HL7 news

-> -> ->

New Access Database Lets You Take Control of HL7 Standards

By Frank Oemig and Larry Stotz

Now interface developers can harness the power of HL7 with the newly developed HL7 Access Database.

Developed by Frank Oemig of HL7 Germany and now being marketed by HL7, the new database system is designed to support the application of the Standard and assist with the interface process.

A "must-have" for every software vendor actively developing HL7 compliant systems, the database is packaged in different modules that can be purchased to fit the particular needs of developers, is comprehensive and user-friendly. Developers can choose a database that represents a single version of the standard (V2.1, V2.2, V2.3 or V2.3.1); a database that represents two consecutive versions; or the comprehensive, four-version edition of the database.

The Access Database lets you take control of HL7 data by providing

A user-friendly GUI interface that ensures ease of use by technical and non-technical staff alike

Powerful search capabilities that provide quick and easy access to trigger events, message structures, segments, data elements (and their components), data types (and their components) and table values (both HL7 and user).

A single-keystroke report generator that allows you to see how the data are linked (e.g. which segments comprise a particular message, to which data elements comprise a particular segment).

Reusable HTML reports of the trigger events, message structures, segments, data elements, data tpyes and tables and their values.

An optional copy of the Standard(s) in HTML format.

A user's guide.

The Dream of an HL7 Database

The new HL7 Access Database is the result of Oemig's dream to provide a comprehensive HL7 database which contains the complete set of information about the HL7 Standard. Oemig envisioned a database that would contain information about events, segments, data elements, tables and their values, as well as message. Putting such a database together proved challenging because all of the relevant information was hidden in natural language text and ill formatted.

Fortunately, however, the standard contained an appendix containing a set of tables, which Oemig was able to use to compile his database.

"At least it was a good starting point to build up such a database" Oemig explains. "But during the implementation process it became apparent that a lot of inconsistencies were hidden within the different chapters of the standard."

Dealing with the inconsistencies was the next major hurdle. With version 2.2, the standard got bigger and so did the number of inconsistencies, although some of the already detected ones were eliminated. The easiest way to face all those problems was to enhance the database to contain both versions - 2.1 and 2.2 - and if feasible also to include a mapping between the different data elements. And it would be great if one could get a possibility to compare/check for differences. Using SQL on the database appeared to be the best solution. So the database was refined according to the new meta model of the HL7 Standard and to maintain a mapping table for the data elements from 2.1 and 2.2.

Because of referential integrity, data can only be entered into the database if it is consistent. So if the standard hides inconsistencies, the newly created database would expose them, a feature that has enhanced development of the standard.

The next major hurdle was determining how to make the huge amount of data accessible to users and how best to represent all of the different views.

With the release of Version 2.3, a VBA module was added to generate HTML files out of the database, which allowed for hyperlinking between documents for quick browsing. According to Oemig, having all this information available in a structured form as well as HTML representation in integration of the original documents is the highlight of this development process.

"The tables in this database - i.e. the meta model of HL7 - can be regarded as a network," explains Oemig. "A network should only provide information belonging together. Everything else should be presented if necessary. The mechanism of HTML files appeared to be the best means for doing so, so I created a generator which allows for extracting the data out of this database. Any available browser can interpret these files, so a quick investigation of the data became possible.

The database contains now four different versions of the standard starting with 2.1 and ending with 2.3.1, including data elements, events, messages, segments, tables, table valus, components, data structures and a mapping of data elements among the different versions. This complete set of information can now be used to create a set of HTML files, which also allows for comparing the different segments with the previous ones in a visual form.

Oemig has also extended the editing functionality of the database to allow users to insert individual information for distinct implementations or site specific negotations. The database is open to introduce personal data, such as asdding example values or to create a new generator e.g. for XML/DTDs.

All in all, Oemig sees his database as a "dream come true." For more information,see the HL7 Catalog of Resources or visit the HL7 Bookstore on the website www.hl7.org.

Last Update : March 7, 2000