Evidence suggests that CDS can improve health care quality, safety, and effectiveness1. Prior work has demonstrated multiple failures and potential harm from CDS improper executions2.

Visualization of this complex data is necessary in rapidly assessing the impact and effect of the rules execution. The visualization can be layered upon automatic methods for detecting CDS failures and identifying the key characteristics of failure.

Failures in these systems occur for various reasons including logic errors, data corruption, and poor design. Systematic design, prospective failure prediction, and retrospective results evaluation are keys to building a effective and resilient clinical decision support system.