Predictive Analytics Work Lists for Healthcare Risk Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Healthcare providers are overwhelmed by the volume of work items generated from predictive models used to predict disease recurrence or hospital readmission, as these models aggregate results in a single place, leading to inefficient distribution of tasks.
Innovation Solution
A mechanism that uses predictive models to generate risk scores and categorize work items based on contributing factors, distributing them to appropriate healthcare providers according to predicted risk scores, predictive models used, and factors contributing to those scores.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If predictive models aggregate all work items in a single place, then comprehensive patient risk assessment is achieved, but healthcare providers are overwhelmed by information overload and inefficient task distribution
Solution Approach 1:
The patent segments the aggregated work items into separate categorized work lists based on contributing factors (e.g., clinical domain, urgency, patient condition). Each work list is assigned to specific healthcare providers or departments, transforming the single overwhelming list into multiple manageable, targeted lists that maintain comprehensive assessment while improving provider efficiency
Solution Approach 2:
The system introduces an intermediary layer (the work item distribution mechanism) between the predictive model output and healthcare providers. This intermediary automatically routes work items to appropriate providers based on predefined rules and contributing factors, eliminating the need for providers to manually sort through all aggregated items while ensuring comprehensive coverage
2Measurement precision
If comprehensive analytics are performed on all patient records data, then accurate predictive risk scores are generated, but extensive computation time and resources are required
Solution Approach 1:
The system performs preliminary actions by pre-calculating and storing contributing factors and risk scores during patient encounters and data collection. When work items need to be generated, the system retrieves pre-computed results rather than performing full analytics from scratch, maintaining accuracy while significantly reducing computation time at the point of care
Solution Approach 2:
The system applies local quality by focusing computational resources on calculating only the specific contributing factors relevant to each patient's risk assessment rather than uniformly processing all possible data elements. The analytics engine selectively computes risk scores based on patient-specific conditions and data availability, improving efficiency while maintaining precision where needed
Data Source
AI summary
A mechanism is provided in a data processing system for generating healthcare work item recommendations based on predictive analytics. An analytics engine executing on the data processing system performs analytics to discover patterns in patient records data and to generate one or more risk scores using one or more predictive models. Each of the one or more risk scores represents a probability of a respective healthcare consideration. Each of the one or more risk scores has an associated set of contributing factors. A decision system executing on the data processing system generates a healthcare recommendation for a given patient having a given risk score based on the given risk score, a predictive model used to generate the given risk score, and a given set of contributing factors associated with the given risk score.