Robustly Extracting Medical Knowledge from EHRs: A Case Study of Learning a Health Knowledge Graph
This work addresses the need for more reliable medical knowledge extraction from EHRs for healthcare applications, but it is incremental as it builds on prior graph construction methods.
The paper tackled the problem of evaluating the robustness of a health knowledge graph learned from electronic health records, identifying sample size and unmeasured confounders as key error sources and introducing a method using non-linear functions to improve causal graph assumptions.
Increasingly large electronic health records (EHRs) provide an opportunity to algorithmically learn medical knowledge. In one prominent example, a causal health knowledge graph could learn relationships between diseases and symptoms and then serve as a diagnostic tool to be refined with additional clinical input. Prior research has demonstrated the ability to construct such a graph from over 270,000 emergency department patient visits. In this work, we describe methods to evaluate a health knowledge graph for robustness. Moving beyond precision and recall, we analyze for which diseases and for which patients the graph is most accurate. We identify sample size and unmeasured confounders as major sources of error in the health knowledge graph. We introduce a method to leverage non-linear functions in building the causal graph to better understand existing model assumptions. Finally, to assess model generalizability, we extend to a larger set of complete patient visits within a hospital system. We conclude with a discussion on how to robustly extract medical knowledge from EHRs.