Improving Clinical Decision Support Reliability using Anomaly Detection Methods

R01 LM011966 (PI – Adam Wright, PhD  2014-2018) Clinical decision support systems, such as drug-interaction alerts and preventive care reminders, when used effectively, have been shown to the quality, safety and efficiency of care. However, such systems are complex and sometimes fail – these failures are often not noticed for a long period of time and can lead to patient harm. In the proposed project, we will study the causes of such failures and develop and test anomaly detection systems to detect such failures and alert knowledge engineers about them with the goal of improving the safety and reliability of clinical decision support systems.

For more information see project profile on NIH REPORTER.