There are so many aspects of OSCAR that as yet do not exist but in listing them it will provide a little bit of structure for road map development. As always really happy to get comments.
- The nature of textual repositories for Clinical Decision Rules
- The unique identification of CDRs – textual – within a Repository
- The unique identification of CDRs – non-textual – within a Repository
- The process of creating CDR Algorithms from CDR Textual Rules
- Role of text or journal publisher in ensuring standardization of textual CDR presentation
- The overarching governance model
- Determining the safety of any endorsed or repository referenced CDRs
- Ensuring the review of CDRs and deprecation of invalid or no longer accurate CDRs
- Developing a level of confidence suitable for machine interpretation to determine use of particular rules in particular problems
- Development of various header and footer metadata structures
- Metadata to incorporate suitable nationalities, domains for use
- The nature of the CDR header to allow machine language understanding of programming structure
- Audit processes to ensure semantic interoperability between text and machine interpretation
I will be developing or ‘fleshing out’ each of the points above and reviewing the literature in these areas.
There are many other things to consider but its a start!