OBX segment / ObservationValue items separator

[adellieux]

Hi,

i'm new to HL7, and am starting using Clear-HL7 classes, using version 2.1.1

I have a question about OBX segment and its property ObservationValue.
I have to conform to an implementation guide and the requirement is to add five strings to ObservationValue array.

The problem i am facing is about the separator used by ObxSegment with ObservationValue.
Here is some sample code i use :

        {
            SetIdObx = (uint?)GetNextObxSequenceNumber(),
            ValueType = _helper.EnumToCode(CodeValueType.EncapsulatedData),
            ObservationValue = [
                "source app",
                _helper.EnumToCode(CodeTypeOfReferencedData.MachineReadableTextDocument),
                _helper.EnumToCode(CodeSubtypeOfReferencedData.XmlExtensibleMarkupLanguage),
                _helper.EnumToCode(CodeEncoding.Base64),
                "xyzxyzxyz" // base 64 encoded document
            ],

and the resulting HL7 string :

OBX|1|ED|||source app~TEXT~XML~Base64~xyzxyzxyz||||||

From the guide i have to conform to, here is a sample of what is awaited :

OBX|1|ED|||^text^XML^Base64^RG9jdW1lbnQgbWVkY2lhbCBhdSBmb3JtYXQgQ0RBIG5pdmVhdSAx|||||||

As you can the see, the separator is not the same.
In my case, FieldRepeatSeparator is used, and in the guide, ComponentSeparator is used.

What is the proper way of achieving this ?
As a workaround, i will use Join and ComponentSeparator and use a single string.
Any advice is welcome ;-)

[davebronson]

Hi @adellieux

OBX segment slot 5 is Observation Value, like you already have, which is classified as "Varies" data type. So it can be a collection of any strings, including (of course) the string output(s) from any other HL7 type. I think what they're doing in the guide you have is using something like a CodedWithExceptions in that slot. Here's an example using CodedWithExceptions that should get you started.

Notice the .ToDelimitedString() which is used to output the delimited string from that CodedWithExceptions.

Message outboundMessage = new()
{
    Segments = new ISegment[]
    {
        new ObxSegment()
        {
            SetIdObx = (uint?)GetNextObxSequenceNumber(),
            ValueType = _helper.EnumToCode(CodeValueType.EncapsulatedData),
            ObservationValue =
            [
                new CodedWithExceptions()
                {
                    Text = _helper.EnumToCode(CodeTypeOfReferencedData.MachineReadableTextDocument),
                    NameOfCodingSystem = _helper.EnumToCode(CodeSubtypeOfReferencedData.XmlExtensibleMarkupLanguage),
                    AlternateIdentifier = _helper.EnumToCode(CodeEncoding.Base64),
                    AlternateText = "RG9jdW1lbnQgbWVkY2lhbCBhdSBmb3JtYXQgQ0RBIG5pdmVhdSAx"
                }.ToDelimitedString()
            ]
        },
   }
}

Outputs:
OBX|1|ED|||^TEXT^XML^Base64^RG9jdW1lbnQgbWVkY2lhbCBhdSBmb3JtYXQgQ0RBIG5pdmVhdSAx

[adellieux]

Hi @davebronson ,

thank you for your quick answer.

That makes sense : even if the ObservationValue has a ValueType = EncapsulatedData, its content seems to be CWE.
from the implementation guide, other OBX segments either have a ValueType = CE or CWE (strangely CE for Coded Entry).

About CE, i can see that it is deprecated ("withdrawn" as mentioned in 2.A.6 WITHDRAWN (CE - coded entry), here: https://www.hl7.eu/refactored/ch02a.html)
Do you think i should use CNE for ObservationValue when ValueType = CE ? (from the source, they both seem to be identical except for naming).

[davebronson]

You're welcome!

CWE and CNE are the replacements for CE, so one of those is what I'd use, yes. The 6 slots of CE match up nicely with the first 6 slots of CWE and CNE too, so it's simple to make that switch. And the output should be the same, all else being equal.

[adellieux]

Hi
sorry for this thread created here
feel free to move it to Discussions ;-)