Data-Integration
HL7 DATA INTEGRATION IN A HYBRID CLOUD & ON PREMISE ENVIRONMENT
THE CUSTOMER
The customer is among the nation’s largest non-profit, academic, healthcare networks, providing coordinated clinical and hospital patient care, all electronically connected to provide convenience for its patients, across its network of 40 owned, managed and affiliated hospitals and specialty hospitals and more than 100 health centers and urgent care centers.
Close collaboration between its clinicians and scientists helps to bring innovative medical discoveries from the laboratory to its patients’ bedsides. The customer has been recognized by U.S. News & World Report as “Best Hospital” in two specialty categories and is a national leader in medical research and among the largest non-university-based physician training centers in the U.S, conducting over 700 clinical research studies.
THE PROBLEM
The customer experienced an unacceptable large lag in the most recent patient data becoming visible to its Help Desk personnel. This caused frustration with patients as well as with its Help Desk personnel. As a consequence the customer faced low patient satisfaction as well as low Help Desk personnel morale and productivity. A major reason for the lag was the time it took to ingest and process the HL7 patient data files for access by the Help Desk cloud-based application.
THE CHALLENGE
The main challenge to solving the problem was to perform the ingest and processing of new HL7 patient data files in not more than 5 minutes from the time the new data files became available. This time frame included ingesting the HL7 data files, splitting the data into subsets while maintaining inherent hierarchical relationships, and culling duplicate data records in order to maintain data integrity.
THE SOLUTION
The customer selected PartnerBO among a set of larger competitors because of PartnerBO’s subject matter expertise, technical expertise and deep knowledge in data integration, and building on-premise, cloud and hybrid data management solutions with leading enterprise cloud data management tool provider Informatica.
We approached the problem systematically, starting with a joint systems analysis with the customer, design of a dataflow and data processing architecture, implementation of the architecture and finally, testing, verification, bug resolution, performance fine tuning and knowledge transfer.
A high-level diagram of the architecture is shown below. The HL7 data file ingest is triggered by the arrival of a new HL7 patient data file in a local landing folder. The file is ingested by transferring it into a processing staging location which in turn triggers the data processor service running on Informatica PowerCenter , which splits the data records into clean sub data set records with no duplicate data records present, next parses and maps the data records according to a dictionary of specific events with all hierarchical relationships preserved, and finally, uploads the processed data to the cloud-based Help Desk platform.
For the implementation we choose Informatica Intelligent Data Platform version 10.2, with Informatica PowerExchange for integration with the Help Desk cloud platform. The Informatica Intelligent Data Platform provides outstanding data integration agility, data quality assurance and enterprise scalability. The platform ran on the customer’s premise for maximum control over patient data record protection.
| KPI | Previously | Target | Result |
|---|---|---|---|
| Consultant performance | N/A | 6 weeks | < 3 weeks for original scope of work Implemented additional HL7 data formats in remaining time |
| Latency | Over 10 Min | 5 minutes max | < 4 minutes. |
| Implementation | Hard to maintain | Future proof | Easy to add additional and future HL7 data formats. |
THE RESULTS
Through the use of the Informatica Intelligent Data Platform, Informatica PowerExchange, and the Informatica PowerCenter data processing B2B libraries, we managed to deliver the fully implemented and tested product to the customer in less than half the time that was allotted by the customer. Together with the customer we were therefore able to expand the scope to include additional HL7 data formats to be ingested and processed for access by the Help Desk platform.
Access to the latest patient information became near real-time, with every new HL7 patient data file now being ingested, processed and loaded to the Help Desk cloud platform in under 4 minutes; thereby beating the target of 5 minutes that was set by the customer.
Finally, our implementation achieved the secondary, but nonetheless important goal of being future-proof by keeping the Informatica Data Processor source code fully reusable. This allows the hospital to implement additional and future HL7 data formats as needed by simply replicating the existing Informatica PowerCenter mappings.
CONCLUSIONS
We were able to implement a fully scalable solution that exceeded every requirement, beating target latency of 5 minutes by delivering less than 4 minutes latency, beating allotted delivery time by completing the original scope in under half of the allotted time, and engineering the solution to be fully future-proof through full code reuse to support adding additional and future HL7 data formats.
All this was possible because of the combination of the powerful capabilities of the multithreaded Informatica Inteligent Data Platform, Informatica PowerCenter and Informatica PowerExchange products on the one hand, and on the other hand because of PARTNERBO’s expertise and experience architecting and implementing flexible and high-performance data integration solutions, as well as our expertise and experience developing with the Informatica suite of data management and integration tools.
___________________________________________________________________________________________________________________________________________
About PARTNERBO
PartnerBO is a data management and analytics System Integration & Solution hospital, transforming the way companies manage and leverage data. Through strong partnerships, innovative and purpose-built solutions, PartnerBO helps companies in their data integration journey.
At PartnerBO we support your growth through System & Data integration, focusing on increasing your strength from inside your organization. We help you empower yourself with new capabilities such as data quality, data management, cloud data integration, big data architecture, hospital-wide data quality governance, data security, and data archiving.
Our Intelligent Data Platforms are a powerful combination of metadata intelligence and artificial intelligence, fully integrated across all aspects of data management. PartnerBO helps you in your journey to leverage cloud, big data, data management, governance, archiving transformations, data asset monetization and data-driven decisions, for your business growth goals.
Visit us at www.partnerbo.co or contact us via email at sales@partnerbo.co
