Download - Validating and Transforming HL7 CDA Documents for i2b2 Import · 4. Majeed RW, Rohrig R. Automated realtime data import for the i2b2 clinical data warehouse: introducing the HL7 ETL

Transcript
Page 1: Validating and Transforming HL7 CDA Documents for i2b2 Import · 4. Majeed RW, Rohrig R. Automated realtime data import for the i2b2 clinical data warehouse: introducing the HL7 ETL

Contact:

Tingyan Xu · [email protected] · Division Medical Informatics · Dept. Of Health Services Research

VI - School of Medicine and Health Services · Carl von Ossietzky University Oldenburg · 26111 Oldenburg

Validating and Transforming HL7 CDA Documents for i2b2 Import Tingyan XU1, Volker S. THIEMANN1, Rainer RÖHRIG1 and Raphael W. MAJEED1 1) Department of Medical Informatics, Carl von Ossietzky University of Oldenburg

www.uol.de/Medizininformatik

The AKTIN Project

We are building a decentralized emergency care research IT-

Infrastructure to ensure patient privacy and comply with the many

and varied privacy laws in Germany. (Refer to [1] for more details.)

Structure

Each participating hospital will store their own data in a data

warehouse located on its own premises. The data originates from

the emergency room protocol (ER note) as designed in [2]. The

protocol is documented in HL7 v3 Clinical Document Architecture

(CDA). The data warehouse is based on the open source clinical

data warehouse system i2b2. Customized configurations and

additional tools are installed to validate and import the ER-CDA

documents, organize the centralized research queries and manage

the answers from each participating hospital.

No. 01KX1319B

References

1. Ahlbrandt et. al. Balancing the need for big data and patient data privacy--an IT infrastructure for a decentralized emergency care research database. PMID: 25160287.

2. Kulla M et. al. Nationaler Datensatz "Notaufnahme": Entwicklung, Struktur und Konsentierung durch die Deutsche Interdisziplinäre Vereinigung für Intensivmedizin und Notfallmedizin.

DOI: 10.1007/s00101-014-2295-x.

3. Pecoraro F et. al. Data Warehouse Design from HL7 Clinical Document Architecture Schema. PMID:26152975. A conceptual framework to design a dimensional model based on the HL7 Clinical Document

Architecture. PMID: 25160190. Designing ETL Tools to Feed a Data Warehouse Based on Electronic Healthcare Record Infrastructure. PMID25991292.

4. Majeed RW, Rohrig R. Automated realtime data import for the i2b2 clinical data warehouse: introducing the HL7 ETL cell. PMID: 22874194.

Transmission and Validation

We offer two standard interfaces for transmitting the CDA

documents: Via the IHE XDS.b document repository for

SOAP API calls or with HL7 FHIR binary endpoint for

RESTful data transfers. All incoming documents have to be

validated before further transformation and usage. The

automatically generated Schematron rules from the CDA-

ER specification (available at http://aktin.art-decor.org/)

are used for this purpose, ensuring only valid documents

of this one specification are loaded into the DWH.

Transformation

After validation the complex CDA XML structure of the incoming

documents have to be broken down into the key-value (also Entity-

Attribute-Value, EAV) format to fit the i2b2’s database structure.

Although not a new function, we need a configurable function in

this case, designed and implemented for the AKTIN environment,

to transform from ER-CDA to i2b2: Some facts in our case may be

empty, others may not be necessary in the researching context and

can be removed. Documents can be retransmitted to update the

data in i2b2.

Results and discussion

Although there are other ETL tools and methods to load

data from CDA into a DWH [3], our approach is highly

adaptive and we provide a specified interface and

software to handle the CDA documents and manage the

i2b2 DWH.

Health Exploring Complexity 2016 28th August – 2nd September 2016 - Munich