Provides an anonymized integration layer for e-SUS PEC APS data with Large Language Models.
Sponsored
This Model Context Protocol (MCP) server enables secure and optimized querying of clinical data within the Prontuário Eletrônico do Cidadão (PEC) database. Built with FastMCP, it exposes specialized tools for Large Language Models (LLMs) to interact with the PostgreSQL-based PEC data in a read-only fashion. The server facilitates anonymized searches for patients, analysis of health conditions, tracking of attendance histories, and generation of key health indicators, all while enforcing data security through anonymization, read-only access, and query limits.
Key Features
01Anonymized patient search by initials, age, sex, and health unit
02Health condition analysis and patient counting by CID-10/CIAP codes
03Metadata lookup for CID-10/CIAP codes and registered health units
040 GitHub stars
05Retrieval of historical SOAP attendance records for specific patients
06Monitoring of population health indicators (e.g., patients without recent consultations)
Use Cases
01Monitoring population health indicators for public health management and strategic planning
02Empowering LLMs to safely query and analyze clinical patient data for insights
03Developing AI applications that require secure, anonymized access to electronic health records
