Explaining Lens Options and Eye Test Results With AI Help

Use AI to draft plain-English explanations that your optometrist or dispensing optician checks before any patient sees them: a small library of lens-option explainers (varifocals, bifocals, thinner lenses, coatings) and a template that turns the clinician's findings into a short patient summary. Keep diagnosis and recommendations human, and keep identifiable records out of consumer AI tools.

The common mistake is letting AI write the claims as well as the wording. Ask a chat assistant about lens coatings and it will happily tell patients that blue-light filters reduce eye strain, yet a 2023 Cochrane review of 17 randomised trials found they probably make no difference to eye strain, visual performance or sleep quality. Your explainers need a claims check as well as a readability check.

Follow me on Instagram@sagnikteaches

What patients go home confused about

Ask the front desk which questions come back by phone after appointments. In most practices the list looks something like this:

Connect on LinkedInSagnik Bhattacharya

• "What do the numbers on my prescription mean? Has it got worse?"
• "Why are varifocals so much more than my old glasses?"
• "The optometrist said something about the back of my eye. Is that serious?"
• "Do I need the anti-reflection coating, or is it an upsell?"
• "My new varifocals make the floor look strange. Is that normal?"

Each is a gap in explanation, not a clinical problem. That's where AI helps: it's quick at producing clear wording from technical input, and your clinicians check it once rather than explaining it forty times.

Subscribe on YouTube@codingliquids

Step 1: Build a library of lens explainers, checked once

Start with eight to ten short explainers covering what patients ask most: single vision, bifocal and varifocal lenses; thinner (higher-index) lenses; anti-reflection coating; photochromic lenses; tints and polarised lenses; getting used to varifocals; lens options for children. Each should be under 200 words and written for a reading age of about 11.

Write a patient information note for an independent optician's practice.
Topic: getting used to new varifocal lenses.
Audience: adults collecting their first varifocals. Plain English,
reading age about 11, under 180 words, short paragraphs.
Include: what feels different at first, simple tips, when to contact
the practice.
Rules: no medical claims or statistics; no brand names; do not say how
long adaptation takes in days or weeks (we will add our own wording);
end with "If anything worries you, call us on [number]."

An illustrative output, trimmed:

Your new varifocals have different zones for distance, middle distance and reading, all in one lens. At first, you may notice the floor looks a little closer when you walk, or that things at the edges look slightly blurred. This is normal while your eyes and brain get used to the lenses.

Wear them as much as you can, rather than switching back to old glasses. Point your nose at what you want to see, and move your head rather than just your eyes. Take care on stairs until you feel confident: look down through the top part of the lens.

Notice the rule about adaptation time. Assistants tend to invent a figure ("most people adapt within two days"), and patients hold you to it. Your dispensing optician adds the practice's own wording and a line about when to come back for an adjustment. Once approved, the note goes into your library and is reused, not regenerated each time, so the checked version is the one patients get.

The coatings explainer is where the claims check earns its keep. An illustrative first draft for anti-reflection coating read: "Anti-reflection coating reduces glare, eases eye strain from screens and protects your eyes from harmful blue light." Only the first claim is safe. "Eases eye strain" is a health benefit nobody has sourced, and "harmful blue light" imports the blue-light claim the Cochrane review doesn't support. The approved version: "Anti-reflection coating cuts down reflections on the front and back of the lens, including glare from headlights at night and from screens. People looking at you see your eyes rather than a shine on your glasses, which helps on video calls and in photos. Ask us which coating your lenses have, as some add extras such as a layer that makes them easier to clean." Every sentence in that version describes something the lens does, not something it does for the patient's health.

Step 2: A prescription explainer that reads the numbers

Patients get a slip of numbers and no key. A one-page guide covering each term, plus an AI-drafted note for each patient, closes the gap. The terms, in plain words:

Given a prescription of R −2.25 / −0.75 × 180, L −2.00 / −0.50 × 170, ADD +2.00, and the optometrist's note "small increase in distance power since last test; ADD increased from +1.75," an assistant produced this illustrative note, which the optometrist approved with one change:

You are short-sighted, which means things in the distance look blurred without glasses. Both eyes also have a small amount of astigmatism, which we correct in your lenses. Since your last test, your distance prescription has gone up slightly, and you now need a little more help for reading. These are common, gradual changes.

The change: the draft originally said "these are normal changes for your age." The optometrist replaced that with "common, gradual changes," because whether something is normal for this patient is a clinical judgement, and the words "for your age" land badly with some people. That kind of edit is exactly what the review step is for.

Never let the assistant work out the change itself by comparing two prescriptions, because the same prescription can be written two ways. Take a patient whose previous prescription, from another practice, was written in plus-cylinder form: R −2.75 / +0.75 × 90. Your practice writes minus-cylinder: R −2.00 / −0.75 × 180. Those describe the same lens. An assistant comparing the SPH figures would tell the patient their distance power has "improved by 0.75", and the patient would reasonably wonder why they need new glasses. The fix is in the input, not the prompt: the optometrist states the change in the findings sheet ("distance: no change") and the assistant only puts it into words.

Step 3: Turning eye test findings into a short summary

A written summary of the test helps patients remember what they were told and share it with family. The safest way to produce it is from a structured findings sheet the optometrist completes, never from free-form notes or the full record:

Vision with current glasses: slightly reduced in the distance
Prescription change: small
Eye health: healthy, except early lens clouding in the right eye
  (monitor, no treatment needed now)
Advice: update glasses; bright light for reading
Next test: 2 years, sooner if vision changes
Referral: none

Ask the assistant to turn this into a 120-word summary for the patient, keeping every finding exactly as stated and adding nothing. Words like "cataract" or "glaucoma suspect" frighten people, so tell the assistant to use the term the optometrist chose, followed by a one-line plain explanation from your approved library, rather than writing its own. Referrals and anything urgent are explained by the optometrist in person and in writing, following your professional guidance; AI can help format a referral letter, but it never decides or softens one.

Here's an illustrative first draft from that findings sheet:

Your vision with your current glasses is a little reduced in the distance, and your prescription has changed slightly, so new glasses will help. Your eyes are healthy. There are early signs of a cataract in your right eye, which is very common and nothing to worry about. We'll keep an eye on it. Good lighting will help with reading. We'd like to see you again in two years, or sooner if your vision changes.

Two corrections before it goes to the patient. The optometrist wrote "early lens clouding", and the draft swapped in "cataract", which is accurate but is the word the optometrist chose not to lead with; the approved wording keeps "early clouding of the lens" and adds the library's one-line explanation. And "nothing to worry about" is a reassurance only the optometrist can give, so it becomes "this doesn't need any treatment at the moment, and we'll check it at your next test". Everything else matched the sheet line for line, which is what the monthly sample check should confirm.

On data: use a tool your practice has approved for patient information, on a business plan that doesn't train on your content by default, or leave the name and identifiers out of the prompt and add them in your practice system afterwards. The patient data confidentiality checklist for small practices sets out what to check before any of this starts.

Step 4: Comparing options at the dispensing desk

A simple comparison, drafted by AI from your own price list and checked by the dispensing optician, makes the lens conversation easier and less like selling. An illustrative sheet for the prescription above:

A realistic mistake from a first draft: the assistant recommended the thinnest high-index lenses for this prescription. For a mild prescription like −2.25, the thickness difference is small and the extra cost is hard to justify, so the dispensing optician removed it. Your dispensers know when an option helps; the AI only knows that thinner sounds better. For a broader look at AI in the dispensing area, see whether AI frame advice and virtual try-on are worth it.

Claims to keep out of AI-written patient material

• Blue-light filters as eye-strain or sleep fixes. The 2023 Cochrane review found they probably make no difference to eye strain, visual performance or sleep quality. If you sell them, describe what they do without implying benefits the evidence doesn't support.
• "Prevents" anything. Myopia management lenses for children are described by manufacturers as slowing progression, supported by their own evidence. Use the manufacturer's approved wording, not the AI's.
• "Normal" and "nothing to worry about." Only the optometrist can say this about a specific patient.
• Answers to symptom questions. Sudden changes in vision, flashes, new floaters or eye pain go straight to a person, never an automated reply.
• Invented statistics. "90% of people adapt in a week" and similar lines appear in drafts regularly. Delete any number you can't source.

Checking AI content for plain English and readability covers the readability side of the same review.

An illustrative two-optometrist practice

Consider an independent practice with two optometrists, one dispensing optician and about 22 eye tests a day. Currently, patients leave with a prescription slip and whatever they remember. The front desk estimates about six calls a day asking what the prescription means, what was said about eye health, or whether a varifocal problem is normal.

Setup takes about two evenings: one to draft the ten lens explainers and the prescription guide, one for the clinicians to check and approve them. From then on, the optometrist completes the findings sheet during the test (about a minute), the assistant drafts the summary, and the optometrist reads and approves it before the patient leaves the chair or while they are at the dispensing desk (about a minute). That adds roughly 20 minutes a day across both clinicians.

What to expect in return is fewer follow-up calls and calmer dispensing conversations, but measure it rather than assume it: keep the front desk's tally of explanation calls for a month before and a month after. If the calls don't fall, the summaries probably aren't answering the question patients actually have, and the tally will tell you which one.

A quick sum keeps expectations honest. Six explanation calls a day at around four minutes each is 24 minutes of front-desk time. The summaries add about a minute of clinician review per test, so across 22 tests the practice spends roughly 20 minutes to save up to 24. On time alone, that's close to even, and clinician minutes cost more than front-desk minutes. The case for doing it rests on patients understanding their results and trusting the dispensing conversation, so judge it on the call tally and the say-it-back check, not on minutes saved.

Checking the explanations actually work

• Ask patients to say it back. "Just so I know I explained it well, what will you tell your family about today?" If they can't, the wording needs work.
• Log the questions that still come in, and update the matching explainer each quarter.
• Keep a version date on every library item, and re-check claims when manufacturers change their product information.
• Review a sample of summaries monthly against the findings sheets to catch anything the assistant added.

The question log is only useful if it leads to a change. An illustrative quarter's log for the two-optometrist practice, with what was done about each:

The driving row is worth copying. Some questions look like explanation gaps but are really clinical, and the log is where you decide which is which before an assistant ever drafts an answer.

The same approach works across healthcare. Explaining dental treatment plans with AI uses a similar findings-first structure, and eight practical AI uses for an independent optician shows where patient explanations sit among the other jobs worth automating.

Questions from optometrists and dispensing opticians

Can we use AI to translate explanations into patients' own languages?

Yes for the general lens explainers, which contain no personal data, as long as a fluent speaker checks the translation before it goes into your library. For individual summaries, translate only in a tool your practice has approved for patient information, and ask the patient to confirm they understood the key points.

Who is responsible if an AI-drafted summary is wrong?

The clinician who signs it off, exactly as if they had typed it. That is why every summary is checked against the record before it leaves the practice. Speak to your professional body or indemnity provider if you want specific guidance on how AI-drafted patient material fits your obligations.

Can patients ask an AI chatbot about their results?

Not about their own results or symptoms. A website assistant can answer general questions about lens types, prices and appointments, but questions about a patient's own eyes, or any new symptom, should go to a person at the practice. Put that rule in the assistant's instructions and test it.

Further reads

• How Independent Opticians Can Use AI for Recalls and Bookings — Get patients back in and booked with AI reminders.
• AI Hallucinations Explained for Business Owners: Causes and Fixes — Why AI states wrong things confidently, and how to catch it.
• How Accountants Use AI to Explain Tax to Clients in Plain English — Another profession turning technical findings into plain English.
• What to Ask Before Buying Any AI Tool for a Medical Practice — What to ask before buying an AI tool for a clinical setting.
• How a Small Clinic Can Use AI at the Front Desk Safely — Using AI safely at a clinic's front desk.
• Writing Pet Owner Emails and Treatment Explanations With AI — How vets write treatment explanations with AI.
• AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.

Sources: Cochrane Database of Systematic Reviews, blue-light filtering spectacle lenses (Singh et al., August 2023); College of Optometrists news on the review.