Cohort Extracting Apparatus Using Bit String History Tables
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cohort extraction methods from Clinical Data Warehouses are inefficient, requiring repeated operations and increased query volumes as criteria change, leading to significant time wastage and unnecessary operations.
Innovation Solution
A method and apparatus that generate a history table for each step of cohort extraction, using bit strings to indicate criterion satisfaction, allowing for efficient tracking and updating of patient events, thereby enabling quick calculation of patient numbers and cohort adequacy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If cohort extraction is repeated as criteria change, then cohort adequacy can be determined, but extraction time increases significantly
Solution Approach 1:
The system performs preliminary extraction of all patient events from the clinical data warehouse into a history table before criteria evaluation begins. This preliminary action stores all necessary data in advance, eliminating the need for repeated extraction operations when criteria change, thus resolving the contradiction between determining cohort adequacy and minimizing extraction time
Solution Approach 2:
The system creates a copy of patient data in the form of a history table that mirrors the clinical data warehouse structure. This copy contains all patient events with bit strings indicating criterion satisfaction, allowing rapid evaluation of different criteria without querying the original warehouse repeatedly, thereby reducing extraction time while maintaining accuracy
2Measurement precision
If the number of criteria increases, then cohort selection precision improves, but query volume increases
Solution Approach 1:
The system pre-evaluates all criteria for all patients during the initial history table generation, storing the results as bit strings in each patient record. When criteria increase, the system simply queries these pre-computed bit strings rather than re-evaluating criteria against raw data, thus maintaining high selection precision while preventing query volume from increasing with additional criteria
Solution Approach 2:
The history table creates a compressed copy of criterion evaluation results for each patient, where bit strings encode satisfaction status for multiple criteria. This copying mechanism allows the system to handle increased numbers of criteria without proportionally increasing query volume, as all criterion evaluations are stored in the copied history table structure
3Reliability
If repeated extraction operations are performed, then updated cohort results are obtained, but unnecessary operations are repeated
Solution Approach 1:
The system maintains a copied history table that contains all necessary patient event data and criterion evaluation results. When criteria change, the system queries this existing copy and updates only the necessary bit strings rather than performing complete re-extraction operations, thus ensuring reliable updated results while eliminating unnecessary repeated operations
Solution Approach 2:
The system discards the inefficient approach of complete re-extraction and recovers only the necessary updates by modifying bit strings in the existing history table. This selective update approach maintains result accuracy while dramatically improving extraction efficiency by avoiding repetition of operations that have already been performed
Data Source
AI summary
Disclosed is a method of operating a cohort extracting apparatus, the method including: obtaining cohort entry criteria and extracting events corresponding to the cohort entry criteria from a clinical data warehouse; generating an initial history table including an event identifier, a patient identifier, and a bit string indicating satisfaction of criteria of an initial step for each extracted event; obtaining criteria of a current step, identifying, among patients included in a history table of a just previous step, current step patients having an event corresponding to the criteria of the current step, and updating a bit string for each event of the current step patients included in the history table of the just previous step, and generating a history table of the current step by adding new events extracted in the current step; and sequentially generating a history table for each step, and then generating a cohort table by using a history table of a final step.