From a6bab6095f3ac3c952ffe48bedb9304bf9eead02 Mon Sep 17 00:00:00 2001 From: jamesagnew Date: Mon, 12 Mar 2018 06:25:30 -0400 Subject: [PATCH] Loinc updates --- LOINC_NOTES.txt | 53 ++++++++++++++----------------------------------- 1 file changed, 15 insertions(+), 38 deletions(-) diff --git a/LOINC_NOTES.txt b/LOINC_NOTES.txt index e5930368ab7..83ffb09561f 100644 --- a/LOINC_NOTES.txt +++ b/LOINC_NOTES.txt @@ -8,54 +8,31 @@ TODO: Comments for Loinc: Overall -- ValueSet and ConceptMap resources have a spot for copyright and - contact information. Are there official values for these? +- ValueSet and ConceptMap resources have a spot for copyright and contact information. Are there official values for these? Answer Lists -- Per the notes, there is no way in FHIR currently to map answer lists to - codes based on context. For this reason, I am ignoring any entries in - LoincAnswerListLink_Beta_1.csv where the "ApplicableContext" context is - not empty. Is this correct? +- Per the notes, there is no way in FHIR currently to map answer lists to codes based on context. For this reason, I am ignoring any entries in LoincAnswerListLink_Beta_1.csv where the "ApplicableContext" context is not empty. Is this correct? Parts -- Only parts with a status of "ACTIVE" are being imported, any others are - ignored. -- The PartTypeName (e.g. "ADJUSTMENT") is ignored as there is no corresponding - property in loinc.xml -- PartDisplayName is not mapped -- Part links are not currently processed (it's not clear to me how to model - these in FHIR, as CodeSystem.hierarchyMeaning has to be only one of 'is-a' - or 'part-of' and presumably the 'is-a' relationship is more important. +- Only parts with a status of "ACTIVE" are being imported, any others are ignored. Does this make sense? +- The PartTypeName (e.g. "ADJUSTMENT") is ignored as there is no corresponding property in loinc.xml +- PartDisplayName does not have an obvious mapping to FHIR +- Part links are not currently processed (it's not clear to me how to model these in FHIR, as CodeSystem.hierarchyMeaning has to be only one of 'is-a' or 'part-of' and presumably the 'is-a' relationship is more important. Part Mappings -- I have made LOINC the source and SCT the target for the mappings in the - ConceptMap resource. Does this seem like the appropriate orientation? -- A canonical URI should be defined for the LOINC->SCT mapping ConceptMap - resource. I have hardcoded "http://loinc.org/loinc-to-snomed" for now, but - we should discuss what is appropriate. +- I have made LOINC the source and SCT the target for the mappings in the ConceptMap resource. Does this seem like the appropriate orientation? +- A canonical URI should be defined for the LOINC->SCT mapping ConceptMap resource. I have hardcoded "http://loinc.org/loinc-to-snomed" for now, but we should discuss what is appropriate. RSNA Playbook -- A canonical URI should be defined for the "all RSNA playbook codes" ValueSet. - I have hardcoded "http://loinc.org/rsna-codes" for now but we should discuss - what is appropriate. +- A canonical URI should be defined for the "all RSNA playbook codes" ValueSet. I have hardcoded "http://loinc.org/rsna-codes" for now but we should discuss what is appropriate. - A name for the "RSNA Playbook" ValueSet is needed. -- Just to confirm, the "all RSNA playbook codes" ValueSet should contain the - loinc codes (e.g. "17787-3") and not the part codes (e.g. "LP199995-4")? -- A codesystem URI for radlex RID and RPID codes is needed (currently - "http://rid" and "http://rpid" are used as placeholders since I'm assuming - these exist somewhere. -- For mappings from loinc part codes to RadLex RIDs, are the codes considered - equivalent (or would they be wider/narrower). They look equivalent to me. +- Just to confirm, the "all RSNA playbook codes" ValueSet should contain the loinc codes (e.g. "17787-3") and not the part codes (e.g. "LP199995-4")? +- A codesystem URI for radlex RID and RPID codes is needed (currently "http://rid" and "http://rpid" are used as placeholders since I'm assuming these exist somewhere. +- For mappings from loinc part codes to RadLex RIDs, are the codes considered equivalent (or would they be wider/narrower). They look equivalent to me. Document Ontology -- Per the SOW, "A value set containing terms in the LOINC Document Ontology - will be created". Just to confirm, entries in this ValueSet are therefore - LOINC terms (such as "11488-4 / Consultation Note") as opposed to part - codes? -- Need to define a URI for the document ontology ValueSet. Currently I am - using "http://loinc.org/document-ontology-codes" +- Per the SOW, "A value set containing terms in the LOINC Document Ontology will be created". Just to confirm, entries in this ValueSet are therefore LOINC terms (such as "11488-4 / Consultation Note") as opposed to part codes? +- Need to define a URI for the document ontology ValueSet. Currently I am using "http://loinc.org/document-ontology-codes" Top 2000 -- Need to define a URI for both ValueSets. Currently I am using - "http://loinc.org/top-2000-lab-results-us" and - "http://loinc.org/top-2000-lab-results-si" +- Need to define a URI for both ValueSets. Currently I am using "http://loinc.org/top-2000-lab-results-us" and "http://loinc.org/top-2000-lab-results-si"