No clinician would tolerate a thermometer that reads a degree higher in one exam room than another. Yet plenty of practices run visual acuity testing with exactly that kind of drift built in – different charts, different distances, different lighting, all producing numbers that look precise but aren’t comparable. Visual acuity isn’t a fixed property you extract from a patient’s eye. It’s a result the eye and the test produce together, and if the test changes, the result changes regardless of what’s actually happening to the patient’s vision.
That distinction matters more than most clinics treat it. A patient’s “20/25” this year and “20/30” next year might reflect real progression of disease, or it might reflect a chart that was hung eighteen inches closer to the wall, or a bulb that’s dimmer than it was twelve months ago. Without standardization, you can’t tell the difference.
The Snellen chart’s built-in inconsistency
The Snellen chart, created in 1862, remains the most familiar visual acuity test worldwide. It’s also structurally inconsistent in ways that compromise the very comparisons it’s meant for.
Snellen lines don’t have equal numbers of letters. The top line might have one giant letter; a middle line might have five or six. The step size between lines isn’t geometric, either – the jump from 20/40 to 20/30 isn’t the same visual angle change as the jump from 20/30 to 20/20. (That means a “one line” improvement or decline means something different depending where on the chart it happens.) And, hey, if you heard people don’t look letter by letter but by word shape, the game’s up: One study found the maximum stroke width was a good predictor of letters of equivalent legibility, not letter size.
None of this is a huge problem if you’re doing a rough screening check. It becomes a real problem when you’re trying to track a patient’s progression over multiple visits, over different rooms, comparing results between two clinicians in the same practice etc. Letter stroke width, step size and number, and the angles subtended are all coming for you, and Snellen was never built to control for any of that.
Why logMAR and ETDRS charts exist
The logMAR chart, born out of the Early Treatment Diabetic Retinopathy Study (ETDRS) protocol, was designed to eliminate this variation. Each row contains five optotypes, typically Sloan letters, chosen because their legibility matches. Many clinics now source a professionally graded eye chart Australia from a specialized ophthalmic equipment supplier specifically to guarantee the optotype design, contrast, and calibration meet the relevant standard. The spacing between lines is based on a strict 0.1 log unit geometric progression, so each step up or down the chart represents the same proportional change in visual angle.
This makes acuity not a rough “which line can you read” test but a continuous, letter-by-letter score. Instead of jumping from one line to the next, you can score acuity to a fraction of a line by counting individual letters read correctly. That’s a real measurement, not an estimate. It’s also so accurate that it can be the primary endpoint in ophthalmic clinical trials, and the relevant authorities will demand exactly that type of repeatable, statistically rigorous result before approving the trial.
The ISO 8596 standard formalizes this. It specifies optotype design, luminance range, contrast ratio, and line progression. A chart that complies with ISO 8596 is not just “better,” it is built to a standard that ensures the results obtained using one clinic’s chart are directly comparable to another’s. That’s the whole point of standardization.
Test distance changes the number, not the vision
Visual acuity is, in essence, about visual angle. The 6/6 (or 20/20) line is constructed so that its letters subtend 5 minutes of arc at the stated test distance. Change the distance without altering the chart, and you’ve changed the visual angle each letter subtends, which changes the measured acuity, even though the patient’s actual vision hasn’t shifted an iota.
A 6-metre chart used at 5 metres will make letters smaller than intended, and a patient may read a line worse than their true acuity. Take the chart to 7 metres and the opposite happens. This is why ETDRS protocols stipulate test distance, with standardized options at 4 metres, 2 metres, and 1 metre for patients that can’t cope with the full-length test, while maintaining the logarithmic relationship between lines. Reduce the distance without adjusting the chart design, and the whole scale becomes meaningless.
In the real world, this involves knowing what the test distance is, and sticking to it. A chart hung “about six metres away” in one room and “roughly six metres” in another isn’t standardized. It’s two different tests in a lab coat.
Lighting is the variable nobody checks
Illumination gets less attention than distance, but it has just as much influence on the result. Chart luminance needs to sit within a defined photopic range, generally cited around 80 to 320 cd/m², to produce a reliable reading. Drop below that range and contrast sensitivity drops with it, meaning patients read fewer letters correctly not because their acuity has changed but because the chart is harder to see.
Glare is a related issue. A chart lit unevenly, or positioned near a window with shifting daylight, will produce different readings depending on the time of day. Clinics that have used the same wall-mounted paper chart for years, under the same fluorescent tube that’s slowly dimming, may not notice the acuity readings creeping downward, and might read that drift as a change in patient health rather than a change in equipment.
The clinical stakes: missing real disease progression
This is not a technical detail that doesn’t matter. People’s health depends on these measurements. Age-related macular degeneration, diabetic retinopathy, and glaucoma are all assessed largely by repeated acuity measurement. A one-line shift can be the difference between “stable” and “progressing” in the chart note, which is the decision point for referral, re-imaging, or treatment escalation.
Snellen overestimates acuity relative to ETDRS in AMD patients by an average of 0.12 logMAR in the better-seeing eye (Falkenstein et al., 2008 _Ophthalmology_). That’s more than a full chart line of overestimation. In a monitoring context, that’s large enough to blot out real visual loss, or to prompt inappropriate treatment based on false decline versus a previous Snellen reading. Refractions are thrown off as well: the baseline acuity directly informs the prescription, so a poorly standardized chart isn’t just randomly incorrect, it actively causes the wrong answer.
Pediatric testing needs the same rigor
Children are not always able to read letters, which is why charts use tumbling E or Lea symbols. These optotypes are meant to maintain the same logMAR spacing and visual-angle logic as letter charts. So, the measurement principle remains unchanged, even when the symbols do.
This becomes extremely important for amblyopia and strabismus management since the treatment decisions rely on repeated, comparable acuity measurements. If a clinic switches between different pediatric chart types, or varies the test distance for a squirming toddler, the resulting acuity trend becomes unreliable just when it’s most important to be solid. Amblyopia treatment depends on small, incremental improvements, and a test that can’t detect those increments reliably isn’t suitable for the task.
It’s also important here to separate acuity from contrast sensitivity since the two are often conflated. Acuity measures the smallest resolvable detail; contrast sensitivity measures how well someone distinguishes an object from its background at varying contrast levels. Both are clinically important, but they are not interchangeable, and a chart designed for one should not be assumed to work well for the other.
The examiner is part of the instrument too
The mere fact of a human doing the testing is also a problem with nearly every other test you can think of. One person’s definition of “good effort” varies from someone else’s; one tester is overly reassuring while another is gruff; one decides that a lower score is close enough; another lets the patient slowly home in on the better choice while sitting near the poster board. Were the lights dimmer at one visit than the other? Did we remember to use the pinhole?
A practical standardization protocol
Applying this in practice doesn’t need a costly overhaul, just some basic discipline around a few variables.
Use a chart that’s ISO 8596 or ETDRS in design, and actually know the test distance it was designed for, rather than just assuming the previous chart happened to use the distance your room layout allows. Measure and mark the actual test distance in each room. Standardize illumination as far as possible, with a digitally-illuminated or backlit chart providing a constant luminance which doesn’t rely on human judgement of room lighting, or a chart which is printed on materials that don’t lose their brightness over years.
Record chart type and test distance directly in the patient’s notes alongside the acuity result, so anyone reviewing the record later – including in medico-legal contexts like disability claims or surgical candidacy assessments – can verify the measurement was taken under known, defensible conditions. A solid, defensible record looks like “this patient’s visual acuity was recorded as 6/6h using an ISO 8596 chart at 6m.” Those are the sort of records we ought to be keeping, and better records mean less chance of these frustrations happening.
Standardized testing conditions also make staff training simpler, since new team members inherit a fixed protocol rather than picking up whatever habits the previous examiner happened to have.
Treat the chart like the instrument it is
Acuity testing is a measurement, and every measurement is only as good as the instrument producing it. A chart that varies by room, by distance, by lighting, or by examiner isn’t giving you a patient’s vision, it’s giving you a number shaped by whichever version of the test happened to be running that day. Fix the instrument, and the numbers finally mean what everyone assumes they already mean.
