Lead Institution: University of Illinois at Urbana-Champaign
Project Leader: Carl Gunter
This project is based on the hypothesis that care pathways can be represented as the progression of a patient through a system and that a model of the patient’s flow as a sequence of accesses defined over a graph can be used to detect anomalous access events. We defined the sequence to correspond to features associated with the access transaction (e.g., reason for access). Based on this motivation, we modeled patterns of patient record usage and evaluated our approach using the NMH audit data set.
- Focus of the research/Market need for this project
- Project Aims/Goals
- Key Conclusions/Significant Findings/Milestones Reached
Our results show that this framework finds a small portion of accesses that may be considered outliers in a scoring system. We also learned that the violation patterns deviate for different types of medical services. Analysis of our results suggests greater deviation from normal access patterns by non-clinical users. We simulated anomalies in the context of real accesses to illustrate the efficiency of the proposed method for different medical services. For example, the area under the ROC curve for the Pediatrics service was found to be 0.9166. Our simulation results suggest that our approach is competitive with, and often better than, state-of-the-art alterative techniques its outlier detection performance. At the same time, our method is more efficient, by orders of magnitude, than previous approaches, allowing for detection of thousands of accesses in seconds.
- Available Materials for Other Investigators/Interested parties
Code implementing PFADS is available by request from the authors.
- Market entry strategies
Mining Deviations from Patient Care Pathways via Electronic Medical Record System Audits
He Zhang, Sanjay Mehrotra, David Liebovitz, Carl A. Gunter, and Bradley Malin
ACM Transactions on Management Information Systems (TMIS), volume 4, number 4, article 17, December 2013