Patient Insight is working to drive more informed medical decisions by designing a real-time interface that dynamically captures and presents prior history for complex patients in less time and with more completeness than using their electronic health record (EHR) system. The company is developing proprietary data mining and natural language processing algorithms that show promise for identifying, aggregating, and displaying the most essential EHR data for patients with complex conditions while at the same time working with clinicians to design the visual interface to reduce the cognitive burden and support diagnosis and treatment decisions. The specific goals of this phase I project are to test the feasibility of developing Patient Insight?s core solution to (1) use advanced analytics and natural language processing to mine and prioritize the most relevant information for use by (ED) physicians and (2) develop an clinician-informed, visual display of that essential information to timely support clinical decision-making sufficient to reduce medical errors, improve health, and control costs. In phase I, Patient Insight will work with Universal Health Systems, the parent company of George Washington University Hospital, to co-develop and test the platform for efficacy in the ED as a precursor to continued development of this core solution/product. This will involve using EHR data to assess comprehensiveness to reconstruct history for five groups of chronically ill patients (e.g., advanced cancer, coronary heart disease, insulin-dependent diabetes, chronic pain, and schizophrenia) who are admitted with high-frequency chief complaints of abdominal pain, chest pain, and fever. This will include mining, prioritizing, and presenting information to clinicians using a total of 30 dimensions of data regarding patient?s prior ED and hospital visits including both structured (e.g., diagnosis, medications, labs, imaging test types, EKG date/time) and unstructured data (e.g., images, physician notes, consult notes) to establish physician consensus on the data most valuable to medical decision-making and co-design data visualization tools to render a clear patient history.
Our aims i nclude:
(Aim 1) Establishing the feasibility of leveraging provider-as-user-centered design to guide development of a data visualization platform that enables providers to readily ?see? the most complete patient record for the five groups of complex patients, and (Aim 2) Conducting usability testing and compare the completeness and efficiency of using the enhanced platform on the five patient categories with usual care. Success in phase I will enable the company to validate a ?proof of concept? for developing its clinician-as-user-centered design process to inform a phase II research and development effort to design a commercially viable clinical decision support tool for hospitals and health systems.
Patient Insight seeks to determine the feasibility of its proposed platform to aggregate and present the clearest picture possible of a patient?s history for time-constrained emergency department (ED) providers sufficient to drive improved medical decision-making, safety, and improved health outcomes for complex patients at the point-of-care and across disparate EHR systems. The strategy is to leverage our capability to extract discreet and unstructured data from select EHR systems and engage providers in the design of key elements of information display to reduce duplicate testing, lower medical errors related to misinformed treatment choices, and reduce unnecessary treatment and costs. Proving feasibility will enable Patient Insight to structure a partnership with Universal Health Services and inform further development and testing of its core solution across multiple ED facilities.
