How Artificial Intelligence Is Transforming Patient Safety
AI-powered drug interaction checkers, language simplification, and specialized chatbots: the future of digital package inserts has already begun.
Digital package inserts, as defined by RDC 885/2024, solve fundamental access and accessibility problems. But the next frontier is already being explored: using AI to make medication information not just accessible, but truly understandable and actionable for patients.
Drug interaction checking
Polypharmacy (simultaneous use of multiple medications) is the reality for millions of Brazilians. Data shows 75% of Brazilians over 65 use at least 2 medications continuously, and 30% use 5 or more. AI-powered checkers can cross-reference all known interactions, classify by severity, explain in accessible language, and suggest when to seek medical guidance.
Language simplification
Package inserts are written for healthcare professionals. Terms like "bioavailability" and "renal clearance" are incomprehensible to most patients. Language models can generate simplified versions maintaining technical accuracy but using everyday vocabulary. The simplified version never replaces the official insert: it's presented as an additional help layer with clear disclaimers.
Specialized insert chatbots
Imagine patients asking: "Can I take this with milk?", "I missed a dose, what should I do?" or "Does this medication cause drowsiness?" A chatbot trained specifically on insert content can answer directly, citing the relevant section as source. Unlike generic health chatbots, responses are based exclusively on ANVISA-approved content with full traceability.
Regulatory considerations
ANVISA has not yet specifically regulated AI use in digital inserts. However, the resolution requires RIEP content to be "identical to approved." This means AI features must be clearly separated from regulatory content, presented as "help tools" with appropriate disclaimers.