HL 7 s National Library of Medicine NLM Contract

HL 7’s National Library of Medicine (NLM) Contract Introduction Three year contract with 2 major projects: 1. Develop and apply methods to ensure UMLS Meta. Thesaurus is in step with HL 7 Vocabulary Standards 2. Develop and pilot test an HL 7 Implementation Guide for the transmission of patient information between disparate electronic health record (EHR) systems using HL 7 Messaging Standards 27 January 2005 Copyright 2005, HL 7 1

Project Management Team • Management: – Phil Barr, HL 7 Project Manager – Chris Chute and Stan Huff, Co-leaders for Vocabulary – Bill Braithwaite and Ed Hammond, Co-leaders for EHR • Work: – voluntary and contract personnel using the approved HL 7 Policies/Procedures – results will feed into the HL 7 standards creation processes 27 January 2005 Copyright 2005, HL 7 2

Vocabulary Project - briefly • Make HL 7 vocabularies available from within the Universal Medical Language System (UMLS) Where possible replace HL 7 vocabularies with designated Consolidated Health Initiative (CHI)* vocabularies Specify vocabulary bindings from specific HL 7 messages to the UMLS (first v 2. x, then v 3) • • – – Vocabulary access, maintenance and collaboration tooling will also be developed as required for the project. All activity will be coordinated using established HL 7 processes and documented in implementation guides. * Specified by (US) Departments of Health and Human Services, Defense, and Veterans Affairs 27 January 2005 Copyright 2005, HL 7 3

EHR Project - briefly • Develop implementation guides that: • enable the transmission of patient information between disparate electronic health record (EHR) systems. • maintain clinical meaning of information in computable form. – leverage existing work within HL 7. – increase the utility/uniformity/usability of HL 7 messages. – each successive task producing a useable working product. 27 January 2005 Copyright 2005, HL 7 4

Two Phases • Phase I will produce – a short-term solution that will allow implementation of information flow between EHR systems – using existing HL 7 formats in a standardized, simple, queryresponse message set – including instructions and tools for implementation of a secure transmission mechanism • encompassing encryption, authentication, and transport verification. • Phase II will expand Phase I – messages will enable • more semantic interoperability and • a richer query set. 27 January 2005 Copyright 2005, HL 7 5

EHR Project – Tonight’s Agenda • Project Description – Bill Braithwaite • Phase I Status Report – Liora Alschuler • Reference Table of Potential Data Elements – Ed Hammond • Related HL 7 Activities – Dan Russler, Michael van Campen • Discussion of Next Steps (RFI & RFQ) 27 January 2005 Copyright 2005, HL 7 6

EHR Project Description Technical Co-Leaders Bill Braithwaite Ed Hammond HL 7 Project Manager Phil

Additional Work • Formation of a stakeholder focus group • Identification and analysis of various alternative local and international models for the exchange of EHR data • Definition of a clinical data element dictionary • Live data pilot project • Balloting of new artifacts as necessary • Implementation guides will be prepared for all functional activity. 27 January 2005 Copyright 2005, HL 7 8

EHR Project: Phase I Create a simple, 2 step Query/Response between 2 EHR systems: 1. Query for the type of information available on a given patient… • Response describes what type of information is available. Followed by 2. Query (by type or date range) for any of the available information… • • Response uses CDA to ‘wrap’ the information ‘payload’. Provides “varying” levels of granularity, from (required) “text” to (optional) HL 7 structured data in any HL 7 format. Includes secure, verifiable transmission over the internet To be demonstrated by February 2005 27 January 2005 Copyright 2005, HL 7 9

EHR Project: Phase II Adds additional specificity at all levels: – robust instance identification paradigms for all objects (e. g. patients, practitioners, documents, and origin of all information) – allows (arbitrarily) greater specificity in query and response definitions (what information is requested/what is available) – Uses CDA release 2 with clinical statements to support variable levels of semantic interoperability (from plain text to fully structured clinical information) depending on the capabilities of the ‘record holder’ system 27 January 2005 Copyright 2005, HL 7 10

Phase II: Review Contract • Review current EHR information exchange implementations – GP 2 GP, open. EHR, CCR, NICTIZ (National ICT Institute for Healthcare in The Netherlands), med. Emed (a project of the Net. Lab group in the Department of Computer Science, University of Victoria), CEN 13606, CDA release 2, and other closely related initiatives. – High level and a detailed analysis and comparison of an agreed upon set of such implementations to inform the progress of the NLM-EHR project. 27 January 2005 Copyright 2005, HL 7 11

Phase II: Review Contract • Tasks – Summarize the use case for each model. – Identify the level of negotiation and or built-in assumptions required for successful information transfer between participants to occur. – Compare process, tools, and methods used to create each model. – Compare handling of unstructured documents. – Compare higher level constructs in each model such as episode-of-care. – Compare interactions in the model such as patient roles, order & observations roles. – Compare ability of each model to handle the range of data elements specified by the ASTM Continuity of Care Record (CCR). – Compare model maturity – Is it currently in production? How long? Current tooling level supporting model? – Compare implementations where they exist. – Gather Implementation guides for each project. – Explore issues encountered in comparison. – Summarize disagreements between proponents of each model. 27 January 2005 Copyright 2005, HL 7 12

Phase II: Message Set • Develop Phase II EHR Query-Response Message Set – Using the model-based HL 7 Development Framework (HDF), ensure that the information exchanged enables as much semantic interoperability as possible. – Design a message set that can carry full semantic information of as much clinical information as possible and produce an implementation guide and tool set complete enough to enable reliable implementation by a broad range of vendors. 27 January 2005 Copyright 2005, HL 7 13

Phase II: Message Set • Tasks – Identify issues beyond the raw data transactions that need to be addressed for implementation (e. g. , patient ID, document ID, ensuring data integrity, ensuring system security, documenting origin of info, etc. ). – Explore the query-response capabilities of existing HL 7 messages to determine modifications to allow the Phase II query-response messages to be more complete as to the kinds of patient data that can be requested and delivered. – Generate specifications for Phase II query-response messages. – Review results of contract for ‘Current Implementation Review’ and incorporate into a specification for the Phase II delivery mechanism. – Design a standard HL 7 Version 3 ‘payload’ message to deliver an arbitrary number of clinical statements, each of which is able to handle a variable level of semantic interoperability from plain text to fully structured and coded clinical information, based on the capabilities of the record holder. – Draft an implementation guide for transmitting all the data and documents comprising an entire electronic health record (EHR) between systems using these messages, independent of source and destination architectures. – Using the implementation guide, create a test implementation to provide Proof of Concept of the Phase II model by transferring a subset of meaningful clinical data, which demonstrates the feasibility of variable semantic interoperability between independent systems and refines the implementation guide and tool set. – Submit any additional or changed standards through the HL 7 standards approval process. 27 January 2005 Copyright 2005, HL 7 14

Phase II: Message Set • Looking for best way to break this contract up into 2 -3 smaller parts each of which would extend the previous part but produce complete deliverables on it’s own. • Goal is to make rapid progress, limit risk to the project and subcontractors, and reuse and build on existing development. • Comment period of three weeks, then formulate and post the next EHR Query-Response Message Set RFQ(s). 27 January 2005 Copyright 2005, HL 7 15

Phase II: Message Set • Possible ideas include: – First implement subset of data, e. g. , medications, labs, and active problem list. – Consider frameworks other than CDA release 2. – Incorporate examples of use of services (e. g. , terminology, patient identification) into message set. – Expand the phase I query-response message set to be more complete as to the kinds of patient data that can be requested and delivered. – Expand phase I query-response message set as necessary to deal with indexed information/documents. – Separately identify issues beyond the raw data transactions that need to be addressed for implementation (e. g. , patient ID, document ID, ensuring data integrity, ensuring system security, documenting origin of info, etc. ). 27 January 2005 Copyright 2005, HL 7 16

More Information … For more information/to declare interest: – Subscribe to contractwork@lists. HL 7. org – Refer to www. HL 7. org/NLMcontract/requestforquote. cfm • contains a description of the projects and the sub-tasks to be subcontracted to experts. Contacts • Phil Barr – pbarr@hl 7. org • Bill Braithwaite – Bill@Braithwaites. com • Ed Hammond – hammo 001@mc. duke. edu 27 January 2005 Copyright 2005, HL 7 17