Conditional outlier detection for clinical alerting

TL;DR

This paper presents a data-driven method for detecting unusual patient-management actions in electronic health records to identify potential errors, validated on post-cardiac surgical patient data with expert opinions.

Contribution

The paper introduces a novel approach for anomaly detection in clinical management actions, demonstrating its effectiveness in reducing false alerts and correlating anomaly strength with alert rates.

Findings

Anomaly detection can identify potential errors with low false alert rates.

Stronger anomalies are associated with higher alert rates.

Expert opinions support the effectiveness of the proposed method.

Abstract

We develop and evaluate a data-driven approach for detecting unusual (anomalous) patient-management actions using past patient cases stored in an electronic health record (EHR) system. Our hypothesis is that patient-management actions that are unusual with respect to past patients may be due to a potential error and that it is worthwhile to raise an alert if such a condition is encountered. We evaluate this hypothesis using data obtained from the electronic health records of 4,486 post-cardiac surgical patients. We base the evaluation on the opinions of a panel of experts. The results support that anomaly-based alerting can have reasonably low false alert rates and that stronger anomalies are correlated with higher alert rates.