Leveraging structured and unstructured electronic health record data to detect reasons for suboptimal statin therapy use in patients with atherosclerotic cardiovascular disease

Figures

• 

  Fig. 1. Flow diagram used to identify dates for selecting notes more likely to include a rationale for statin therapy guideline non-adherence (AST – aspartate transaminase; ALT – alanine transaminase; CK creatine kinase.

• 

  Fig. 2. Flow diagram demonstrating the process used to develop the NLP system (left) responsible for identifying reasons for VA atherosclerotic cardiovascular disease (ASCVD) patients not being on high-intensity statin therapy (HIST). Also shown (right) is the process used to test the accuracy of the system with respect to classifying patients according to whether such a reason existed in a) structured data stored in the VA adverse drug event (ADERS) system, b) unstructured form within the text of clinical notes, or c) neither, indicating potential clinical inertia.