Help : Multiple segments with same address

[Jeff Manley]

I am moving some interfaces to our new 6.0 server.  Watching the log, I see these new warning that I have never seen before.  

Here is the warning message:

WARNING: Multiple segments with same address were found in message format ‘2.3 epic ADT_A03’ in file /cloverleaf/cis6.0/integrator/live/formats/hl7/2.3/epic/ADT_A03. This may lead to incorrect parse or translation.

[msg :RecP:WARN/0:EPIC_ADT_xlate:–/–/—- –:–:–] Detailed warning information:

[msg :RecP:WARN/0:EPIC_ADT_xlate:–/–/—- –:–:–]     Segment ‘NTE’ with address ‘0(0).NTE’ found at lines 13,16

Here is the message format (from our Epic system):

MSH

EVN

PID

PD1

PV1

PV2

OBX

NTE

GT1

IN1

IN1

Attached is the XLATE.  A simple one that should not cause the issue.

Also attached is the HL7 ADT_A03 definition.

Topic

[Jim Kosloskey]

Jeff,

First of all that layout you provided from the HL/7 Configurator is not HL/7 2.3 or any other base version because non of the base version I have looked at have ANY NTE segments.

So I suspect this is a custom variant.

The reason you are getting the error is there are 2 NTE segments defined at the same level in the variant – one after the PD1 and one after the PV2. While that is legal in HL/7 the Cloverleaf parser has a problem with that.

Based on what you have provided I am guessing the messages are unaffected you are just getting a lot of log entries.

So you can make either of those NTE segments a group of one (that will change the addressing pathing for entries for the change and below). Don’t forget because you are using BULKCOPY it is likely the variants on both sides will need to match – so you might need to change 2 variants of different variants are used.

Of course that may impact other integrations as well.

Another option is to remove both the NTE segments as it appears neiither is being used.

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Jeff Manley]

A05 messages are getting the same error.  Here is a sample message from our Epic vendor.

MSH

EVN

PID

PD1

NTE

PV1

PV2

NTE

GT1

IN1

IN1

[msg :RecP:WARN/0:EPIC_ADT_xlate:10/23/2013 13:36:21] WARNING: Multiple segments with same address were found in message format ‘2.3 epic ADT_A05’ in file /cloverleaf/cis6.0/integrator/live/formats/hl7/2.3/epic/ADT_A05. This may lead to incorrect parse or translation.

[msg :RecP:WARN/0:EPIC_ADT_xlate:–/–/—- –:–:–] Detailed warning information:

[msg :RecP:WARN/0:EPIC_ADT_xlate:–/–/—- –:–:–]     Segment ‘NTE’ with address ‘0(0).NTE’ found at lines 13,17,20

So not much we can do here except have Epic change their messages(unlikely) or just deal with the warning messages?

[Jim Kosloskey]

Jeff,

In this case, make one of the NTE segments in your variant a group of one.

Then you may need to modify any and all references to the variant using address pathing – like Xlates or Tcl procs using GRM functions – as the NTE made part of a group and everything below it will likely have their address pathing changed for referencing.

Or like you indicate you can ignore the loge entries.

If Epic never uses the NTE segments indicated in reality, then it would also be possible to remove one of the NTE segments. Then the log entry disappears and most likely no address path changes will be needed. If using BULKCOPY though make sure both variants match.

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Bob Richardson]

One and All,

Running CIS 5.8.6.0 on AIX 6.1 TL7 SP4

We have Epic Systems implemented here and ran into the same problem of the NTE at the same level.  The NTE after the PD1 segment does serve a purpose:  in our shop it is used to supply information about whether or not a patient wants to decline sharing their health care information.

This is how Epic decided to pass that information.  They may be passing other information too for other customers but apparently not for you!

As indicated in the replies to this post, we did not need the second NTE and so deleted it from several variants to keep the pathing intact in hundreds of xlate files that used the variants.

In other instances with the same segment at the same logical level, we had to create a new group and then repath whatever xlate files used those variants.  Repath means to respecify the source and destination as anything below the new group gets a new group assignment.

Hope this helps on why those NTEs exist from Epic Systems.

[Jim Kosloskey]

I hate to hijack this thread but this is an Epic question (it is a possibility we may have to involve ourselves with Epic). Hoewever, I do think thei is somewhat germane to this discussion.

Is the 2.3 HL/7 version the only version Epic can produce or can it produce later (say at least up to 2.6 maybe even 2.7)?

The reason I ask is many of the HIPAA and AFCA related data have places to be in the 2.6 and beyond HL/7 versions. I would much rather utilize normative placement than use non-normative expression if possible.

Plus it is good knowing this going in.

Thanks

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Bob Richardson]

Greetings,

Epic Systems creates its own versions of the “standard” HL7 V2 variants,

It may say 2.3 but may contain bits and pieces of 2.6 – discoveries by us Cloverleaf developers.  Hence we had to scramble to retro-fit our existing 2.3 or 2.4 variants to contain customized segments/fields only “officially” defined in 2.6.

Once we upped our Cloverleaf version to 5.8 we had the 2.6 and now try as much as possible to create new HL7 V2 variants starting with a standard 2.6 variant.

Just fair warning to those looking to implement Epic Systems feeds via the Integrator.

Have fun!

[Jim Kosloskey]

Thanks Bob.

email: jim.kosloskey@jim-kosloskey.com 29+ years Cloverleaf, 59 years IT - old fart.

[Greg Tataryn]

This may be a dumb question but in 6.1.1, how can you create a group of one? We are having a similar issue with ZOE segments out of STAR. I have each one set as optional and repeating but not sure how to make it a group.

[Robert Kersemakers]

Hi Greg,

Had to try this myself and found a way to do it.

Insert the desired segment (ZOE) twice. Then select both these segments and make them repeating and optional. It will make a new group with two segments. Then select one of the segments and delete it. The group will remain, with only one segment in it.

Maybe there are other ways as well for this.

Zuyderland Medisch Centrum; Heerlen/Sittard; The Netherlands

[Author]

Replies