This picture speaks for itself really (right click and -> "view image" removes the clutter) and aims to illustrate the connection between a bunch of activities that need to go on to create and effectively deploy knowledge bases and clinical decision support. There is nothing awfully complicated in here... the trick is to take it small piece at a time.
The sequence of events is:
1. Clinician activates decision support at point of care, or other event triggers request
2. System interfaces assemble a complete set of data about the individual (in a "vMR")
3. Decision support engine applies rules and decision support algorithms to this vMR
4. Conclusion sent immediately back to the point of care and to other audiences as needed
5. Clinician optionally clicks on link to record feedback on value and accuracy of the decision support conclusions
Architectural principles:
1. Generic system components that do one thing and do it well, with reuse of components across domains to reduce software costs
2. Clean simple defined interfaces between system components to allow ongoing system improvement
3. Leveraging existing HL7 standards
4. Provision for direct clinician feedback into the knowledge engineering process
5. Specialist knowledge formats to support complex rules in the simplest format possible
6. Inspections and test case library to allow knowledge validation and quality assurance
7. Standardised knowledge representation formats to allow better tooling and greater depth of skill across campus
8. Explicit separation of clinical knowledge from software to allow knowledge reuse in different contexts and using different front end applications
No comments:
Post a Comment