If there's any industry that stands to benefit from IT-enabled insights, it's healthcare. However, traditional methods of analyzing healthcare operations data are costly, rigid, and complex. Clinical and operations data must first be logged in applications such as Cerner, Epic, PeopleSoft, LabMed, and Allscripts, and then cleansed and fit into a standard data model. When a change is required, database experts can require weeks or months to adjust the data model and prepare new reports.
It doesn't have to be this hard! By analyzing HL7 transactions on the wire and sending relevant metrics to an unstructured data store such as MongoDB, healthcare organizations can gain full insight into the wealth of clinical and operations data produced by their applications.
Tomorrow, we'll be hosting a webinar featuring Wes Wright, CIO of Seattle Children's Hospital, that will explain the opportunity for HL7 analytics. Wes will talk about how Seattle Children's uses HL7, including for communications between applications such as Epic and Cerner and a new pediatric hospital information network. Erik Giesa, ExtraHop SVP of Marketing and Business Development, will also be on hand to explain how this wire data can yield tremendous insight for patient flow, fraud detection, epidemiology, and other areas.
UPDATE: The webinar is now over, but you can watch the recording below or view the slides on SlideShare.
Why HL7 as the Data Source?
The HL7 protocol is the most widely implemented standard for exchanging clinical data between systems—everything from patient administration, financial management, and materials management to patient care, clinical laboratory automation, and scheduling. The HL7 communications between various healthcare applications contain all of this, and it represents a rich trove of information that could potentially be mined for more intelligent healthcare operations. For all the possibility, there has not been any technology built to analyze these communications between systems. The best that healthcare organizations can do currently is to either attempt to pull HL7 data from proprietary interfaces offered by gateways and interface engines or extract the information from every single application database and store that data in a structured relational database elsewhere for analysis. This approach requires significant work on the part of the IT department, does not necessarily cover all the HL7 communications present in an environment, and is limited to the information that various gateways, interface engines, and databases are set up to either log or compile through a rigid and delayed extraction process. For example, this approach would not likely be able to capture data contained in free-form fields within HL7 transactions.
Wire Data Analytics for HL7
New wire data analytics technology can be applied to this issue. Using a passive network appliance with a real-time stream processor to analyze HL7 messages and transactions
on the wire, healthcare organizations are able to mine, visualize, trend, and alert on any attribute contained within every HL7 transaction regardless of its source or destination. Unlike traditional solutions that are limited to a preconfigured subset of data, this wire data analytics approach can include anything contained in the HL7 transaction payload including acknowledgements, status codes, data contained in the primary and sub-fields, and even data contained in free-form fields. In short, any information that is separated by a delimiter in the HL7 transaction can be parsed on the wire with this method. This method can be applied to other data-interchange protocols used in healthcare, such as the X12 protocol. X12 is basically just XML, a messaging format, which rides on top of the HTTP wire protocol. ExtraHop has several customers that currently analyze X12-formatted EDI feeds over HTTP to extract transactional information that exists within these messages.
Who Does This Apply to, and How to Get Started?
Every large healthcare organization can benefit from HL7 analytics technology. IT organizations that are already invested in Big Data technology, especially NoSQL databases or other unstructured data stores, should augment their solutions with real-time HL7 metrics captured off the wire as the integration is trivial and the potential return is significant. Healthcare IT organizations
that are responsible for the performance of HL7 gateways and interface engines should also deploy wire data analytics technology to reduce mean-time-to-resolution for performance issues and proactively resolve issues before they impact patient care or operations. Watch the video below to see how MEDHOST uses HL7 analytics technology.