Diacritical marks in HL7

[Ken Seachrist]

Has anybody had the “opportunity” to have to send diacritical marks across an interface?  I think I can get them to go across, but I believe the issue will be with the downstream systems!

Topic

[Elisha Gould]

Most systems wont handle them correctly and need to ensure they are encoded for the system they are being sent to.

The “ideal” way of handling them is:

set MSH-18 Character Set to the encoding scheme being used.

escape the characters using the C or M escape

See “Escape sequences supporting multiple character sets for PN, XPN, XCN,

XON, and XAD data types” in the hl7 spec for details.

If the application can’t handle escaped characters properly, then you’ll need to send them in whatever encoding scheme they need it in.

Alternatively use a mapping table to translate them to ASCII characters.

Be careful with extended characters, as cloverleaf uses utf-8 by default, so it may change the characters unless no manipulation is done and the thread is set to binary.

[Ken Seachrist]

Elisha,

  Thank you for the great info!  I agree that it will mostly depend on the downstream system, but since we generally use the doctor number as the key, we may be able to get away with using these in the name field.

  I am going to experiment with the control characters and see how it looks coming out of Cloverleaf.  Thanks again…and …..Oy!  Oy! Oy!   😀

[Author]

Replies