As healthcare organizations look to leverage agentic artificial intelligence technologies, Wolters Kluwer Health is developing the foundational capabilities to support AI agents for medication workflows and automating the prescription process.
The company rolled out Medi-Span Expert AI to provide "AI-ready" expert-curated medication for digital health tech developers. The company is offering a model context protocol (MCP) server to select AI developers designed to cut customer development timelines and complexity by connecting third-party AI applications and agents with Medi-Span content.
Medi-Span offers a suite of embedded drug databases and data solutions to support medication-related decisions across healthcare settings from hospital and emergency room settings to pharmacy dispensing systems. Medi-Span is used extensively by pharmacy benefit managers, health insurers, retail pharmacies, electronic heath record vendors, digital health tech and healthcare organizations to support formulary management, medication safety, claims processing and drug pricing decisions, according to Christian Hartman, vice president, product innovation, pharmacy and health technology solutions at Wolters Kluwer Health.
"What we're announcing today with MCP provides foundational capabilities for medication workflow automation. It delivers the structured, evidence-based medication intelligence through the MCP server that teams need when building and managing their own agents or automation features," Hartman told Fierce Healthcare in an interview as the company gave Fierce Healthcare a first look at the new offering.
Hartman said future releases will build on this foundation with additional logic sets and capabilities that support more ready-to-use automation behaviors, including planned agent functionality built on the same clinically governed structure.
Medication management and prescription processes are ripe for AI technologies to automate manual work, provide faster support and improve efficiency.
Industry forecasts estimate that agentic AI in healthcare will grow 40% to 45% annually and could exceed $5 billion in five years. But effective AI must be grounded in rigorously validated drug and clinical content that can be accessed within clinical workflows, according to Wolters Kluwer Health. An MCP becomes vital to connect AI agents and automation tools to evidence-based content as it acts as a standardized integration layer between AI agents and external tools. The MCP acts as a "universal translator and traffic controller combined," Wolters Kluwer Health executives wrote in a blog post.
Currently, existing integration standards for drug information, such as application programming interfaces and flat files, are not optimized for agentic AI workflows and can slow AI-driven innovation. Wolters Kluwer Health contends its MCP server from Medi-Span connects expert-curated drug content to workflows in a fast, dynamic format ready for use by AI agents.
"As AI systems are increasingly used in medication workflows, accuracy and consistency are essential given the impact these systems can have on medication decisions at scale," Hartman said. "Medi-Span Expert AI aims to address a key limitation of general-purpose large language models by providing an evidence-based medication intelligence layer designed to support safe and reliable AI-driven decision making."
Broadly, Medi-Span Expert AI is designed for digital health technology companies, health IT vendors and organizations that are building AI-driven workflows, Hartman noted.
Initial applications could include AI-ready medication lookup and reconciliation, drug interaction and duplicate therapy screening, medication order validation and patient-specific medication information retrieval, Hartman said. Other possible applications could include formulary and benefit management, pricing and contracting, inventory and supply chain optimization.
"These capabilities can give AI agents access to trusted, evidence-based Medi-Span drug intelligence. We can enable this through structured, machine-readable interface such as MCP, with the aim of safer and more reliable automation of medication decisions," he said.
The initial group of developers with access to the MCP consists of "innovation partners and early adopters" that are actively building agentic or AI-driven medication workflows, according to Hartman.
"These developers are working on use cases such as medication management, clinical decision support, pharmacy automation and patient engagement, where safe and deterministic medication intelligence is critical," he said.
Wolters Kluwer is investing deeply in AI to supercharge its clinical decisions support tools. Back in October, it launched UpToDate Expert AI as a generative-AI-powered version of the widely used clinical decision support solution that's been on the market for 30 years.
"Wolters Kluwer understands both the technology and patient care side of medication. Working with Wolters Kluwer makes it possible to access evidence-based content that is backed by over 30 years of knowledge and experience, coupled with robust AI architectures," said Jean-Claude Saghbini, chief technology officer and president of health tech company Lumeris.
Editor's note: This story has been updated to accurately attribute comments and quotes to Christian Hartman, vice president, product innovation, pharmacy and health technology solutions at Wolters Kluwer Health.