Medication reconciliation is a critical process for ensuring patient safety during transitions of care. However, the manual effort required often leads to inconsistencies and potential errors. **AI in medication reconciliation** is emerging as a transformative tool to enhance the accuracy of medication histories. A recent scoping review analyzed the current landscape of AI applications in this field to determine their effectiveness.

Current Progress of AI in Medication Reconciliation

The review identified three primary tasks: creating medication histories, identifying discrepancies, and resolving them. Most research focuses on the first stage. Specifically, 97.9% of the studies automated information acquisition. Researchers primarily used machine learning to extract data from clinical notes. Additionally, some studies utilized pill images to identify medications through convolutional neural networks. These advancements show how AI can streamline the heavy workload of data collection.

Despite these technological advancements, significant gaps remain. Only a tiny fraction of studies addressed discrepancy identification. Furthermore, the automation of discrepancy resolution is still largely unexplored in current literature. Most models currently exist as proofs-of-concept rather than clinical tools. Therefore, moving from model development to real-world usability is essential for clinical practice. This evolution will require more focus on the integration of these tools into existing workflows.

Future Challenges for AI-Based Tools

Healthcare providers must address data incompleteness to improve these systems. Because many sources lack structured data, AI models often struggle with perfect accuracy. Consequently, future efforts should prioritize discrepancy resolution and real-world implementation. This shift will ensure that AI-driven tools provide meaningful support to clinicians and patients alike. By overcoming infrastructural barriers, the medical community can move toward a safer, automated reconciliation process.

Frequently Asked Questions

What are the three core tasks of medication reconciliation?

The core tasks include creating a best possible medication history, identifying discrepancies between different medication lists, and resolving those discrepancies to ensure patient safety.

How is AI currently used in medication history creation?

Most AI applications focus on extracting medication data from free-text clinical notes using machine learning or identifying medications through pill images using neural networks.

What are the main barriers to AI adoption in medication reconciliation?

Primary barriers include data incompleteness in electronic health records and a lack of studies focused on real-world usability and the resolution of medication discrepancies.

Disclaimer: This content is for informational and educational purposes only. It does not constitute medical advice or a professional recommendation. Refer to the latest local and national guidelines for clinical practice.

References

Tabja Bortesi JP et al. AI-Based Automation for Medication Reconciliation: Scoping Review. J Med Internet Res. 2026 May 11. doi: 10.2196/86760. PMID: 42114154.

World Health Organization. Medication Safety in Transitions of Care. Technical Report. 2019.

Institute for Safe Medication Practices. Medication Reconciliation: The Next Generation of Patient Safety. 2023.