9.2 Core Workflows and Decision Points

9.2 Core Workflows and Decision Points

A reliable biometry session is a sequence, not a single button press. Each step has a decision point where the technician can catch an error before it reaches the operating room.

The standard measurement sequence

Always do keratometry before any drops or contact testing. Topical anesthetic, fluorescein, and especially applanation tonometry or a contact A-scan probe distort the corneal surface and corrupt K readings. The cornea should be measured on an undisturbed tear film.

Choosing the right method

Optical biometry (IOLMaster, partial coherence interferometry) is the default because it is non-contact and reproducible. It fails when light cannot reach the retina, a dense or mature cataract, vitreous hemorrhage, or a corneal scar. In those cases switch to immersion A-scan (preferred over applanation because it avoids corneal compression).

Before any A-scan, set the device to the correct eye status, which selects the ultrasound velocity:

• Phakic (natural lens present): mixed velocity, typical setting ~1550 m/s composite
• Pseudophakic (IOL already implanted): velocity adjusted for the lens material (PMMA, acrylic, silicone)
• Aphakic (no lens): single aqueous/vitreous velocity ~1532 m/s
• Silicone-oil-filled eye: a special low velocity setting, because sound travels much slower through silicone oil

Choosing the wrong mode is a classic exam trap: measuring a silicone-oil eye on a phakic setting gives a grossly wrong axial length.

Quality controls and handoffs

Two high-value sanity checks should be habitual. First, compare the eyes: in most patients the two axial lengths agree within about 0.3 mm and keratometry within about 1 D. A larger asymmetry means a measurement error, true anisometropia, or a real pathology, repeat before accepting. Second, confirm the A-constant matches the exact IOL model the surgeon will implant; the A-constant is a lens-specific number that calibrates the formula. A correct AL and K with the wrong A-constant still produces a refractive surprise.

The output of this workflow is a clean biometry sheet, but the responsibility is shared: the technician acquires and flags, the surgeon selects target refraction and lens. Document outliers and how you resolved them so the chart shows why a value was repeated or overridden.

Patient positioning and acquisition technique

Good data starts with positioning. For optical biometry the patient sits with chin in the chinrest and forehead firmly against the bar, fixating the internal target so the measurement is taken along the visual axis through the fovea, not obliquely. Off-axis fixation lengthens the apparent axial length, the same way a tilted A-scan probe does. Take several readings, most devices acquire and average multiple sweeps, and reject any single sweep with a poor signal-to-noise ratio rather than letting it pull the mean.

For immersion A-scan the patient lies supine, a topical anesthetic is instilled, and a fluid-filled scleral shell (for example a Prager or Hansen shell) is placed; the probe floats in saline or methylcellulose above the cornea so it never touches it. For applanation A-scan, used only when immersion is unavailable, the probe is brought perpendicular to the cornea with the lightest contact that still gives clean spikes, because any indentation shortens the eye.

Selecting the target refraction

The surgeon, not the technician, chooses the target (desired) postoperative refraction, often plano (zero) for distance, or a small minus value such as −0.25 to −0.50 D to hedge against a hyperopic surprise, or deliberate monovision (one eye targeted for near). The biometry sheet must therefore present the predicted refraction for several IOL powers so the surgeon can pick.

A frequent workflow handoff error is recording the patient's wishes, distance versus reading vision, inaccurately; confirm and document the visual goal, especially for monovision and toric (astigmatism-correcting) lenses where the steep-axis keratometry must also be captured.

Order of operations and why it is fixed

The sequence is not arbitrary. Keratometry comes first because it is the most easily contaminated, anything that touches or dries the cornea ruins it. Optical axial length and ACD are captured in the same non-contact sitting. A-scan, if needed, comes after K because the contact probe distorts the cornea. Drops for dilation are generally fine for axial-length acquisition but the keratometry should already be done.

Document the device and software version used, because an A-constant is optimized for a specific device and lens pair; an A-constant tuned for ultrasound is not interchangeable with the optical-biometry A-constant for the same lens, a subtlety that causes systematic refractive surprises if mixed.

When to repeat versus when to escalate

Repeat a reading yourself when the cause is mechanical, alignment, fixation, tear film, or device mode. Escalate to the surgeon or a senior technician when the cause is anatomic or beyond your scope, an unexpected staphyloma, a corneal scar that no fixation fixes, suspected silicone oil not in the chart, or a post-refractive history that needs special formulas and counseling. The exam reliably distinguishes "recheck your own technique" answers from "flag it up the chain" answers; the right choice depends on whether the problem is the operator or the eye.