2016-12-28PC DrugAllergySubstance Call Minutes
Jump to navigation Jump to search
|HL7 Negation Requirements Meeting Minutes
Time: 9:00-10:00 ET
|Facilitator||Jay Lyle||Note taker(s)||Jay Lyle|
|y||Jay Lyle||JP Systems / VA|
|Rob Hausam||Versaggi Consulting|
- review table design
- review plan for completion & QA
- Froukje provided further information about G-SRS. It currently uses UNII codes as keys, but will generate a new set of international identifiers, available in about 2 years. IDMP is the model. Participants are national drug regulation agencies (including EMA); ISO seems to be involved but leadership & stakeholder organization unclear at this time.
- Include reactions? We'll look at the possibility of classifying substances by reaction type, but even if they classify cleanly, we only have that information in one data set.
- Should we look into WHO International Non-proprietary Names (INN)?
- Question of cross-reactivity. We envision a two-step process. In the first step, we capture what is known about the patient, e.g., allergic to substance X. The second step happens at a future point in time, where a drug is considered for the patient: the order system needs a drug check service that uses a knowledge base to check for risk, whether from containment (proposed drug contains a substance on the patient's list) or cross-reactivity (proposed drug may cross react with a substance on patient's list). That second step is not in scope for our project. In the US, it is handled by the private sector, and we have not indication that it's not being done as well as possible.
- Table design:
- one table listing each string, source, and count
- one table listing each map (string, mapped system, mapped code, mapped label)
- A string may occur multiple times in the frequency table, if it is reported by multiple sources.
- A string may occur multiple times in the map table, if it is mapped to multiple systems.
- A query will consolidate the sum of frequencies for a string with the set of mappings.
- If a string is mapped to different concepts in the same system, we will need to resolve the conflict.
- Completion plan
- Jay to complete table deconstruction
- Jay to map using UMLS & RxMix
- Michael Tan may have a tool
- Rob and Elaine to look into mapping using automated tools
- Quality plan
- Each string will be mapped at least twice, preferably 3-4 times.
- All cases where maps disagree will be resolved by joint review (on the call)
- All concepts in the top frequency list (whatever that number is) will be reviewed by multiple experts
|Next Meeting/Preliminary Agenda Items|
© 2012 Health Level Seven® International. All rights reserved.