Duration
7 months
Role
Product Designer
Team
Dan Friedman, Cormac Rada, Caleb Pase
Billions were set aside for the US Healthcare system to undergo rapid computerization to set up Electronic Medical Records. Despite widespread adoption, EMRs are more often charged with hurting healthcare than aiding. Since the adoption of EMRs, the majority of physicians agree that EHRs greatly contribute to physician burnout, detract from their clinical effectiveness, and put hospitals in great financial
Finding relevant fields, clicking on options, and Typing data is slower than writing it out. Current options for charting including typing in front of patients, using scribes and dictation software, typing in between patient interactions, and typing at the end of the day.
Some physicians enter the note with the patient present, while others wait until the patient has left the exam room. Those with busy schedules tend to wait until the end of the day to enter their notes for all patients trying to reconstruct what occurred hours earlier. Current methods to easily and seamlessly capture encounters real-time lead to unreadable, rigid, and templated notes, un-useable for future value-based care analytics.
Creating a review process that condenses critical parts of a patient's chart for a quick and intuitive review workflow.Increases in use, clicks per chart, and training times, mean that users are looking for the easiest and quickest way to get to an accurate and approved note.
Creating a representation of a NLP pre-populated chart and note, that clearly communicates its degree of completion.
Creating an experience that transforms how doctors and nurses feel about EMR updates. Pushing the experience from medical software to a productivity app that allows users to measure their output versus induction, manage workload more appropriately, all the while presenting an extrinsic reward system across the practice and providing evidence of the improved outcomes from our tool.
A patient's chart and associated note is a structured record of clinical and historical patient data. In this way an encounter represents a stereotyped situation, in which a doctor divides extracted or abstracted knowledge and "slots" them into "frames" as AI Pioneer Marvin Minsky explains in "A Framework for Representing Knowledge". In the medical setting, the completeness of a chart can be measured across several dimensions, as exemplified in a diagram developed by MITRE's Standard Health Record Initiative. From a complete frame, a comprehensive note can be easily generated following the typical Subjective, Objective, Assessment, Plan model.
Mobile devices allow the user to do menial tasks whenever they have a free moment, while integrated gestural features reduce time between highly repetitive tasks, making the process go lightning fast. A review process -mediated by a mobile device and mobile best practices - would start at the conclusion of a patient exam as a user receives a notification that the chart is ready. The user then enters a review phrase where they validate or revise partially filled, incomplete, or potentially inaccurate entries. The conclusion of this process would push the corrections to the chart and note, and push the user to review the note before they ultimately sign.
The email model gives the user visibility into the entire reviewable landscape, so that they can proactively choose which reviews to edit. The user would be able to toggle contexts between weighted suggestions and the primary source transcript pertaining to that entry. However, in terms of economy of interactions, this layout would require an additional tap or swipe-down to access quick-correct items that could potentially save input time for users.
The quadrant and card layout immediately places the user into reviewing an individual entry. The cards represent a different suggestion to fix the inaccuracy. Swiping or selecting a quadrant would confirm the submission. This design could support multiple directions for swiping. Swiping down on the card container would reveal a drawer of even more quick add items and search, while a swipe up would reveal a drawer related to the primary source transcript. While this design benefits in terms of immediacy and finger reach, it decreases in learnability and constrains the user’s choice in correction.
In these designs a percentage-based visualization presents the user a simple and quick guide to a note's or section of a note's degree of completion.
A power-user wants to optimize the amount of completed, quality notes with time spent trying to complete them. Successful designs should create accessible features to inform the user on the note's relative degree of completion, across several possible dimensions, and assist in the prioritization of the the doctor's backlog.
The population-based visualization gives the user a better understanding of the the location and type of missing information, and perhaps the correlated time associated in filling out said entries.
Creating an experience that transforms how doctors and nurses feel about EMR updates. Pushing the experience from medical software to a productivity app that allows users to measure their output versus induction, manage workload more appropriately, all the while presenting an extrinsic reward system across the practice and providing evidence of the improved outcomes from our tool.