| Anna A. Yudovin, James Joseph |
This presentation introduces an innovative method of utilizing Synthea™, an open-source system, developed by the MITRE Corporation, for generating clinical trial data. Despite Synthea's original design not targeting clinical trial simulation, its modifiability allows it to be adapted for this purpose.
Synthea's authors have created an additional open-source tool, Synthea Module Builder, to extend its functionality. This allows users to define a variety of physiological phenomena (such as medical conditions) or actions that the simulated individuals themselves might undertake, essentially as a collection of states and state transitions, which may be simple, or governed by conditional logic, statistical distributions – or both.
Another way to augment data produced by Synthea is through creating additional Health Record Editors, which are intended to manipulate information recorded in a synthetic individual's electronic health record – such as providing additional information, or mimicking errors that may take place during manual data entry into an electronic health record system.
We demonstrate how Synthea's flexibility enables the generation of realistic clinical trial data. Future developments target further refinement of simulation techniques, using additional factors like patient drop-off, missing or out-of-window data, and out-of-range values, to pre-define data for real clinical trials, in order to program and validate trial deliverables ahead of actual data collection.
No specific operating system or software version dependencies are indicated, and the intended audience includes individuals with beginner to intermediate software development experience. |