6.3 Manual & Automated Keratometry: Technique, Readings and Interpretation

What keratometry actually measures

A keratometer treats the anterior cornea as a convex mirror and measures the size of the reflected image of an illuminated mire of known size. From image size it derives the radius of curvature, and from radius it derives dioptric power using the keratometric index of 1.3375:

Power (D) = (1.3375 − 1) × 1000 ÷ radius in millimetres = 337.5 ÷ radius (mm)

A 7.5 mm radius therefore gives 337.5 ÷ 7.5 = 45.00 D; a 7.8 mm radius gives about 43.27 D.

Three limitations follow directly from the method and all three are examinable:

1. It samples only the central 3 to 4 mm of the cornea. Peripheral change, including most keratoconus in its early form, is invisible to it.
2. It measures only the anterior surface and assumes a fixed relationship to the posterior surface. That assumption fails after refractive surgery, which is why standard keratometry over-estimates corneal power after myopic LASIK and produces hyperopic intraocular lens surprises.
3. It assumes the cornea is a regular sphero-cylinder. Irregularity makes the mires distort and the reading meaningless.

Normal ranges. Adult keratometry is typically 42.00 to 44.00 D, radius about 7.6 to 8.0 mm. Below about 41 D the cornea is flat (post-hyperopic surgery, some large eyes); above about 47 D it is steep and keratoconus should be suspected, particularly with high or oblique astigmatism.

Manual keratometry technique (Bausch and Lomb type)

The Bausch and Lomb instrument is a one-position keratometer: both meridians are measured without rotating the instrument body, using doubling prisms.

1. Focus the eyepiece first. Point the instrument at a plain white surface and turn the eyepiece until the reticle crosshair is sharp. Skipping this step is the single most common source of error — every subsequent reading will be focused to the operator's uncorrected refraction rather than to the instrument.
2. Seat the patient, adjust the chin rest so the outer canthus aligns with the marker, and occlude the fellow eye.
3. Ask the patient to look at the reflection of their own eye in the centre of the barrel.
4. Align the instrument until the mires are centred and the reticle sits in the lower right mire.
5. Focus by moving the joystick until the mires are crisp.
6. Rotate the instrument body so the plus signs align along one principal meridian (removes the axis obliquity).
7. Turn the horizontal drum until the plus signs just touch — this reads the horizontal meridian power.
8. Turn the vertical drum until the minus signs just touch — this reads the vertical meridian power.
9. Read both powers and the axis from the scale.

The Javal-Schiotz keratometer is a two-position instrument: the body is rotated to measure each meridian separately, with the operator aligning a stepped and a rectangular mire until they just touch.

Recording and interpreting

Record both meridian powers with their axes, for example:

K's OD: 43.00 @ 180 / 44.50 @ 090

The corneal cylinder is the difference: 44.50 − 43.00 = 1.50 D. Because the steep meridian is vertical, this is with-the-rule astigmatism. In minus-cylinder convention the axis written is that of the flatter meridian, so this cornea is roughly −1.50 × 180.

The principal meridians should be 90 degrees apart. If they are not, the cornea is irregular and the reading must be flagged.

Extended range

Standard manual keratometers read roughly 36 to 52 dioptres. Beyond that range — very flat post-surgical corneas or steep keratoconic corneas — an auxiliary lens extends the scale:

• A +1.25 D lens taped over the aperture extends the steep end; the reading is converted using the manufacturer's table (commonly measured reading × 1.166 as an approximation for the classic conversion).
• A −1.00 D lens extends the flat end.

Always document that an auxiliary lens was used and which conversion was applied, because an unlabelled converted reading entered into an intraocular lens formula is a refractive surprise waiting to happen.

Automated keratometry and when to distrust either

Automated keratometers and the keratometry built into autorefractors, optical biometers and topographers are fast, require little cooperation and remove operator focusing error. They are the practical default for screening and for biometry. Their weakness is that they will produce a plausible number from a poor tear film or a poorly fixating eye without telling you the reading is unreliable.

Signs that keratometry should not be trusted:

Tear film is the number one practical cause of bad keratometry. Before any biometry for intraocular lens calculation, treat surface disease, have the patient blink fully, and repeat until readings are reproducible. Contact lens wearers must discontinue wear before keratometry for surgical planning — commonly at least one to two weeks for soft lenses and longer for rigid lenses, per the surgeon's protocol — because lens-induced corneal warpage distorts the readings.

When keratometry is irregular or suspect, corneal topography or Scheimpflug tomography supersedes it, because those instruments map thousands of points across the whole cornea rather than four points in the central few millimetres.