Your workup report, number by number
Written August 2026 · Informational — candidacy, procedure choice and every planning parameter are determined by the operating surgeon
A refractive workup hands you two or three pages of numbers produced by five or six different machines, in four different units, and nobody sits down and explains them. So people photograph the report, post it online, and get an answer computed with the wrong flap thickness for the wrong procedure. This page is the decoder: what each number is, which instrument made it, what unit it comes in, and what it has to clear — plus the part almost nobody states, which is that the same cornea gives three different depth answers depending on which procedure you are being measured for. If you are flying in for surgery, the last section is the one to act on: which of these values you need in English before you board.
Why this page exists. On the forums where people bring their refractive decisions, one thread shape now recurs weekly: a patient posts their own biometry — 518 microns, ACD 3.40, ISV 21, K 40.5 to 41.5 — and asks what it means. The replies are confident and frequently wrong, usually because someone has run a depth calculation using a flap figure from a mechanical microkeratome era, or applied a LASIK formula to a SMILE plan. The numbers themselves are not secret and not especially hard; what is missing is a single place that says which machine produced each one and what it is being weighed against. That is what follows. It will not tell you whether you can have surgery — nothing written down can do that, and the final section explains precisely why.
Five machines, one report
Before the numbers, the instruments — because knowing which box produced a value tells you how much weight it carries and whether two figures on your report are even comparable.
| Instrument | What it does | The numbers it produces |
|---|---|---|
| Scheimpflug tomographer (e.g. Pentacam class) | Rotating camera builds a three-dimensional model of the cornea — front surface, back surface and thickness everywhere, not just the centre | Central and thinnest pachymetry, front and back elevation, keratometry, the shape indices, anterior chamber depth |
| Placido topographer | Reflects illuminated rings off the tear film to map the front surface curvature only | Keratometry, astigmatism magnitude and axis, front-surface irregularity |
| Optical biometer | Measures the eye along its axis with low-coherence interferometry | Axial length, anterior chamber depth, white-to-white, lens thickness |
| Specular microscope | Photographs and counts the single layer of cells on the inner face of the cornea | Endothelial cell density, coefficient of variation, hexagonality |
| Phoropter, autorefractor, aberrometer | Measure the refractive error itself — subjectively, automatically, and as a full wavefront | Manifest and cycloplegic refraction, higher-order aberrations, mesopic pupil diameter |
Instrument classes as encountered at Chinese refractive centres, researched August 2026. Different brands within a class do not always agree to the micron — which is why a value measured on one machine should not be compared against a threshold quoted for another without saying so.
The thresholds table
Every number that routinely appears on a refractive workup, with what it is weighed against. Read the last column as published guidance and common practice, not as a pass mark: the whole point of the workup is that these values are interpreted together, by someone who has examined the eye.
| Number | Unit | What it is | What it is weighed against |
|---|---|---|---|
| CCT — central corneal thickness | µm | Thickness at the corneal centre | Nothing on its own. It is the denominator of the depth budget; commonly quoted population average is in the 540–560 µm region, and figures below about 500 µm are widely described as pushing a flap-based procedure out of consideration |
| Thinnest point + its location | µm, plus x/y in mm | The thinnest pachymetry reading anywhere on the map and where it sits | Both the value and the displacement matter — a thinnest point sitting well below and outside the centre is one of the classic early ectasia patterns |
| Manifest refraction | D (sphere / cylinder / axis) | Your prescription as you subjectively accept it | Should agree closely with the cycloplegic figure; a large gap suggests accommodation is masking the true number |
| Cycloplegic refraction | D | The same measurement with focusing paralysed by drops | This is what surgeons plan from. It is the reason your appointment involves several hours and blurred near vision afterwards |
| K readings (K1, K2, Km) | D | Corneal curvature in its flattest and steepest meridians, and the mean | Very flat and very steep corneas both complicate laser treatment; a post-operative cornea that would end up unusually flat is one reason a large correction gets rerouted |
| Corneal astigmatism | D | The difference between K2 and K1, with its axis | Drives whether a toric lens is needed, and how much extra ablation depth an astigmatic treatment will cost |
| Mesopic pupil | mm | Pupil diameter in dim light | Compared with the planned optical zone — the comparison that predicts night-time glare. Covered properly on the night-vision page |
| Axial length | mm | Front-to-back length of the eye | Not a surgical threshold but a lifelong one: longer eyes carry higher retinal risk regardless of what is done to the cornea, which is why high myopes get a dilated retinal examination |
| ACD — anterior chamber depth | mm | Space between the corneal endothelium and the front of the natural lens | The gate for an implantable lens: CE mark from 2.8 mm, FDA approval from 3.0 mm. Measured from the endothelium, not the epithelium |
| WTW — white-to-white | mm | Horizontal visible corneal diameter | Feeds the manufacturer's sizing nomogram for an implantable lens. Not a pass/fail number — a sizing input |
| ECD — endothelial cell density | cells/mm² | Population of the single non-regenerating cell layer that keeps the cornea clear | Age-dependent minimums for lens implantation, published in the roughly 1,900–3,875 cells/mm² range depending on age and lens power. Also your lifetime baseline — see the cell-loss guide |
| Tear film: break-up time, Schirmer, osmolarity | s, mm, mOsm/L | How stable and how plentiful your tears are | A gate in its own right, and a distorting influence on every other measurement — an unstable film degrades topography. See the dry-eye guide |
| IOP — intraocular pressure | mmHg | Pressure inside the eye | A screening value now, and a baseline you should keep forever, because corneal surgery changes how honestly this measurement reads afterwards |
| Shape indices: ISV, IVA, KI, BAD-D | unitless | Statistical descriptions of how irregular, asymmetric and atypical your corneal shape and thickness distribution are | Deviation from a normal population. In the standard display each component flags as suspicious at roughly 1.6 standard deviations from the mean — a prompt for closer inspection, never a diagnosis by itself |
| Posterior elevation | µm | How far the back surface of your cornea rises above a best-fit reference sphere | Often the earliest place a developing ectasia shows, which is why a Placido topographer alone — which cannot see the back surface — is not considered sufficient screening |
| Higher-order aberrations (total HOA RMS, coma, spherical) | µm | Optical imperfections that spectacles cannot correct | Feed custom treatment profiles; elevated vertical coma in particular is one of the signals used in early ectasia detection |
Threshold guidance compiled from published ophthalmic sources, researched August 2026. Values and cut-offs are quoted as commonly published guidance for orientation. They are not eligibility criteria for any individual, they differ by device, jurisdiction and surgeon, and no combination of them constitutes a candidacy assessment.
The arithmetic people get wrong
Two calculations run underneath most refractive planning, and both are simple enough that patients attempt them. Doing so is not a bad instinct — understanding the shape of the trade-off is genuinely useful — as long as you know where the estimate breaks.
Residual stromal bed
What is left of the load-bearing corneal stroma after the procedure, and the traditional safety figure. Roughly:
residual bed ≈ CCT − flap thickness − ablation depth
Femtosecond flaps are commonly planned between 100 and 120 µm. Ablation depth follows, approximately, the classic relationship in which depth per dioptre equals the square of the optical zone diameter in millimetres divided by three — so roughly 12 µm per dioptre at a 6.0 mm zone, and appreciably more as the zone widens. The historic floor for what should remain is 250 µm, and 300 µm is now widely described as the more comfortable figure.
Where the estimate breaks, in the order it usually does. Modern aspheric and wavefront-guided profiles remove more tissue than the plain formula predicts, because they deliberately treat a wider blend zone. Astigmatic correction costs depth the sphere-only formula ignores. And the single commonest error on the forums is a flap figure of 160 µm or more, carried over from the mechanical microkeratome era, which can add fifty microns of imaginary tissue loss and turn a workable plan into an apparent refusal. Your surgeon plans from the laser platform's own calculation for your specific eye and profile, which is not this formula.
Percent tissue altered
PTA = (flap thickness + ablation depth) ÷ CCT
Work published from 2015 onward found this ratio to be the strongest single predictor of post-LASIK ectasia in eyes with otherwise normal topography — correlating more closely with ectasia than residual bed thickness, patient age, corneal thickness or the older ectasia risk scoring system taken individually — with 40 per cent emerging as the cut-off, and later work validating it. It is now the number most often cited when a large correction on a moderately thick cornea gets rerouted away from LASIK.
The limit that belongs with the number, and it is routinely dropped. PTA was derived for flap-based LASIK. It describes a cornea in which a lamellar flap has been cut and lifted, and it does not transfer unchanged to a procedure that does not do that. A reply that computes one PTA figure and applies it to LASIK, SMILE and TransPRK alike is making a category error even when its arithmetic is flawless — see the next section. And a figure sitting just under or just over 40 per cent settles nothing on its own: it is a population risk marker being applied to one eye, alongside a topography map that can outrank it in either direction.
Why the same cornea gives three different answers
This is the part of the report that patients most often mis-read, because the depth arithmetic everyone quotes describes only one of the four procedures on offer.
| Procedure | What it does to the depth budget | What that means for your numbers |
|---|---|---|
| Flap-based LASIK | A lamellar flap of roughly 100–120 µm is cut and lifted; the ablation happens beneath it and the flap is replaced | Both flap and ablation are spent. This is the situation the residual-bed and PTA formulas were written for — and only this one |
| SMILE | A lenticule is cut inside the intact cornea and drawn out through a small incision; the anterior cap is never lifted as a flap | The cap is not a flap, so the LASIK formula does not carry over unchanged. The relevant comparison is what the lenticule removes and what the intact anterior layers still contribute |
| TransPRK | The epithelium is removed by the laser and the surface is ablated directly — no flap, no cap | Different accounting again: no flap comes out of the budget, and the epithelium regrows over the following days. The cost is moved from tissue to recovery time |
| EVO ICL | Nothing is removed. A lens is placed in front of the natural lens | Your corneal thickness stops being the constraint entirely and a completely different set of numbers takes over: ACD, white-to-white, endothelial cell density |
Procedure mechanics as generally described, researched August 2026. Procedure selection for any individual eye is determined by the operating surgeon after examination; this table explains why the numbers differ, not which procedure anyone should have.
The practical consequence: when a number appears to rule you out, ask which procedure it rules you out of. A depth budget that fails for LASIK can leave a surface procedure open, and a cornea that closes every laser route says nothing at all about a lens-based one. The thin-cornea guide maps that fork in detail; the four-way comparison covers the trade-offs between the routes themselves.
When one number vetoes the rest
The thresholds table can give the impression that the numbers are peers, each contributing a vote. They are not. Corneal shape outranks corneal thickness, and this is the hierarchy patients find most counter-intuitive: abnormal topography is described in the literature as the single most significant identifiable risk factor for ectasia after LASIK, and a pattern indicating keratoconus or another ectatic disease is treated as an absolute contraindication however generous the pachymetry looks.
That is why a report showing 560 µm of comfortable thickness alongside a flagged asymmetry index is a harder conversation than a 505 µm cornea with a textbook-normal map. It is also why a single flagged index is not the disaster it reads as: these indices are deliberately sensitive, they flag at a modest deviation from a normal population, and the ordinary response is a closer look — repeat imaging, the family history, the other eye — rather than a refusal. What a flag most often changes is the route, not the answer.
What the report cannot tell you
Three honest limits, stated here because the rest of the page is a table of numbers and tables of numbers invite the belief that they add up to a decision.
- A number without its instrument and its date is not a number. Pachymetry by ultrasound and by Scheimpflug do not always agree; anterior chamber depth means two different distances depending on where the measurement starts. A value copied out of context and compared against a threshold published for a different device is a guess wearing a decimal point.
- Contact lenses corrupt the measurements you most rely on. A cornea still moulded by a lens produces curvature and topography figures that are not its own — and the resulting distortion usually pushes the reading toward a worse verdict than the eye deserves. Washout requirements differ by lens type and are set by the examining clinic; the dry-eye guide covers the tear-film side of the same problem.
- The thresholds are guidance, and guidance moves. Two hundred and fifty microns of residual bed was the standard for years and three hundred is now widely preferred; a cut-off that only entered the literature in 2015 now reroutes cases weekly. A page like this one dates its sources for that reason, and a threshold you read anywhere — including here — is worth checking against what your own surgeon works to.
Which numbers to obtain in English before you fly
If your surgery is a flight away, one thing changes. The hospital will repeat the entire workup on arrival — that is not negotiable and it should not be, since the plan has to be built on measurements the operating surgeon trusts. So the pre-flight set is a screening set, and its only job is to stop you boarding a plane for a procedure your own numbers already rule out. That is a low bar and a valuable one: it puts a routine local measurement between you and a wasted international trip.
The set that earns its place, each value labelled in English with its unit, its date, and the instrument that produced it:
- Cycloplegic refraction as well as your everyday prescription, per eye — the one your future plan is built on, and the one that stops existing in comparable form the moment anyone operates.
- The full tomography export — the multi-map report as a document, not a phone photograph of one screen. The screen most people photograph is the curvature map; the posterior elevation and thickness-distribution maps are the ones that carry the screening.
- Central corneal thickness, the thinnest point, and where the thinnest point sits. The location is a data point, not a footnote.
- Keratometry, flat and steep, with the astigmatic axis.
- If an implantable lens is a possibility: anterior chamber depth — stated with the reference surface it was measured from — white-to-white, and an endothelial cell count from specular microscopy.
- If you are a high myope: the axial length and a dilated retinal examination report, which is about the health of the eye rather than the eligibility of the cornea.
- An intraocular pressure reading with its date and method — the one value on this list whose usefulness is mostly in the future rather than the present.
Two of these become unrepeatable. Your pre-operative refraction and your pre-operative corneal shape stop existing the moment you are operated on — thereafter they live only in whichever file recorded them. If that file is in a hospital on another continent, in another language, the practical answer to "what was I before?" may be no answer at all. The same discipline is applied to a different reader in the occupational-medicals guide and the aviation record set: records written for the person who will have to act on them, in the language they act in. Getting the English record set out of a Chinese hospital is covered on the language and records page.
What this set is not is a substitute for the consultation. Nothing on it authorises a procedure, none of it is read in isolation by anyone competent, and the interpretation — which route your eye can take, at what parameters, or whether it can take one at all — is determined by the operating surgeon after examining you. This page exists so that when they explain it, you already know what they are pointing at.
Questions people actually ask
What corneal thickness do you need for LASIK?
There is no single pass mark, which is why a number quoted on a forum is close to meaningless without the rest of the report. Central corneal thickness matters only in combination with how much tissue your prescription requires the laser to remove and how thick a flap the surgeon plans, and the commonly published guidance is expressed as what must be left behind rather than what you start with: a residual stromal bed historically floored at 250 microns, with 300 microns now widely preferred, and a percent tissue altered figure under 40 per cent. A cornea around 500 microns can be comfortable for a small prescription and unusable for a large one, and a thick cornea with abnormal topography can still be a refusal. Thickness is one input to an arithmetic, not a verdict, and the arithmetic is the operating surgeon's.
What is percent tissue altered (PTA) and why does 40% matter?
Percent tissue altered is the share of your cornea's central thickness that a LASIK procedure disturbs: flap thickness plus planned ablation depth, divided by central corneal thickness. Research published from 2015 onward found that a PTA of 40 per cent or more was the strongest single predictor of post-operative ectasia in eyes with otherwise normal topography, correlating more closely with ectasia than residual bed thickness, patient age or corneal thickness taken alone, and later work validated that cut-off. Two honest limits belong with the number. It was derived for flap-based LASIK, so applying it unchanged to SMILE or to a surface procedure is a category error people make constantly online. And it is a population risk marker, not a personal verdict — a figure just under or just over the line does not settle anything on its own.
Can I work out my own residual stromal bed from my report?
You can estimate it, and the estimate is useful for understanding rather than for deciding. Residual stromal bed is roughly your central corneal thickness minus the flap the surgeon plans minus the depth the laser will ablate. Femtosecond flaps are commonly planned between 100 and 120 microns, and the classic Munnerlyn approximation puts ablation depth per dioptre at roughly the square of the optical zone in millimetres divided by three — about 12 microns per dioptre at a 6.0 mm zone, more at a wider one. The estimate goes wrong in three predictable ways: modern aspheric and wavefront-guided profiles remove more tissue than the plain formula predicts, older microkeratome flap figures of 160 microns and above still circulate and inflate the answer against a modern flap, and astigmatism and treatment-zone blending add depth the formula ignores. Your surgeon plans from the actual laser's own calculation, not from this arithmetic.
What does ACD mean on an ICL report, and why is 2.8 mm the number?
ACD is anterior chamber depth — the space between the inner surface of your cornea and the front of your natural lens, which is the room an implantable Collamer lens has to sit in without crowding either structure. For lens sizing it is measured from the endothelium, not from the epithelium, and reports that quote it the other way produce a number roughly half a millimetre larger that is not comparable. The published minimum differs by jurisdiction: the European CE mark allows implantation from 2.8 mm, while the US FDA approval requires at least 3.0 mm, so the same eye can be inside the criteria in one country and outside them in another. ACD never travels alone — white-to-white corneal diameter feeds the sizing nomogram and endothelial cell density has its own age-dependent minimum — and the lens choice remains the operating surgeon's after examination.
Which workup numbers should I get in English before flying to China for surgery?
Treat the pre-flight set as a screening set whose job is to stop you boarding a plane for a procedure your own numbers already rule out — not as a substitute for the workup, which the operating hospital will repeat regardless. The set that earns its place: a cycloplegic refraction as well as your everyday prescription; the full tomography export as the multi-map report rather than a phone photograph of one screen; central corneal thickness with the thinnest point and where it sits; keratometry; and, if an implantable lens is on the table, anterior chamber depth, white-to-white and an endothelial cell count from specular microscopy. High myopes should add a dilated retinal examination report and the axial length. Have every value labelled in English with its unit and the date and instrument that produced it, and stop wearing contact lenses well before the measurements are taken, because a cornea still moulded by a lens produces curvature and topography figures that are not yours.
My topography indices are flagged but my corneal thickness is fine — which one decides?
Shape usually outranks thickness, and this is the part patients find most counter-intuitive. Abnormal corneal topography is described in the literature as the single most significant identifiable risk factor for ectasia after LASIK, and a pattern that indicates keratoconus or another ectatic disease is treated as an absolute contraindication however generous the pachymetry looks. The Scheimpflug indices — index of surface variance, index of vertical asymmetry, the keratoconus index and the Belin/Ambrósio deviation value — are statistical descriptions of how far your cornea's shape and thickness distribution sit from a normal population, and a single flagged value is a prompt for a closer look rather than a diagnosis. What it usually changes is the route rather than the answer: a surface or lens-based procedure may remain open where a flap-based one does not. Which of those applies to your eye is determined by the operating surgeon after examination.
Why do SMILE, TransPRK and LASIK give different answers from the same cornea?
Because each procedure spends your corneal thickness differently, and only one of them spends it the way the familiar arithmetic assumes. Flap-based LASIK cuts a lamellar flap of roughly 100 to 120 microns and then ablates beneath it, so both the flap and the ablation come out of the depth budget — which is the situation percent tissue altered was derived to describe. SMILE creates a lenticule beneath an intact cap and removes only the lenticule, so the cap is not a cut-and-lifted flap and the LASIK formula does not transfer to it unchanged. TransPRK removes the epithelium and ablates the surface directly with no flap and no cap, and the epithelium regrows, so its depth accounting is different again. The same cornea and the same prescription therefore produce three different depth answers, which is why a forum reply that runs one calculation for all three is usually wrong even when the arithmetic itself is correct.