Public Health / Data GovernanceBaoshan Municipal Health Commission2026-07-15

Baoshan Health Commission: A City-Wide Health Data Set Across 17 Domains

A platform is only worth what it can collect. Rather than buying interfaces, Baoshan uses MOI to extract raw data directly from each institution's database or standby replica, then integrates it into the 17 business domains of the municipal shared data set.

Baoshan Municipal Health Commission

Baoshan is a prefecture-level city in western Yunnan with a resident population of around 2.4 million. Its municipal health informatization program builds a city-wide population health information platform linking the health commission, hospitals and primary care institutions.

2.4M
Residents covered
17
Shared data set domains
No interfaces
Direct standby extraction
City-wide
Hospitals and consortia

The challenge

The platform is only worth what it can collect. Data has to come out of every hospital and county medical-consortium system in the city and arrive in one shared, standard-conformant data set-but those systems belong to more than a dozen healthcare IT vendors. The conventional route is to pay each vendor to build a custom reporting interface for each system: across 41 hospital systems and seven secondary platforms, interface fees alone approached RMB 20 million-and the platform would still not control delivery schedule or data quality.

The solution

Rather than buy interfaces, the program collects the data itself. OmniFabric extracts raw data directly from each institution's database-or its standby replica, with a nightly offline pull where hospital load rules out CDC-and runs integration, cleansing and governance centrally, producing output conforming to the municipal medical service shared data set and reporting it to the city platform through one governed pipeline. Stream-and-batch unification connects heterogeneous sources (Oracle, PostgreSQL, MySQL, SQL Server, MongoDB, ElasticSearch, JSON, XML) with pipelines developed in SQL, and native multi-tenancy puts each hospital and county platform in its own tenant and resource group.

The outcome

Collecting and governing data centrally removes the need to pay a dozen-plus healthcare IT vendors for custom reporting interfaces across 41 hospital systems-fees priced near RMB 20 million.

At the Changning pilot hospital the source scope narrowed from 4,578 tables across 10 databases to 185 tables across 5, and the pipeline passed the municipal platform's quality inspection.

One governed pipeline replaces a dozen vendor-built interfaces, so the platform controls timing and data quality itself instead of waiting on each vendor.

Solution Architecture

Data sources
  • Hospital HIS, PACS & Physical Exam Databases
  • County Medical-Consortium Platforms
  • Municipal & Provincial Health Platforms
  • SQL Server / Oracle / MySQL / PG / Mongo / ES
MatrixOne Intelligence
  • Direct Extraction from Institution Databases — No Vendor Interface
  • Stream-and-Batch Unified Pipelines Developed in SQL
  • Mapping to the Municipal Shared Data Set (130 Target Tables)
  • Multi-Tenant Isolation per Hospital and County Platform
  • Metadata, Data Quality, Encryption & Masking Built In
Business applications
  • Standard-Conformant Data Reported to the City Platform
  • 17 Business Domains Across 130 Target Tables
  • Data Quality Validated at Ingest
  • Path to Level 4A Interoperability Certification

Related solutions and products