4.2 Core Workflows and Decision Points

4.2 Core Workflows and Decision Points

A reliable automated field begins long before the first stimulus. The COA controls setup, and most unreliable tests trace back to a setup error rather than the patient. Treat every field as a small procedure with a fixed order: correct the refraction, position the head, explain fixation, run a practice, then test - and read the reliability header before you trust the map.

Step-by-step Humphrey setup

1. Enter demographics and pupil size. Age normalizes the result; a pupil under ~2 mm can cause artifactual depression.
2. Add the correct trial lens. Use the patient's distance refraction plus the appropriate near add for the bowl distance (about +2.50 to +3.25 for presbyopes on the 24-2/30-2). Without the near add, the whole field reads dim - a classic artifactual generalized depression.
3. Patch the non-tested eye completely; light leaks cause false positives.
4. Position the patient. Forehead against the bar, chin in the rest, eye centered on the monitor camera. Keep the lens close to the eye to avoid a lens-rim artifact (a dense ring scotoma in the periphery).
5. Explain fixation. The patient stares at the central target and presses the button whenever a light is seen, even a faint one - and does not chase the lights.
6. Run a brief demo/practice, then start the recorded test.

Reading the four reliability indices

The printout header reports reliability before the physician trusts the map. Memorize these cutoffs.

How fixation losses are measured

The Humphrey uses the Heijl-Krakau method: periodically, instead of testing a normal point, the machine projects a stimulus into the patient's mapped physiologic blind spot. Because there are no photoreceptors there, a properly fixating patient should never see it. If the patient presses the button, the eye must have moved off the central target, and the machine tallies a fixation loss. This is why correctly plotting the blind spot at the start matters - a mis-plotted blind spot inflates the fixation-loss count and can wrongly condemn a good test.

The decision rule

• High false positives mean a trigger-happy patient who presses without seeing a light; sensitivity values are inflated and the field looks better than it truly is. White, supranormal-looking points appear. Reinstruct ("press only when you actually see a light") and repeat.
• High false negatives or high fixation losses mean the field looks worse than reality, usually from fatigue, inattention, drowsiness, or poor head positioning. Offer a short break, reposition the patient, re-explain fixation, and repeat.
• A cloverleaf pattern - seeing only the four central points the machine tests first, with the rest of the field blank - indicates the patient understood the task initially but stopped responding partway through, often from fatigue or confusion.

Confrontation and Amsler workflows

Not every field is automated. For confrontation testing, you sit about one meter in front of the patient, have them cover one eye while you cover your opposite eye, and present fingers or a target in each of the four quadrants, asking the patient to count fingers or say when they see motion. Your own normal field is the reference standard - if you see the target but the patient does not, a defect in that quadrant is likely.

For the Amsler grid, the patient holds the grid at normal reading distance, wears their reading correction, covers one eye, and fixates the central dot while reporting any missing, blurred, or wavy lines; wavy lines (metamorphopsia) or a central gap suggest macular disease and warrant referral for OCT or dilated exam.

Documentation and handoff

Record the eye tested, pupil size, the trial-lens power used, the test pattern and strategy, and any reliability concerns directly with the printout. If you repeated a test, keep both runs so the physician can see the artifact and your correction. Never silently discard a bad field; the fact that the first attempt was unreliable is itself clinically useful information about the patient's ability to perform the test, and the physician may need it to judge whether a future field is comparable.

Worked example

A 68-year-old glaucoma patient finishes a 24-2 SITA test. The header shows fixation losses 5 percent, false positives 28 percent, false negatives 8 percent, and several scattered white supranormal points across the map. Two of the three indices are fine, but false positives are elevated above the SITA flag of roughly 15 percent, so the result overstates the patient's true vision. The correct action is to reinstruct the patient and repeat the test, not to record it as an improving field. Reporting an inflated field as genuine improvement could wrongly suggest the glaucoma is stable and delay needed treatment changes.