A reference guide to rapidly implementing an institutional dashboard for resource allocation and oversight during COVID-19 pandemic surge.

OBJECTIVES: We develop a dashboard that leverages electronic health record (EHR) data to monitor intensive care unit patient status and ventilator utilization in the setting of the COVID-19 pandemic. MATERIALS AND METHODS: Data visualization software is used to display information from critical care data mart that extracts information from the EHR. A multidisciplinary collaborative led the development. RESULTS: The dashboard displays institution-level ventilator utilization details, as well as patient-level details such as ventilator settings, organ-system specific parameters, laboratory values, and infusions. DISCUSSION: Components of the dashboard were selected to facilitate the determination of resources and simultaneous assessment of multiple patients. Abnormal values are color coded. An overall illness assessment score is tracked daily to capture illness severity over time. CONCLUSION: This reference guide shares the architecture and sample reusable code to implement a robust, flexible, and scalable dashboard for monitoring ventilator utilization and illness severity in intensive care unit ventilated patients.

Deep Learning on Electronic Health Records to Improve Disease Coding Accuracy.

Characterization of a patient's clinical phenotype is central to biomedical informatics. ICD codes, assigned to inpatient encounters by coders, is important for population health and cohort discovery when clinical information is limited. While ICD codes are assigned to patients by professionals trained and certified in coding there is substantial variability in coding. We present a methodology that uses deep learning methods to model coder decision making and that predicts ICD codes. Our approach predicts codes based on demographics, lab results, and medications, as well as codes from previous encounters. We are able to predict existing codes with high accuracy for all three of the test cases we investigated: diabetes, acute renal failure, and chronic kidney disease. We employed a panel of clinicians, in a blinded manner, to assess ground truth and compared the predictions of coders, model and clinicians. When disparities between the model prediction and coder assigned codes were reviewed, our model outperformed coder assigned ICD codes.