Current Version - Under Revision
The form in which records are represented may have a substantial impact on the efficiency, accuracy and completeness of retrieval. The forms that best suit retrieval may differ from the forms that are required to meet the principles of clinically safe and legally valid electronic health records .
Storing information as entered
This option leaves information in the form entered in the electronic health record with no additions to assist future retrieval. The application must do all the work needed to locate the required records and compute subsumption and equivalence when a request is made to retrieve data.
Using an Expression Repository
An innovative approach to the issues raised by literal storage of postcoordinated expressions is to implement an expression repository. Each unique expression used in the system is stored in a referenced database table and assigned an internal unique Identifier (e.g. a UUID). When an expression is used in a clinical record entry the unique expression id is used to reference the expression in the repository.
The key advantages of this approach of this approach are:
- The expression Identifier can have a fixed size whereas a postcoordinated expression in of variable and indeterminate size. This significantly improves storage and index efficiency.
- The expression repository can also be used to store normal form representations of each expression and to relate these to the original expression. This optimizes performance for expression normalization during retrieval.
- The expression repository could also be processed by a Description Logic Classifier and a transitive closure table of all the expression used in the application could then be generated. postcoordinated retrieval would then be highly optimized by using the transitive closure to test a single join between each predicate and the candidate expressions .
- 7 Testing and Traversing Subtype Relationships
- Optimisation of normalization and expression subsumption testing
- 7.5 Optimizing concept subsumption testing
- Testing subsumption and equivalence between expressions
The objective of this approach is to allow the generation of simple indices for the precoordinated representation. It is then possible to undertake most retrievals using the 116680003 |is a| subtype hierarchy to compute whether Concepts in the record are subtypes of the Concepts used to specify retrieval. Where postcoordination is required, the minimum number of additional tests are required to confirm that a Concept in the record meets the specified retrieval criteria.
One difficulty with this approach is that there may be more than one representation that requires the same degree of postcoordination. This is discussed in more detail and illustrated in Transforming expressions to normal forms .
If this approach is adopted additional rules need to be applied to determine the choice between alternatives with a similar number of postcoordinated components.
"steel pedal bicycle" + " color " = "red"
Both are equally close to the objective of minimizing postcoordination. A rule is needed to determine which of these is preferred. There is no obvious right or wrong solution to this but a simple rule that places the attributes in an order will, if applied consistently, allow all postcoordinated representations to be reduced to a single minimized form.
An alternative approach is to expand any precoordinated concepts in the record to their fullest possible postcoordinated forms. This general type of transformation is illustrated in Transforming expressions to normal forms .
This approach requires a richer record structure but has the advantage that there are three possible end-points to postcoordination, each of which ensures that any computably equivalent representations of Concepts will expand to an identical postcoordinated form. The three end-points are summarized here:
- Short canonical form:
- This option is more verbose as it includes some redundancy.
- A concept is represented as the combination of:
- Exhaustive postcoordinated form:
If the retrieval criteria are expressed in a similar form, a relatively simple query can interrogate the record for all entries with a matching set of primitive Concepts and specified characteristics.