Patient Counseling and Rational Drug Use — Artificial Intelligence in Pharmacy

Gains:

Ability to prepare understandable drug use and rational drug use (RAİK) information texts appropriate to the patient level with artificial intelligence
Ability to confirm and simplify every medical claim in the consultancy text with current SmPC/KT and guidelines within safe limits
Ability to consider the limits of artificial intelligence in diagnosis, treatment recommendation and over-the-counter medication guidance and the need to refer to a physician

A patient should leave the pharmacy not only with a box of medicine, but with the knowledge of how to use that medicine correctly. Patient counseling is the most visible and valuable job of the pharmacist: explaining in a language the patient can understand how, when, how long to use the medicine, and what to pay attention to. Artificial intelligence is a powerful aid in this work because it is quick to translate complex medical information into simple sentences, adapt text for different literacy levels, and remind you of a forgotten warning. But consulting is a relationship of trust; Every sentence produced by artificial intelligence must be filtered with the knowledge and responsibility of the pharmacist before reaching the patient. In this unit, you will learn to use artificial intelligence safely for patient counseling and rational drug use.

What is rational drug use?

Rational drug use (RUD), as defined by the World Health Organization, is the patient's use of the right medicine, in the right dose, for the right duration, at the most affordable cost and in the right way. Pharmacist consultation is the most important tool of AİK. The practical elements of AIC are:

The patient understands what the medicine is used for.
Compliance with dose, frequency and duration (e.g. not stopping the antibiotic).
Method of administration (fasting/full, swallow without chewing, storage condition).
Expected effect and possible side effects.
Interactions (other drug, food, alcohol).
When to consult a doctor.

AI can produce understandable draft text for each of these elements; But content accuracy and suitability for the patient belong to the pharmacist.

Tip: Good counseling text allows the patient to understand "what I need to do" in one reading. Ask the AI to write the text at a 6th grade level, with short sentences and numbered steps; Then you control medical accuracy.

Where does artificial intelligence stand in consultancy?

Artificial intelligence has a clear limit: it cannot diagnose, recommend treatment, or replace the physician. When a patient says “I have a headache, what should I take”, AI can produce a list; But presenting this list directly to the patient is dangerous. A headache can be a simple tension or a sign of a serious condition. Even when recommending an over-the-counter (OTC) product, the pharmacist; should evaluate the patient's age, pregnancy, chronic diseases and other medications. Artificial intelligence cannot make this assessment.

When it comes to red flags (symptoms that require immediate referral to a physician), AI can be a reminder (“refer to a physician if you have these symptoms”), but the decision is still made by the pharmacist.

There is also an easily overlooked aspect of counseling: whether the patient actually understands the text. Giving an instruction is not the same as having the patient understand it correctly. That's why a good pharmacist uses the "teach-back" method: he asks the patient to repeat in his own words how to use the medicine. Artificial intelligence is a preparation tool at this point; It can generate in advance a few simple control questions to be asked to the patient ("how many times a day will you take it, is it hungry or full?"). But it is the pharmacist who conducts the retelling, reads the hesitation on the patient's face, and explains it again in a different language when necessary; No text producer can make this human evaluation.

three mini cases

Case 1 — Inhaler use. A 58-year-old patient diagnosed with asthma was prescribed a new inhaler. The pharmacist asked the artificial intelligence for a step-by-step, simple instruction text. Artificial intelligence produced a 6-step instruction. The pharmacist compared the steps with the device's IFU, corrected the order of the "exhale" step, and added the mouth rinse warning. The patient received the text on one page; Artificial intelligence saved time, the pharmacist guaranteed accuracy.

Case 2 — Antibiotic compliance. A 30-year-old patient took 7 days of antibiotics but was inclined to "stop when I get better." The pharmacist asked the artificial intelligence for a short 4-item compliance text focused on RAC: why it should be used for the full time, what to do if the dose is missed, what to do if there are side effects, storage. The pharmacist confirmed the text with the SmPC and read it to the patient. The sentence describing the risk of resistance increased the patient's compliance.

Case 3 — Elderly polypharmacy user. A 75-year-old patient using 6 medications needed a simple "daily medication schedule." The pharmacist asked the AI for a morning/noon/evening chart. Artificial intelligence set the table; The pharmacist corrected the fasting/full status and time of each drug according to the SmPC and separated the intake time of two drugs that affect each other. The result was a clear chart that the patient could hang on the refrigerator.

Step by step consultancy text production

Define the context. Patient's age, literacy level, language, special circumstances (without identification).
Prepare the source. Have the relevant SPC/KT and the guide, if any, at hand.
Request a draft. Request simple, step-by-step text appropriate to the target level from artificial intelligence.
Check medical accuracy. Compare each dose, warning and instruction to the SmPC/CT.
Adapt to the patient. Personalize by taking into account allergies, pregnancy and other medications.
Reinforce verbally. While giving the text, repeat the most critical 2-3 points verbally.

Weak prompt / Strong prompt

Weak: "Write down how to use this medicine."

Güçlü: "Write an instruction text with short sentences and numbered steps, at the 6th grade level, to be given to the patient for the following drug. Include the following: what it is used for, how many times a day and when, fasting/full, missed dose, 2-3 common side effects and when to consult a physician. Drug: [name]. Mark any information you are not sure of as 'Must be confirmed from the SmPC'. Dose fitting."

The target literacy level, content titles and verification requirement are clear in the powerful prompt.

Four copyable templates

Task: Patient instructions (DRAFT).Medication: [name/form]. Patient level: 6th grade, plain Turkish. Content: 1) What for 2) Dosage and time 3) Fasting/full 4) How to apply5) Common side effects 6) When to see a physician. Numbered, short sentence.Rule: Mark the dose/warning with [CONFIRMATION] to be confirmed in the SmPC.

Task: Rational drug use compliance text (antibiotic).Emphasis: full term use, resistance, missed dose, storage.Tone: non-judgmental, motivating, 4 items. Drug: [name].No made-up information; [CONFIRMATION] sign in uncertain.

Task: OTC consultancy pre-assessment generate QUESTIONS (NOT suggestions).Complaint: [...]. Output: questions to ask the patient + red flag symptoms that require referral to a physician. Do not recommend medication; Just give the evaluation framework.

Table of the role of artificial intelligence in consulting

advisory role

artificial intelligence does

Man (pharmacist) makes

Convert to plain text

draft language

Medical fact checking

Literacy adaptation

Level adjustment

Cultural/personal adaptation

List of side effects

General list

Suitability for this patient

OTC product selection

Evaluation question

Product decision and guidance

Diagnosis

nothing

Referral to physician

Common mistakes

Giving the artificial intelligence text to the patient without reading it. Each sentence must be checked with SPC/KT.
Leaving OTC recommendation to AI. The product decision belongs to the pharmacist who evaluates the patient.
Bypassing red flags. Symptoms that require referral to a physician must be observed.
Overly technical language. If the patient does not understand, counseling is unsuccessful; Simplify the level.
Forgetting to personalize. Allergies, pregnancy and other medications should be reflected in the text.

In summary

Artificial intelligence is a powerful aid in translating complex drug information into plain text that the patient can understand, saving time for counseling. But counseling is a matter of trust and responsibility: diagnosis and treatment decisions are shaped by the pharmacist's evaluation and, when necessary, referral to the physician. Check each text with the SPC/CT, personalize it to the patient, and verbally reinforce the most critical points. Artificial intelligence writes the text; The pharmacist guarantees accuracy and suitability.

Application task

Ask the AI for a 6th grade level patient instructions for a drug you encounter frequently. Compare each dosage, warning and instruction in the text with the relevant SmPC/CT and mark any errors. Then, create a "daily medication calendar" table for the same drug and fill in the fasting/full and time information according to the SmPC. Finally, for an OTC complaint, have the AI produce “assessment questions and red flags, not suggestions” and compare them with your actual practice.

checklist

[ ] I defined the patient level and their specific situations.
[ ] I produced the text in a simple, step-by-step manner and at the target level.
[ ] I confirmed each dose/warning/instruction with the SmPC/CT.
[ ] I personalized the text to the patient (allergy/pregnancy/medication).
[ ] At OTC I based the product decision on evaluation.
[ ] I looked out for red flags and referrals to the physician.
[ ] I reinforced the most critical 2-3 points verbally.