eHOP is a clinical data warehouse designed to consolidate and reuse real‑world hospital data for research, quality, and care optimization. It orchestrates acquisition, transformation, and integration from heterogeneous HIS sources (e.g., HL7, HPRIM, PN13, LOINC, documents) into a unified store, and provides tools to organize and query both structured and unstructured data. Deployments described in the literature run on virtualized infrastructure with an Oracle database integrating patient demographics, encounters, diagnoses, labs, clinical notes, and medication administration, exposed through eHOP query services and regulated data marts for controlled, time‑limited investigator access. Governance and GDPR compliance are supported through dedicated modules and procedures for consent/opposition management, pseudonymization, access traceability, and de‑identification workflows aligned to institutional frameworks.

Key Features

Heterogeneous data acquisition and integration Consolidates multi‑source hospital data into a single warehouse.

• Ingests HL7, HPRIM, PN13, LOINC, and document formats (Word, PDF, CSV, text)
• Automated, scheduled, and monitored feeds via interoperability stack

Comprehensive clinical domain coverage Enables reuse across care and research use cases.

• Tables/resources spanning patients, consultations, diagnoses, labs, notes, and inpatient drug administration
• Simple textual and structured query tools for investigators

Regulated data marts and controlled access Delivers parsimonious, time‑limited datasets.

• Data marts published on hospital intranet with de‑identified data and scoped cohorts
• Investigator self‑service queries within governed environments

GDPR‑aligned governance and security Built‑in controls for privacy and traceability.

• Opposition management module excludes opposed patients at datamart creation
• Pseudonymization with collision‑free identifiers; per‑study re‑pseudonymization; robust access logging

De‑identification of unstructured data Progressive NLP/ML for documents.

• Effective patient data de‑identification; ongoing improvements for clinician data with high F‑scores (0.96–0.99) in testing

Scalable technical architecture Proven hospital‑grade deployment model.

• Virtual machines with Oracle database integrating multiple sources
• Operated with enterprise interoperability (Enovacom Suite lineage)

Benefits

Accelerated research and real‑world evidence generation Faster cohort discovery and study execution.

• Consolidated, queryable warehouse reduces time to access multi‑source data
• Parsimonious, de‑identified data marts streamline IRB‑governed projects

Improved data quality and standardization Reliable inputs for analytics and AI.

• Standardized ingestion across diverse formats and codesets
• Continuous stewardship for integrity and compliance

Stronger privacy, compliance, and trust Operationalizes GDPR within workflows.

• Opposition handling, pseudonymization, and access traceability embedded in the CDW
• De‑identification pipelines for documents to expand compliant reuse