Running Reconciliation

Multi-agent analysis in progress

Gathering patient data from FHIR

Retrieving medications from providers

Drug Interaction Agent: checking conflicts

Guidelines Agent: validating dosing

Building reconciliation report

Medication Reconciliation

Intelligent Medication
Reconciliation, Automated.

Multi-agent AI system that coordinates across FHIR data sources, checks drug interactions, validates clinical guidelines, and produces actionable reconciliation reports.

How It Works

System Architecture

Three specialized agents coordinate through standardized protocols to perform comprehensive medication reconciliation.

Orchestrator Agent

Coordinates the reconciliation workflow. Delegates tasks to specialist agents and assembles the final clinical report.

A2A Protocol

FHIR MCP Server

Exposes medication, allergy, condition, and lab data from FHIR R4 servers as standardized MCP tools.

MCP Server FHIR R4

Drug Interaction Agent

Analyzes medication combinations for drug-drug conflicts, duplicate therapies, and allergy cross-reactivity.

A2A Agent

Clinical Guidelines Agent

Validates dosing against evidence-based guidelines with renal adjustments and age-based considerations.

A2A Agent

SHARP Context

Propagates patient identity and FHIR server credentials securely across the entire agent chain.

SHARP Extension

Prompt Opinion

Published to the marketplace for discovery and invocation by healthcare organizations on the platform.

Marketplace

Patient Information

—

Post-surgical patient with complex polypharmacy

Summary

Conflicts Identified

Reconciled Medication List

Status	Medication	Current Dose	Prescriber	Action / Reason

Recommendations

Agent Audit Trail

Clinical Disclaimer: This report is generated by an AI-based multi-agent system and is intended for clinical decision support only. All findings and recommendations require review and validation by a qualified healthcare professional before implementation. This system does not replace clinical judgment.