A study of the array of rule-authoring tools used to convert medical knowledge into machine-executable clinical decision support rules across Partners Healthcare in Boston found many limitations--and frustrations.
Most limit the ability to create CDS interventions that are standardized, sharable, interoperable and extensible, the authors found. None was deemed ideal.
The study, published at BMC Medical Informatics and Decision Making, reviewed through meetings with users of the clinical rule-authoring environments at Partners, who manage more than 7,000 CDS rules. It included an ad hoc collection of tools, some implemented enterprise-wide, some for ambulatory care only and others for specific systems. It focused on the process for creating and using reminder and medication rules.
None of the existing tools, for example, facilitate the entire process of transforming free text into structured information that can be used to create executable rules. And most rule editors, embedded in either inpatient/ambulatory EHR systems or terminology editors, do not interface with a centralized knowledge repository. This can lead to overlaps or conflicts in rules applicable to a given clinical scenario, the authors found.
They say rule logic will become more complicated in the future, but the tools lack the capacity to grow with the need. Accommodating genomic data and personalized medicine, for instance, would require modification of an existing editor or development of a new one.
Formal knowledge representation and standards, metadata support, terminology integration, collaboration support and integration with EHR systems are among the elements required for success, they found, concluding that "better management is needed as institutions race headlong into CDS for a mature EHR and Meaningful Use."
Indeed, CDS tools topped healthcare leaders' wish lists for the next 12 months in a survey by Black Book Rankings, but as with all health IT systems, flexibility is key. More personalized CDS is on the way, though, according to a study recently published in the Journal of the American Medical Informatics Association. The new approach--dubbed ADAPT--relies on confidence intervals drawn from individuals to make predictions.
To learn more:
- read the research
Adaptive clinical decision support improves patient risk estimation
Decision support, predictive modeling may speed clinical research
FDA's approach to clinical decision support software: A brief summary
Clinical decision support a double-edged sword
Clinical decision support tools top hospitals IT leaders' wish lists