Ringholm-Logo Ringholm Ringholm page header

Frequency of use of HL7 message types

Publication date: Jul 24, 2012

If we were to create a diagram that depicts the frequency of the number of message types supported by software applications, what would it look like?

This is a relatively difficult question because of the lack of reliable data. After conferring with a number of knowledgeable implementers the shape of the diagram (not to any particular scale) is as follows:

Diagram with two peaks, camelhumps.
Diagram #1

HL7 version 2

The diagram shows that most applications support a certain number of HL7 v2 message types (or rather, trigger events). There are some applications that only support one or two message types, and there are a lower number of systems that support a very large number of message types.

The curve is not a Gaussian Bell curve: the two peaks in the curve reflect the fact that most applications are either ancillary/departmental applications (which support relatively few message types) or HIS/PAS and EHR applications (which have to support a larger number of message types). The value of X (as shown in the diagram) probably lies between 4 and 6, and Y between 15 and 25.

For an implementer the main element of re-use in HL7 version 2 is the segment. So what's the frequency of the number of segments supported by software applications?

Diagram with 1 peak, a skewed bellcurve.
Diagram #2

Even a simple application that receives ADT messages has to support (however minimalistically) at least 3 segments (MSH, PID, PV1). The mean value probably lies around 15.

HL7 version 3 messaging

Note: this section focuses on HL7 version 3 messaging, and doesn't include CDA.

Diagram #1 shows that most applications support a certain number of HL7 v3 message types. There are some applications that only support one or two message types, and there are a lower number of systems that support a very large number of message types.

Given that HL7 version 3 messages tend to be used to support inter-organizational workflows (e.g. regional and/or national data exchanges) we again have two peaks: most applications are either used to support a smallish part of the workflow (and support relatively few message types) or they provide some kind of EHR-like functionality (a large number of message types).

The main element of re-use in HL7 version 3 messaging is either the message type itself (the R-MIM), or a RIM class. HL7 version 3 is based on a methodology (called MDF 99) whereby the message types are object oriented specializations of the reference information model (RIM).

So what?

However interesting these diagrams may be, the big question obviously is what conclusions one should or could draw from them, especially when it comes to the implementability of HL7 version 3 (and CDA). This will be the subject of an upcoming column.

-Rene

PermaLink to this page: http://www.ringholm.com/column/HL7_message_types_per_system.htm

Index of columns:


About Ringholm bv

Ringholm bv is a group of European experts in the field of messaging standards and systems integration in healthcare IT. We provide the industry's most advanced training courses and consulting on healthcare information exchange standards.
Rene's Column (English) Rene is the Tutor-in-chief of Ringholm.
[e-mail]