June 4, 2014
11 EDT/10 CDT/9 MDT/8 PDT
Attending: Susan Albright, Hugh Stoddard, Co-Chairs; Terri Cameron, Sascha Cohen, Kristi Ferguson, Scott Kroyer, Noel Rios, Robby Reynolds, Paul Schilling, Valerie Smothers and Walter Fitzwilliam.
1 Review minutes of last meeting
The minutes were accepted as submitted.
2 Discuss null values of id attributes
Valerie explained that every event, sequence block and integration block has an ID which is given using an ID attribute. Although use of the ID attribute is required, the field can be empty or null and the XML document still validates. If the ID attributes are empty, the XML is totally unusable. Terri brought this to Valerie’s attention and suggested changing the specification so that it would no longer be valid to have a null attribute and ID would be required. Valerie recommended changing it to use a non-null string data type. Sascha asked the reason for the current datatype. Valerie replied it was an oversight. Susan asked for an example of what happened. Terri commented there was a document that passed XML validation, was loaded in the Curriculum Inventory, and then they discovered the data was unusable. Hugh asked how the uniqueness requirement would be enforced. Valerie will research that, and suggested using ID type as a way to provide uniqueness. Terri noted that the AAMC system checks for uniqueness. Duplicate ids are not allowed. Valerie agreed to research appropriate data types for ID attribute.
3 Review revised specification (and schema for the technically inclined)
Valerie continued with the changes related to the discussion on longitudinal clerkships and flexible scheduling. Both may be courses that start and end in different academic levels. She and Sascha developed a draft data model. Currently sequence blocks point to a single academic level. They propose having instead sequence block levels that indicate the starting and ending academic levels for the sequence block. Page 43 shows what that would look like. The scheduling attribute would be optional and would provide a way to indicate if the sequence block is longitudinal (ie. It spans multiple academic levels) or is flexible (ie. It may occur in one or two more academic levels). Valid values are longitudinal and flexible. If the sequence block is neither longitudinal nor flexible, you would omit the attribute.
Terri asked if this discussion was for future enhancements for the standard because she was not sure they could get this coded at this time. Valerie clarified all changes showing in the document were intended to be for the next version 1.1 Susan suggested including the null value in this version. Terri noted the AAMC wanted that change implemented as soon as possible.
Scott asked if there was an allowance for standard bug update to be expedited. Valerie will check with ANSI to see if they have any recommendation for bug fixes. Paul asked if ANSI has a Syntax standard for versioning and version release. Valerie will research what ANSI recommends and share with the working group.
Susan asked if academic level must be discreet or can they be overlapping? Valerie answered that academic levels cannot overlap. Terri suggested using consistent terminology in the revised spec; sequence block levels is confusing. Sascha commented that competency based overlapping levels is difficult to conceptualize; levels are discreet units that happen in a certain order. Hugh added that you still have PGY1, 2, 3 and there is money attached to each of these. The question becomes do we care where students are exposed to content.
Terri stated they have considered not having Academic levels anymore because they don’t have a lot of meaning. Schools are using some sort of structure. Is knowing the structure and progression of all the schools important and useful data to us? Susan recommended getting feedback from vendors and implementers. Valerie agreed to work with Susan on phrasing that question. Scott recommended getting feedback from pharmacy and nursing as well; he will work to get feedback from other health professions.
Valerie noted the only other change was a technical change to using the non-null string datatype defined in the common.xsd file. Scott asked if it was possible to do a schema that reflects the AAMC business rules layering on top of the MedBiquitous schema. Terri can’t answer that question but she can ask someone. Valerie commented that if the AAMC did create a different schema for validation, they would have to post the profile schemas and tell everyone to point to those schemas. It would be a significant change in the AAMC system.
4 Discuss proposal for flexibly timed clerkships
Susan asked if sequence block levels relate to number 4 as well. Valerie answered yes; it can appear in different years and different academic levels. Susan asked Hugh if he had a 2nd year clerkship in 4th year would this solve your problem. He agreed it would.
5 Discuss use cases and issues modeling clerkships
Susan asked Sascha to share his thoughts on this document. Sascha mentioned this evolved from discussions internally and with the working group about the best way to capture information for clerkships. Sascha thought it would be helpful to hear from Terri after this past year and what she has seen in terms of data coming through. Terri noted it’s likely going to be awhile before they can run content reports. She suggested tabling the concept for a while. Susan asked if there were other issues on the roadmap that we can tackle. Valerie replied that best practices around documenting flipped classrooms and comprehensive reports for academic levels were still pending. Terri will share statistics for the next meeting in July.
- Valerie and Susan will develop and circulate a question related to the utility of academic levels .
- Scott will query other health professions schools regarding the perceived value of modeling academic levels.
- Terri will share some Curriculum Inventory statistics for the next meeting in July.