What an Allergy Practice Should Look For in an AI Copilot (A Visit-by-Visit Checklist)

The short answer: an AI copilot for an allergy practice should be judged

on allergy visits, not on a generic demo. Check five things. Does it draft

an allergy-structured note (history, review of systems with pertinent

negatives, assessment and plan, workup)? Does it have **templates for the

visits you run**: new patient, follow-up, skin testing, immunotherapy, asthma,

and food and drug challenges? Can the clinician **check a line against its

source before signing? Does the clinician sign everything**, with nothing

filed on its own? And will the vendor sign a BAA and work with your EHR?

The clocks go back this weekend, and the drive home gets darker earlier. In

a lot of allergy practices the note still goes home too. That is the problem

an AI copilot is supposed to solve, and it is why the choice deserves more

than a five-minute demo on a generic primary-care visit.

Physicians are already deciding this. In the AMA's 2026 survey, 81% of

physicians said they use AI in their practices, up from 38% in 2023. The same

survey found that 86% emphasize data privacy, 88% want robust safety and

efficacy validation, and 85% want to be consulted or directly involved in

decisions about AI adoption. The AAAAI's work group on augmented intelligence

put it in specialty terms: "specialists should be involved in the design,

validation, and implementation of AI in allergy and immunology."

This checklist turns that into questions you can ask on a trial with your

own visits.

1. Start with your visit mix, not the feature list

Write down the five visit types that fill your week, then make every vendor

show each one. For most allergy practices that list looks like this:

| Visit | What the note has to carry | Ask the vendor |

|---|---|---|

| New patient consult | Symptom timeline, triggers, home and work exposures, prior testing, medications tried, family history | Does the draft keep the allergy history structured, or flatten it into a paragraph? |

| Skin-test day | Order, controls, sites tested, results the staff charted, who signed | Does the note sit in the same record as the skin-test sheet? |

| Immunotherapy | Build-up or maintenance, dose, reactions since the last injection, the pre-shot check | Is there a template for the shot visit, or only for a full consult? |

| Asthma follow-up | Control since the last visit, rescue use, triggers, spirometry if done, the plan | Does it capture what the patient does not have? |

| Food or drug challenge | Indication, doses and times, observations, outcome, instructions | Is there a challenge template at all? |

The rows after the first one are where allergy work is its own specialty,

so give them most of the trial.

2. Pertinent negatives are the allergy note

Allergists work by exclusion. "No fever, no body aches, symptoms ease

indoors" is what separates seasonal rhinitis from a lingering virus, and it is

what the next clinician needs to read. A copilot that only writes down what

the patient has leaves out half the reasoning.

Test it: run a visit where most of the important history is negative, and

see whether the draft keeps those negatives in the review of systems. Our

guide on allergic rhinitis vs. cold, flu or RSV

shows which negatives matter in that visit.

3. You need to see where each line came from

The clinician signs the note, so the clinician needs a fast way to check it.

Ask whether lines in the draft link back to the moment in the visit

conversation where they were said. A copilot that only hands you finished

prose leaves you two options: trust it, or re-listen to the visit in your head.

Ask: "Show me how I check this line before I sign." If the answer is

"it's very accurate", that is not a way to check.

4. The clinician signs every note

This should be non-negotiable. The copilot drafts; the clinician reviews,

edits and signs. Nothing should be filed, submitted or finalized without the

clinician. Be careful with any tool that says it diagnoses, recommends

treatment or decides on immunotherapy. Those decisions are yours.

Ask: "What happens to a draft nobody signs?" The right answer is that it

waits for the clinician.

5. Skin testing should live in the same record as the note

On skin-test days the note and the test results describe the same visit. If

they live in two systems, someone re-types the results, and that is where

errors creep in. Look for a skin-test workflow in the same patient record as

the note: the order, controls first, the sites tested, the values staff

charted or confirmed, and the allergist's sign-off. Unchecked sites should

stay open or be marked not tested, never saved as "no reaction".

For the documentation itself, see our

skin-test documentation checklist.

Ask: "Who confirms each value, and where is that recorded?"

6. The follow-up visit should start where the last one ended

Allergy care is longitudinal. The patient who tested positive to dust mite in

the spring comes back in the fall with the furnace on. Ask whether the

copilot shows a snapshot of the prior visit, keeps the patient's allergen

profile across visits, and carries chronic conditions forward so they are not

re-entered every time.

Ask: "Open a second visit for the same patient. What do I see before I

walk in?"

7. Codes are suggested, not decided

A copilot can suggest ICD-10 codes from the note, but the clinician reviews

them. Look for codes attached to each assessment item so the review is quick,

and check the suggestions on your own allergy diagnoses. Our

allergy ICD-10 coding guide

lists the codes most allergy notes use.

8. Privacy, the BAA and your EHR

Before a single real visit goes through the tool:

• BAA: the vendor signs a business associate agreement.
• Where the data lives: which cloud, which region, how it is encrypted at

rest and in transit.

• Security program: ask for the current status in plain words. "In

progress" and "certified" are different answers.

• EHR: does it pull the schedule and work with your chart, or are you

copying and pasting?

9. Run the trial on your own visits

A demo shows the vendor's best visit. A trial shows yours. Give it a week of

real clinic days and judge it on three things you can see without a stopwatch:

1. How much you change in each draft before you sign.
2. Whether the negatives, the exposures and the plan come through.
3. Whether your notes are signed before you leave the building.

Where Medora fits

Medora is an AI copilot built only for allergy and immunology practices, and

it is built together with a working allergy practice, Allergy Affiliates in

Bradenton, Florida.

• Medora Copilot listens to the allergy visit and drafts an

allergy-structured SOAP note: HPI, review of systems with pertinent

negatives, assessment and plan, and workup. It has visit templates for new

patients, follow-ups, skin testing, immunotherapy, asthma, and drug and food

challenges, and it suggests ICD-10 codes for the clinician to review.

• Evidence Mapping links lines of the note, including review of systems

and history, back to the moment in the visit conversation where they were

said, so the clinician can check the source before signing.

• Patient Intelligence shows a snapshot of the prior visit at follow-up

and keeps a structured allergen profile that builds across visits.

• The clinician reviews, edits and signs every note.
• Medora runs on HIPAA-eligible AWS in us-east-1, covered by AWS's Business

Associate Addendum, and we offer a BAA to each practice. Every connection uses

TLS 1.2 or newer. We are not SOC 2 certified, and we will say when that changes.

In a 60-day pilot at Allergy Affiliates, the time an allergist spends signing

off each skin-test visit dropped sharply

(the case study).

Medora Scribe is $99 per provider per month after a 14-day free trial, and no

card is needed to start.

Frequently asked questions

What should an allergy practice look for in an AI copilot?

An allergy-structured note with pertinent negatives, templates for the visits

allergists run (skin testing, immunotherapy, asthma, food and drug

challenges), a way to check each line against its source, clinician sign-off

on everything, a BAA, and EHR integration.

Which visits should we test first?

The ones that make allergy different: skin-test days, shot visits, food and

drug challenges, and visits where the key history is negative.

Who is responsible for the note an AI copilot drafts?

The clinician. The copilot drafts, and the clinician reviews, edits and signs.

Nothing should be filed without them.

How should we run a trial?

Use a week of your own clinic days. Judge how much you edit before signing,

whether the negatives and exposures come through, and whether notes are signed

before you leave.

Does the AAAAI say anything about AI?

A 2022 AAAAI work group report, *A Framework for Augmented Intelligence in

Allergy and Immunology Practice and Research*, says specialists should be

involved in the design, validation and implementation of AI in allergy and

immunology.

Try it on your own visits

• See how Medora handles allergy documentation
• Start the 14-day free trial of Medora Scribe (no card needed)
• Book a demo
• Read the buyer's guide to AI for allergy and asthma practices
• The state of AI in allergy, 2026

*Sources: AMA, "AMA: AI usage among doctors doubles as confidence in

technology grows" (press release, March 12, 2026),

https://www.ama-assn.org/press-center/ama-press-releases/ama-ai-usage-among-doctors-doubles-confidence-technology-grows.

Khoury P, Srinivasan R, Kakumanu S, et al. A Framework for Augmented

Intelligence in Allergy and Immunology Practice and Research: A Work Group

Report of the AAAAI Health Informatics, Technology, and Education Committee.

J Allergy Clin Immunol Pract. 2022;10(5):1178-1188

(doi:10.1016/j.jaip.2022.01.047). Figures are as published; check the

originals for updates.*