Assessment, referral and lens power calculation
Deciding to operate on a cataract is a conversation about how the person sees and lives, not a number on a chart. Once surgery is chosen, the refractive result depends on measuring the eye and choosing the right implant power. This lesson covers NICE NG77 on referral, the preoperative assessment, biometry and the logic of lens power formulae.
From symptoms to referral
Most cataracts in the UK are found by community optometrists. A GP who suspects cataract should send the person to an optometrist, who measures acuity, confirms the lens opacity and looks for other causes of poor vision. The optometrist may then refer directly to the hospital eye service, depending on local arrangements.
NICE NG77 bases the decision to refer on a discussion with the person. It is not triggered by a test result.
- How the cataract affects their vision and quality of life.
- Whether one or both eyes are affected.
- What surgery involves, including its risks and benefits.
- How their quality of life may be affected if they choose not to have surgery.
- Whether they want surgery.
NICE then states plainly: do not restrict access to cataract surgery on the basis of visual acuity. The reason is that Snellen acuity is measured with high-contrast letters in a dim room and misses glare and loss of contrast. A person with a posterior subcapsular cataract may read 6/9 in clinic and be unable to drive at night. CKS notes that some local commissioning policies still set acuity thresholds, which is not in line with NICE.
Driving is a common reason to operate. All drivers must be able to read a number plate at 20 metres and have an acuity of at least 6/12 with both eyes open, wearing glasses or contact lenses if needed. Glare from a cataract can cause someone to fail the number plate test even when their clinic acuity looks adequate.
The preoperative assessment
The assessment has three aims: confirm that the cataract explains the symptoms, find anything that makes surgery riskier or the outcome poorer, and gather the measurements for the implant.
In the history, ask about medicines and practicalities. Alpha-blockers such as tamsulosin must be recorded because they cause intraoperative floppy iris syndrome. Note anticoagulants, diabetes control and previous corneal laser refractive surgery. Check that the person can lie flat and still, and can manage drops afterwards or has someone to help.
On examination the surgeon looks beyond the lens. A relative afferent pupillary defect is not caused by cataract and indicates retinal or optic nerve disease that will limit the result. The corneal endothelium is inspected, since an eye with Fuchs endothelial dystrophy is at risk of corneal decompensation. A pupil that dilates poorly, pseudoexfoliation and a wobbling lens all warn of a difficult operation. The fundus is examined for macular degeneration, diabetic retinopathy and glaucoma. If the cataract is too dense to see through, a B-scan ultrasound excludes retinal detachment.
These findings shape the consent discussion. In eyes with no other disease, about 95% reach a corrected acuity of 6/12 or better after surgery. Coexisting eye disease lowers that figure, and the person should hear this beforehand.
Biometry: measuring the eye
Removing the lens takes a large amount of focusing power out of the eye. The implanted intraocular lens (IOL) must replace it with exactly the right power for that eye. Biometry is the set of measurements used to work this out. Two matter most.
Axial length is the distance from the front of the cornea to the retina. It is the most sensitive measurement: an error of 1 mm changes the required lens power by about 2.5 to 3 dioptres. Keratometry measures the curvature of the central cornea in its two principal meridians (K1 and K2) and converts it to corneal power. The cornea provides about two thirds of the eye's focusing power. An error of 1 dioptre in corneal power produces an error of about 1 dioptre in the lens chosen. The difference between K1 and K2 is the corneal astigmatism.
NICE recommends optical biometry to measure axial length. The machine shines infrared light into the eye and uses interference between the reflections from the cornea and the retina to measure the distance between them, without touching the eye. Ultrasound biometry (A-scan) is used when optical biometry is not possible or does not give an accurate reading. The contact form of A-scan presses on the cornea, which can shorten the measured length.
| Optical biometry | Ultrasound A-scan | |
|---|---|---|
| Principle | Interference of reflected infrared light | Reflected sound, shown as spikes |
| Contact with the eye | None | Probe on the cornea, or a water bath |
| Main source of error | Cannot measure if light does not get through | Corneal compression and operator skill |
| Place under NICE NG77 | First choice | When optical biometry fails or is inaccurate |
NICE also advises considering corneal topography, a map of curvature across the whole cornea, for people with unusually flat, steep or irregular corneas, significant astigmatism or previous corneal refractive surgery.
From measurements to lens power
The logic of every formula is the same. A long eye needs a weaker implant, because the retina is further back. A steep, powerful cornea also needs a weaker implant, because the cornea is already doing more of the focusing. The simplest formula, SRK, states this directly.
P = A - 2.5L - 0.9KP is the IOL power for emmetropia in dioptres, L the axial length in mm, K the average corneal power in dioptres and A the A-constant of the lens model.Worked example: for a lens with an A-constant of 118.4, an axial length of 23.5 mm and an average K of 43.5 D, P = 118.4 - (2.5 x 23.5) - (0.9 x 43.5) = 118.4 - 58.75 - 39.15 = 20.5 D. If the axial length had been wrongly measured as 22.5 mm, the formula would give 23.0 D. That lens is 2.5 D too strong and leaves the patient unexpectedly myopic.
The A-constant is a number supplied by the manufacturer for each lens design. It stands in for the one thing that cannot be measured before surgery: where the implant will finally sit in the eye, called the effective lens position. A lens that sits further forward has more effect. Predicting this position is the main remaining source of error, and it is what successive generations of formulae have tried to improve. Surgeons can also refine the constant from their own results.
The original SRK formula is a regression line that works for average eyes and fails in very short or very long ones. It has been superseded, but it is the clearest statement of the principle. Modern formulae are chosen by axial length. For eyes without previous corneal refractive surgery, NICE recommends the following.
| Axial length | Formula |
|---|---|
| Less than 22.00 mm (short eye) | Haigis or Hoffer Q |
| 22.00 to 26.00 mm | Barrett Universal II if built into the biometer, otherwise SRK/T |
| More than 26.00 mm (long eye) | Haigis or SRK/T |
Target refraction and difficult eyes
The formula gives the power for any chosen outcome, so the surgeon and patient must agree a target refraction. Most people choose emmetropia: clear distance vision without glasses, with reading glasses for near, because a standard monofocal implant cannot change focus. Some lifelong myopes prefer to stay slightly myopic so that they can read unaided. Monovision sets one eye for distance and the other for near. NICE recommends offering it to people who already have monovision or anisometropia and want to keep it. NICE requires the choice to be discussed and recorded at the preoperative assessment.
Eyes that have had corneal laser refractive surgery are a known trap. The laser changes the front surface of the cornea but not the back, so standard keratometry misjudges true corneal power and standard formulae pick the wrong lens. NICE says to adjust for the altered relationship between the two corneal surfaces and to warn these patients that the outcome is harder to predict.
When the second eye is done, the result in the first eye is useful information. NICE suggests considering using 50% of the first eye's prediction error to adjust the lens power for the second.
Risk stratification, the second eye and the right lens
NICE advises considering a validated risk stratification algorithm to identify people at higher risk of complications, and explaining the result to the person. Features that raise the risk of a complication such as posterior capsule rupture include advanced age, a very dense or white cataract, pseudoexfoliation, a small pupil, a shallow anterior chamber, high myopia and previous vitrectomy. High-risk cases, and people with only one seeing eye, should have an experienced surgeon or a closely supervised trainee. People likely to develop a dense cataract should be told that delay makes surgery riskier.
Second-eye surgery is offered using the same criteria as the first eye. There is no rule that the second eye must wait or be worse. Operating on both eyes gives further gains, particularly in depth perception. NICE says to consider operating on both eyes at the same sitting for people at low risk of complications, or for those who need a general anaesthetic that carries added risk. The trade-off is that the first eye's result cannot inform the second, and a complication could affect both eyes.
Implanting the wrong lens is classed as a never event in the NHS. NICE sets out checks to prevent it: confirm identity before biometry, transfer results electronically or fix the printout in the notes and never copy them by hand, have two team members check formula, calculation and constant, and have only one lens in theatre, matching the plan, before the anaesthetic is given.
Summary
- NICE NG77: refer on the basis of a discussion about symptoms, quality of life, risks, benefits and the person's wishes.
- Do not restrict access to cataract surgery on the basis of visual acuity.
- Preoperative assessment looks for alpha-blocker use, corneal endothelial disease, poor pupil dilation, pseudoexfoliation and coexisting retinal or optic nerve disease.
- Biometry rests on axial length and keratometry; a 1 mm axial length error changes IOL power by about 2.5 to 3 D.
- Optical biometry is first choice; ultrasound is used when it fails.
- SRK formula: P = A - 2.5L - 0.9K, so longer eyes and steeper corneas need weaker lenses.
- The A-constant reflects the lens design and its expected position in the eye, which cannot be measured beforehand.
- NICE: Haigis or Hoffer Q for short eyes, Barrett Universal II or SRK/T for average eyes, Haigis or SRK/T for long eyes.
- Previous corneal laser refractive surgery makes standard keratometry and formulae unreliable.
- Second-eye surgery uses the same criteria as the first, and wrong lens implantation is a never event prevented by systematic checks.
Check yourself
Quick recall on this lesson. These are easier than the question bank.
Written for revision. It is not a source of clinical advice, and guidance changes.