Personalized CQL Artifact Generation for Alert Fatigue Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional clinical decision support systems suffer from alert fatigue due to irrelevant or redundant alerts, lack of context-specific recommendations, and biases, and are difficult to update with new clinical evidence, leading to suboptimal patient care.
Innovation Solution
An AI system generates disease-specific CQL models using LLMs trained on real-world medical data, integrating with FHIR and CDS Hooks to provide personalized, contextually appropriate alerts and recommendations, reducing alert fatigue and ensuring adaptability and bias-free decision support.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional clinical decision support systems generate alerts based on general rules, then coverage of clinical scenarios is broad, but alert fatigue occurs due to irrelevant or redundant alerts
Solution Approach 1:
The system generates CQL artifacts that are personalized to specific patients based on their individual characteristics, medical history, and current condition. Instead of applying uniform alert rules to all patients, the system tailors clinical recommendations to each patient's unique context, ensuring relevance and reducing unnecessary alerts.
Solution Approach 2:
The clinical decision support system segments the patient population into distinct groups based on shared characteristics, conditions, or risk factors. This segmentation allows the system to apply appropriate alert rules and recommendations to specific segments rather than generating alerts for all patients uniformly, thereby reducing alert fatigue while maintaining comprehensive coverage.
2Adaptability or versatility
If traditional CDS systems use static rules, then system complexity is low, but the system is difficult to update with new clinical evidence
Solution Approach 1:
The system transitions from static alert rules to dynamic, adaptable CQL artifacts that can be automatically updated with new clinical evidence. The system continuously learns from new data and updates its recommendations accordingly, allowing it to adapt to changing clinical guidelines and evidence without requiring complete system redesign.
Solution Approach 2:
The system automatically updates its own CQL artifacts by processing new clinical evidence and learning from new data. This self-updating capability reduces the need for manual system reconfiguration and allows the system to maintain current knowledge without increasing operational complexity for users.
3Reliability
If traditional CDS systems provide general recommendations, then ease of implementation is high, but patient care quality is suboptimal due to lack of personalization
Solution Approach 1:
The system generates CQL artifacts that are personalized to specific patients based on their individual characteristics, medical history, and current condition. Instead of applying uniform alert rules to all patients, the system tailors clinical recommendations to each patient's unique context, ensuring relevance and reducing unnecessary alerts.
Solution Approach 2:
The system dynamically adjusts alert thresholds, recommendation criteria, and clinical parameters based on individual patient characteristics such as age, comorbidities, medication regimens, and risk factors. By changing these parameters according to patient-specific data, the system provides personalized recommendations that improve care quality without requiring entirely separate systems for different patient types.
Data Source
AI summary
Described is a system for receiving a disease specific treatment algorithm corresponding to a disease of a patient from a first Electronic Health Record (EHR) system, processing data corresponding to the disease specific treatment algorithm using a Large Language Model (LLM), receiving one or more CQL models from the LLM based on the processing of the data corresponding to the disease specific treatment algorithm, receiving an update to a patient record of the patient from the first EHR system, based on the update to the patient record and the one or more CQL models, triggering a Clinical Decision Support (CDS) hook to generate an alert, and causing transmission of the alert for a medical practitioner associated with the first EHR system.