Trauma Registry Project FAQ

Back to Trauma Registry page

Modeling

#	Question	Answer	Status
1	How do we distinguish a patient who has no prehospital section because there was no EMS engagement from a patient who who has no prehospital section because the data is missing?	The section could be populated with a negation indicator, however, the semantics are unclear on negating a section. Instead, the prehospital transport question, in the hospital section, can be used. A "not applicable" value means that there was no prehospital transport--i.e., no professional service rendered--while a null would indicate that the information is unknown. (OK 2/24 - but annotate to make it clear). If not present, unk or na allowed. ACS: Do you need to know why it's missing, if it's missing? If so, how do you do it today?	Open.
2	How are identifiers assigned?	* The provider institution assigns patient identifiers, and ACS uses the facility identifier with the patient identifier to ensure uniqueness. Both of these can be stored as II extensions. The II roots must be OIDs (or UUIDs), though they maybe null. For now, vendor software assigns provider institution OIDs that can be used as II roots (for documents, patients, or any other resource). The Facility Id should be ACS OID root plus the ACS-assigned 5-digit ID. In the event a state or regional group submits on behalf of a facility, that organization would be the custodian, and would need an ACS ID. The document id should be the (software-assigned) institution OID plus locally generated document id. ACS doesn't know what that id is; it's up to the implementer. SetId is for the patient record; document.id shows the version unique identifier; document.effectiveTime puts those versions in order. No versionNumber necessary. xsd datetime property is precise to the minute Facilities may update submissions; use the CDA version number property. The patient ID is functionally the same as the document id.	Open (versioning)
3	Are time zones required on time stamps?	Time zones are recommended. This will be a text recommendation in the guide.	Open (not yet applied)
4	Are time intervals inclusive or exclusive?	The abstract data types specification stipulates no default assumption, but the IVL pattern does support the Boolean properties lowClosed and highClosed that allow the user to specify the boundary type.	Closed
5	confirm validation can include constraint identifier in error message	Grant reports no, but NTDS identifiers will be provided as they are today	closed
6	Do we need a code system for actEncounterCode (FLD)	AMS?	Open
7	Should we specify section title in IG (content, not just presence)	Yes; "should" for title. Text is up to implementer.	Open
8	Put last modified date in version rather than create?	document.effectiveTime is document generation; should not change with new versions. ServiceEvent.effectiveTime may change if record is updated. The time the event occurred is an observation. Version is also needed; synchrony to be discussed on a future call on workflow.	Open
9	May need to break Transport Mode into two templates	AMS	Open
10	Value sets without enumerated members are not listed in appendix: Trifolia issue		Open
11	Missing LOINCs	Most received; some questions on missing items, and address codes requested. Need to choose a deadline for AAAM. Discuss in September.	Open
12	Need question codes for coded address elements	Question requested from LOINC. FIPS has been retired, asserting that "INCITS" is the new steward of these values, but charges for them. However,ANSI now publishes "ANSI FIPS" for the Census: we'll use these. (http://www.census.gov/geo/reference/ansi.html)	Open

Terminology

#	Question	Answer	Status
1	(age units) Minutes in HL7 not in NTDS; Weeks in HL7 not in NTDS	I don't recall deciding to add these, so we should probably remove those two. NEMSIS has minutes but not weeks.	Open
2	Industry incorrect	I don't know why LOINC assigned some codes and not others. We should use the NAICS top-level values, not the LOINC answer codes.	Open
3	Any long list incomplete (e.g. industry)	can we set up Trifolia for this, or do we need to publish a spreadsheet (or use PHIN VADS or VSAC)	Open
4	Code for “Other Race” is in NTDS but not in HL7	We propose to use "OTH" flavor of null	Open
5	LA17105-0 - Patient Chemically paralyzed – In HL7 is not a NTDS code	Included based on harmonization with NEMSIS	Closed
6	LA17388-2 No airbag present – Not an NTDS code and currently the represented NA – null flavor case. This is the only case where I found a NTDS Null coded into a LOINC code.	Included based on harmonization with NEMSIS. And "not present" isn't really a null.	Closed
7	Protective Devices: Code Distinction issue – The lap and shoulder belt codes now contain the ‘Only’ description which is not accurate to the existing NTDS codes.	Included based on harmonization with NEMSIS	Closed
8	Child seat breakdown now in protective devices, no distinct field/element?	Included based on harmonization with NEMSIS	Closed
9	No code for “Airbag Present”	Based on harmonization with NEMSIS--Airbag status contains this information	Closed
10	No code for ‘Other’ protective devices	We propose to use "OTH" flavor of null	Closed
11	Need country codes	Use VS 2.16.840.1.113883.3.88.12.80.63 (HITSP), system 1.0.3166.1.2.2, name Country Codes	Open
12	Need State codes	Use VS 2.16.840.1.113883.3.88.12.80.1 (HITSP), system "NA", name FIPS 5-2 (State) (available at http://www.census.gov/geo/reference/ansi_statetables.html)	Closed
13	Need County codes	Use VS "NA", system "NA", name FIPS 6-4 (County) (available at http://www.census.gov/geo/reference/codes/cou.html)	Closed
14	Need City codes	Use VS "NA", system "NA", name ANSI Codes for Places (available at http://www.census.gov/geo/reference/codes/place.html). Or, NEMSIS uses GNIS, available at http://geonames.usgs.gov/domestic/download_data.htm	Open