeks later can quietly withdraw both.

Two machines, one eye, and a disagreement the size of the decision

The routing on this page turns on tens of microns. So does the gap between instruments.

Ultrasound pachymetry, Scheimpflug tomography and anterior-segment OCT do not return the same number from the same cornea, and the bias has a direction: ultrasound reads highest. One comparison reported mean differences against Scheimpflug and OCT of about 6.5 and 7.5 microns in normal eyes; another put a Scheimpflug-Placido system roughly 16 microns below ultrasound, with 95% limits of agreement running from about −26 to −6 microns. A single study's population means give the flavour of the spread: OCT ≈ 538, Scheimpflug ≈ 546, ultrasound ≈ 556 microns. Scheimpflug and OCT sit close to one another; ultrasound sits apart, and it is also the most operator-dependent of the three. The conclusion those authors draw is the one to carry away — the methods should not be treated as interchangeable.

Now put that beside the arithmetic two sections up. At roughly 12–15 microns per diopter, an instrument-to-instrument disagreement of 16 to 26 microns is the same size as one to two diopters of correction. In a decision being made at the margin, that is not a rounding error.

And the direction is the unhelpful one for precisely this page's reader. The reading most likely to be taken casually — a handheld ultrasound probe at a screening appointment — is the reading that comes out thickest. Someone told 505 microns at home can find 490 on the tomographer at the surgical centre having changed nothing but the machine, and 490 is the side of the line where the conversation changes.

A suspicious pattern is a question, and questions are answered by time

The section above ends the laser conversation for a cornea with a suspicious shape and then moves on, which is where most pages leave it. The reader is left holding a flagged index and no next step, so here is the next step.

Keratoconus and its milder relatives are not diagnosed from a single map. They are diagnosed from shape and its change over time — so a flagged index on one scan is not a verdict, it is an open question, and the instrument that answers it is a repeat scan on the same machine after an interval, read against the first.

Which carries a blunt consequence for a fly-in itinerary: you cannot resolve a suspicious topography during a two-week trip. A stability question needs two points separated by months. If your screening flags shape, the sequence that works is to get the question answered at home, on one machine, over an interval, and then travel carrying both maps — not to book a surgical trip hoping a second opinion abroad reads the same single scan more generously. A second opinion on one scan is another reading of one data point; it is not a second data point.

Two things stay open while the question does. A repeat scan may withdraw the flag altogether — see the warpage section above, since an artefact does not progress. And where a shape question settles as stable while laser reshaping stays closed anyway, the lens route removes nothing from the cornea and remains discussable at the operating surgeon's judgment.

When the flag arrives after you land

The section above assumes the flag reaches you at home, early enough to change the plan. For a large share of the people reading this page it will not — and the reason is two sections up. A shape flag comes from a tomographer: a Scheimpflug or anterior-segment OCT device that maps the front surface, the back surface and the thickness across the whole cornea. What an ordinary high-street screening produces is a handheld ultrasound probe and an autorefractor, which return a thickness and a prescription and no shape at all. A screening that cannot see shape cannot flag it. So for many fly-in patients the first tomographer they have ever sat in front of is the partner hospital's, on day one — which makes “have the question answered at home first” sound advice for someone who already holds a home tomography map, and no advice at all for the person it actually happens to.

Three things follow, and none of them is in the leaflet.

An interval is not a formality — it is the time a slow change needs to outgrow the machine's own scatter

Published work on how progression is judged puts a number on why next week will not do. International consensus treats progression as a pattern rather than a single moving figure: it requires consistent change across at least two of three domains — anterior corneal curvature, posterior curvature, and corneal thinning. The threshold in widest use, and the inclusion criterion in the trials behind the FDA-approved crosslinking platform, is a ≥1.0 D change in maximum keratometry (Kmax) observed within a window of up to 24 months. Repeatability studies on Pentacam HR devices put the point at which sequential measurements likely represent real change at roughly 1.08 D of Kmax and 1.54 D of steep keratometry.

Read those two figures next to each other and the reason for the wait is arithmetic, not caution: the detection threshold and the instrument's noise floor are the same size. Below about a diopter, a difference between two scans is indistinguishable from the device disagreeing with itself. A rescan seven days later cannot answer a stability question on any machine, in any country, at any clinic, however good — not because nobody will do it, but because the quantity being looked for has not had time to become larger than the scatter it must be seen through. Longitudinal series that track these corneas properly run over years; one recent progression cohort followed patients for a median of about 25 months. The interval that applies to you is set by the clinician examining you, and it is counted in months.

“The same machine” is doing more work than it looks

The previous section's requirement of a repeat scan on the same machine is not a convenience. Indices from different tomographers are not interchangeable, and the disagreement is not a constant you can subtract — the page has already shown the same effect on plain thickness, where three instruments return three different numbers from one cornea. Applied to shape, it means a map flagged in Guangzhou and a map taken at home on a different device are not two points on a line. They are two unrelated single points, and a progression question needs a line.

Which produces a conclusion almost nobody expects, and it is the practical heart of this section: being flagged in China puts your baseline on a machine in China. If a return trip is realistic for you, the same hospital is where the pair completes, and that is a genuine argument for going back rather than starting again — the months you have already spent waiting only count on the device that took the first scan. If a return trip is not realistic, then the honest structure is to open a fresh baseline at home on one instrument and treat the Chinese map as corroboration rather than as point one. That is a defensible choice, but it costs you the elapsed months, and it is much better made deliberately on the flight home than discovered a year later at a second consultation.

What to carry out of the building, so that the comparison is possible at all

The default handout is a printed colour map, and a printed colour map is the one form of the result that progression software cannot re-read. Ask, before you leave the department:

A shape flag travels as a dataset, not as a picture. Assembled from the diagnostic handover practice at partner refractive centers, researched September 2026; what any individual department can release is decided by that department.

The itinerary consequence, stated plainly

Look back at the options map on this page. Every other row reroutes you to a different procedure, this week — thin-but-regular sends you to surface ablation, high myopia sends you to a lens. The suspicious-topography row is the only one that reroutes you to a different year, and it is the only row on which nothing at all is scheduled. That asymmetry is worth building the trip around: where the hospital allows it, keep the diagnostic block and the surgical week separable rather than booked as one indivisible fortnight, and avoid committing to anything downstream that only makes sense if an operation happens.

What a flag does not close: your glasses and lenses carry on exactly as before, since nothing has been done to your eye and nothing has been found that was not already true of it yesterday. A lens-based route removes no corneal tissue and stays discussable in stable cases at the operating surgeon's judgment. And the flag itself may not survive the second scan — an artefact does not progress.

High myopia: where the laser range actually ends

Laser correction doesn't hit a wall at a marketing number; it fades out as the arithmetic worsens. Somewhere past −8 D, three things degrade together: the tissue withdrawal gets large, the optical-quality cost of flattening that much cornea grows (night-vision symptoms scale with correction size), and the margin for a future enhancement disappears. By −10 D, most surgeons have long since moved the conversation to the EVO ICL, which is rated to about −18 D (with toric versions for astigmatism) and performs at its best precisely where lasers perform at their worst — the higher the correction, the more the optical quality of an implanted lens pulls ahead of a heavily reshaped cornea. One trade rides along with that recommendation and deserves to be stated where the routing happens: comparative studies tend to report more night halos after ICL than after the laser procedures — usually a discrete ring most patients adapt to within months. What that ring is, why your dim-light pupil diameter predicts it, and when it stops being normal are covered in our night-vision guide.

One thing high myopes should expect at the workup that lower prescriptions skip: a careful dilated look at the retina. An eye at −12 D is a longer eye, and the stretch carries retinal risks — thinning, lattice degeneration, tear-prone areas — that matter more to your future sight than the choice of refractive procedure. Partner hospitals examine, and where needed treat, the retina before any correction is scheduled. If you take one thing from this section: at high prescriptions, the refractive surgery is the second most important thing the workup looks at.

And the reason that exam matters long after your trip is the part most price and procedure pages leave out: none of these procedures shortens the eye. Correcting the refraction does nothing to axial length, so the retinal risk that comes with a long eye is exactly what it was — which is why the findings from that dilated exam are a baseline you should take home in English rather than leave in a hospital record. We have set out what the evidence actually says, and what it genuinely does not settle, in the risk that survives the surgery.

The third asymmetry: when the two eyes are a long way apart

This page has been about two findings that limit what a cornea can afford. A third situation arrives at the same routing map from a different direction and is badly served everywhere: a large difference between the eyes — say −10.50 D in one and −1.25 D in the other. Anisometropia, in the clinic’s vocabulary. People in that position usually describe it in terms of the strong eye, and what comes back is a conversation about whether that eye is correctable. That is the wrong quantity.

What decides how the two eyes get on is image size, not dioptres. A spectacle lens sits roughly twelve to fourteen millimetres in front of the eye, and at that distance a strong minus lens shrinks the image it delivers. The working rule of thumb is about 1% of image-size difference for every dioptre of anisometropia. Symptoms are commonly described as beginning around 0.75–1%, becoming definite between 1% and 3%, and by roughly 5% normal binocular function is generally no longer found. A nine-dioptre gap is a long way past that last line. So the familiar complaint — glasses that are optically correct and still unwearable, with pulling, headache and no usable depth — is not evidence the prescription is wrong. It is aniseikonia: two images of different sizes arriving at a brain that has to fuse them and cannot.

There is a classical objection worth answering, because it is why some people are told their glasses ought to be fine. Knapp’s law predicts that where the difference is axial — one eye longer, rather than one cornea steeper — a spectacle lens at the eye’s anterior focus returns equal retinal image sizes, so glasses should be the comfortable correction. Measured clinically, that prediction does not hold: spectacle correction has been found to produce significantly more aniseikonia than contact lens correction even in axial anisometropia, with retinal stretching in the longer eye, and the micropsia that follows from it, the usual explanation offered. The practical reading is that the axial-versus-refractive distinction is not a reason to expect glasses to be tolerable at a large difference, and that correcting closer to the eye’s own optics helps in either case.

Which is where this connects to the map below. A contact lens, a reshaped cornea and a lens implanted inside the eye all correct at or inside the eye’s own plane, so all three largely remove the spectacle magnification that glasses cannot avoid. On the numbers, one published adult series carrying a mean myopic anisometropia of −10.70 ± 3.02 D reported it at +0.09 ± 1.67 D one month after phakic lens implantation, with contrast sensitivity in the weaker eye improved; related work reports partial recovery of stereopsis in some patients. Read that as a mechanism rather than a promise. An eye that has been the weaker eye since childhood carries an acuity ceiling set by that history, which no correction moves — and whether any of this applies to your eyes is determined by the operating surgeon after examination.

The itinerary consequence is specific to travelling for this, and it is the part worth carrying into your first email. Where the two eyes are far apart and the plan is an implanted lens, ask before you book how the two eyes will be scheduled. If the second lens follows the first after an interval, you spend that interval in a transient anisometropia — published binocular-function work on phakic lenses flags the delay between the first and second implant as exactly the window in which binocular vision can decompensate — and the length of that window is a scheduling decision, which makes it one of the few clinical variables the shape of your trip can actually influence. Our trip-length guide sets out the standard itineraries; this is the question that decides which of them fits you.

The options map

With every caveat above — this is orientation for reading your own situation, not a decision tool — here is how the routing tends to run at partner centers:

Routing patterns as practiced at partner refractive centers, researched July 2026. Every cell is a tendency, not a rule; the operating surgeon decides from your measurements.

What the alternatives cost

The honest irony of thin-cornea routing is that it usually points at the cheapest procedure on our list: TransPRK runs $1,000–1,500 both eyes at partner-hospital standard rates (researched July 2026), below Femto-LASIK's $1,200–1,800. The high-myopia route points the other way — EVO ICL at $3,800–4,800 both eyes, where the custom-manufactured STAAR lens, not surgical time, sets the floor. Both figures are the same dated standard rates shown on the homepage table, and both carry the usual caveat: the written quote is confirmed after diagnostics, at the price of the procedure you're actually matched to. TransPRK's trade-off is never price — it's the slower surface-healing recovery, which is the main reason people with a free choice often pay more for SMILE or LASIK.

The reroute is a budget event too, and it runs in one direction

This page's premise is that people arriving with either of these two findings usually leave the workup with a different procedure than the one they came for. Everywhere else on the internet that is a clinical statement. For someone who has already bought a plane ticket it is also a financial one, and the figures are sitting in the paragraph above: TransPRK at $1,000–1,500 both eyes against EVO ICL at $3,800–4,800 both eyes (partner-hospital standard rates, researched July 2026) is a roughly threefold spread — and the two findings on this page can route you to either end of it depending on where the prescription sits.

So the number to settle before you fly is not the price of the procedure you expect. It is the price of the procedures you could be rerouted to. Ask for the written quote to name every route your own measurements leave open, each priced, rather than the single one your enquiry happened to mention. The quote is confirmed after diagnostics regardless, so asking for it conditionally costs nothing and removes the only surprise on this trip that is genuinely expensive.

One unglamorous logistical point belongs beside it. An implanted lens is sized to your own eye and ordered for it, so a reroute to ICL is not a same-afternoon substitution for a laser slot the way a reroute between two laser procedures can be. If that reroute is a live possibility for you — and on a thin or strongly myopic cornea it usually is — then the measurements ICL sizing depends on (anterior-chamber depth, white-to-white corneal diameter, an endothelial cell count from specular microscopy) are worth having in hand before you travel rather than discovering on arrival that they still have to be taken. The workup decoder lists that set in full, with the units each value should arrive in.

Questions people actually ask

Keep reading