Mining Deviations from Patient Care Pathways via Electronic Medical Record System Audits

摘要: In electronic medical record (EMR) systems, administrators often provide EMR users with broad access privileges, which may leave the system vulnerable to misuse and abuse. Given that patient care is based on a coordinated workflow, we hypothesize pathways can be represented as progression of through introduce strategy model patient’s flow sequence accesses defined over graph. Elements in correspond features associated transaction (e.g., reason for access). Based this motivation, patterns usage, indicate deviations from workflows. We evaluate our approach using several months data large academic center. Empirical results show framework finds small portion constitute outliers such flows. also observe violation deviate different types services. Analysis suggests greater deviation normal by nonclinical users. simulate anomalies context real illustrate efficiency proposed method As an illustration capabilities method, it was observed area under receiver operating characteristic (ROC) curve Pediatrics service found 0.9166. The suggest competitive with, better than, existing state-of-the-art its outlier detection performance. At same time, more efficient, orders magnitude, than previous approaches, allowing thousands seconds.