Clinical Ontology Interface for EHR Data Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current electronic health record (EHR) systems fail to efficiently organize and retrieve clinically relevant information, leading to overwhelming burdens for clinicians, inappropriate resource utilization, and potential compromises in patient safety due to the mismatch between data schema and clinical decision-making needs.
Innovation Solution
A system and method that utilize ontologies to query structured and unstructured data stores, generating a patient abstract by identifying relevant medical concepts and data elements, and retrieving guidelines for interventions, thereby presenting clinicians with concise, clinically relevant information through a user interface.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If clinicians manually search patient electronic health records for relevant information, then they can access medical data, but the time and effort required becomes overwhelming
Solution Approach 1:
The system extracts only the clinically relevant information needed for decision-making from the vast EHR database. It identifies and retrieves specific data elements (medications, allergies, lab results, diagnoses) that are pertinent to the current clinical question, filtering out unnecessary information to reduce search time while maintaining completeness of relevant clinical data
Solution Approach 2:
The system segments the EHR search task into distinct components: extracting clinical questions from the encounter, identifying relevant data elements based on the question type, querying for those specific elements, and presenting results in an organized format. This segmentation allows the system to efficiently handle different types of clinical inquiries without requiring a comprehensive manual search of the entire record
2Loss of information
If EHR systems store comprehensive patient data, then all medical information is available, but the information is dispersed across multiple data stores making retrieval difficult
Solution Approach 1:
The system introduces an intermediary layer between the clinician and the distributed EHR data stores. This intermediary automatically translates clinical questions into appropriate data queries, navigates across multiple data stores (structured and unstructured), and aggregates results. The intermediary handles the complexity of data organization transparently, allowing clinicians to access comprehensive information without directly navigating the complex data architecture
Solution Approach 2:
The system creates a universal interface that handles multiple types of clinical questions (diagnostic, therapeutic, prognostic) through a single integrated process. The same system architecture can retrieve information from various data sources (labs, imaging, notes, medications) regardless of the specific clinical context, making the system adaptable to different clinical scenarios without requiring separate retrieval mechanisms for each data type
3Reliability
If clinicians access all patient medical information, then they have complete data for decision-making, but the information-gathering burden becomes overwhelming
Solution Approach 1:
The system applies local quality by tailoring the information retrieval to the specific clinical question at hand. Instead of retrieving all patient data uniformly, it identifies the specific data elements relevant to the current diagnostic or therapeutic decision. For example, for a medication allergy question, it retrieves only allergy information and related medications, not the entire medical record. This ensures reliability by getting the right information while improving ease of operation by reducing the volume of information the clinician must review
4Ease of manufacture
If EHR systems use structured data schemas, then data is organized for storage, but the schema does not correspond to clinical concepts needed for decision-making
Solution Approach 1:
The system introduces dynamics by allowing the data retrieval process to adapt based on the clinical context. The data elements queried are not fixed by the EHR schema alone but are dynamically determined by the clinical question being asked. The system can flexibly combine structured data from the EHR schema with unstructured data from clinical notes, adjusting the retrieval strategy based on the specific clinical concept being evaluated, thus achieving both storage efficiency and clinical relevance
Data Source
AI summary
Systems and methods for automatically transforming a user interface include a computer system, a first data store containing medical records, a second data store containing one or more data models, and a communications network operatively coupling the computer system, the first data store, and the second data store. The computer system may be configured to automatically extract a term from the medical records, identify a first medical concept related to the term, identify an intervention related to the first medical concept, identify a second medical concept related to the intervention, and transform the user interface by generating an interactive node corresponding to the second medical concept. The user interface may display a first popup in response to an input cursor being placed over the interactive node. The first popup may display detailed information generated based on the second medical concept.