Closed ejancic closed 11 years ago
I'm really hoping I didn't just step on a land mine. I changed this on dev, but the test export does not validate. The validation tool specifically checks for SNOMED codes for problems. There's never been any talk of needing SNOMED codes for problems.
Here's the only possible parachute... If I specify no code (just the description), the CCD validates.
Thoughts?
What did you change on dev?
We should be using ICD-9 codes and not SNOMED codes for problems. There's no requirement to use SNOMED for problems. The export looks wrong because we seem to be claiming that an ICD-9 code is a SNOMED code. The CCR file that we're using identifies ICD-9 as the coding system for problems so it seems like CCD should do the same.
On Apr 2, 2013, at 10:51 AM, Skwerl notifications@github.com wrote:
I'm really hoping I didn't just step on a land mine. I changed this on dev, but the test export does not validate. The validation tool specifically checks for SNOMED codes for problems. There's never been any talk of needing SNOMED codes for problems.
Here's the only possible parachute... If I specify no code (just the description), the CCD validates.
Thoughts?
— Reply to this email directly or view it on GitHub.
Well, yes, the export specifies code system 2.16.840.1.113883.6.96, which is SNOMED, but the code in there is an ICD-9 code. To that extent you're correct.
But if I modify the XML to specify code system 2.16.840.1.113883.6.103, which is ICD-9, the validation fails. The validation tool actually explicitly makes sure that we're using code system 2.16.840.1.113883.6.96 (SNOMED).
CCR validates fine with ICD-9, imports fine, no problem there. But CCD validation seems to require either SNOMED or no code at all.
Well, for now, the CCD export properly specifies that the code system is ICD9. That commit is pretty easy to read.
Here is the error that the validator is throwing though:
Message: Error: HITSP/C83 Condition Problem Code SHALL be coded as specified in HITSP/C80 Section 2.2.3.1.1 Problem. See HITSP/C83 Section 2.2.2.7.5, rule C154-[DE-7.04-1]. Context: /ClinicalDocument[1]/component[1]/structuredBody[1]/component[2]/section[1]/entry[3]/act[1]/entryRelationship[1]/observation[1] Test: not(cda:value[@code and @codeSystem]) or cda:value[@codeSystem="2.16.840.1.113883.6.96" and @code]
To translate the NIST validation tool's syntax, it's failing the following check: "Either code and codeSystem are not present, or code is present along with a codeSystem value of 2.16.840.1.113883.6.96"
Thus it's specifically looking for SNOMED or nothing. So not sure how we should proceed.
Given the option of SNOMED and no code system at all, I vote for no code system at all.
On Apr 2, 2013, at 11:15 AM, Skwerl notifications@github.com wrote:
Well, for now, the CCD export properly specifies that the code system is ICD9. That commit is pretty easy to read.
Here is the error that the validator is throwing though:
Message: Error: HITSP/C83 Condition Problem Code SHALL be coded as specified in HITSP/C80 Section 2.2.3.1.1 Problem. See HITSP/C83 Section 2.2.2.7.5, rule C154-[DE-7.04-1]. Context: /ClinicalDocument[1]/component[1]/structuredBody[1]/component[2]/section[1]/entry[3]/act[1]/entryRelationship[1]/observation[1] Test: not(cda:value[@code and @codeSystem]) or cda:value[@codeSystem="2.16.840.1.113883.6.96" and @code]
To translate the NIST validation tool's syntax, it's failing the following check: "Either code and codeSystem are not present, or code is present along with a codeSystem value of 2.16.840.1.113883.6.96"
Thus it's specifically looking for SNOMED or nothing. So not sure how we should proceed.
— Reply to this email directly or view it on GitHub.
Would that screw up CCD import?
I would have to test it to be sure but I don't think we're using the code set field for anything, just the code itself.
On Apr 2, 2013, at 11:59 AM, Skwerl notifications@github.com wrote:
Would that screw up CCD import?
— Reply to this email directly or view it on GitHub.
Well, they come together. If any code is specified, SNOMED codesystem must also be. It's both or neither.
Well yes we do need the problem code there unless you are pulling it from somewhere else. Since neither one of us know lets leave this alone for now and move on.
On Apr 2, 2013, at 12:08 PM, Skwerl notifications@github.com wrote:
Well, they come together. If any code is specified, SNOMED codesystem must also be. It's both or neither.
— Reply to this email directly or view it on GitHub.
Closing this until we know we need to reopen.
The CCD Export identifies the code system for problems to be SNOMED CT. The code system should be ICD-9
Problem excerpt from XML is attached, full XML is below:
Full XML: