Independent UX project
As a part of an UX interview, I was challenged to design a product experience that helps ED doctors and nurses treat a large volume of patients effectively.
The solution was AssistED, an application plugin that allowed emergency providers to accurately and efficiently make medical decisions by providing summarized patient information and recommended clinical actions.
I conducted research, ideated and designed within a 5-day period, with the goal to demonstrate effective solution in the highest fidelity possible.
UNDERSTANDING PROBLEM SPACE
To start, I drafted 4 guiding research questions about the emergency department experience.
What is the general workflow in an emergency room?
Who are the key stakeholders and what are each of their needs?
What are some steps in the process where efficiency and accuracy can be improved?
What are some ways emergency physicians and nurses interact with technology currently?
To understand the typical workflow in an emergency department, I conducted in-depth interview with an ED medical professional.
Based on my conversation with the medical professional, I created the below flow diagram to illustrate key players, interactions and digital touch points. Non-critical patients usually spend around 2 hours in the ED. Some of the key takeaways I learned about ED workflow are:
Most regular ER visits follow a consistent and standardized process and pattern as illustrated in the diagram below.
Nurses are highly involved in the ER experience, even performing early clinical actions.
Physicians key actions for diagnosis actually take place outside of the exam room at his work desk on a computer.
Using information gathered through research, I uncovered a few key insights about emergency providers’ experience to fuel design.
Insight 1: Emergency physicians have very little time to review a patient’s medical history
A patient’s medical history is critical to making the right diagnosis. Primary care physicians, whose patients are appointment-based, have time to review a patient’s medical record a day in advance. Whereas emergency physicians don’t know their patients ahead of time and often need to care for multiple patients concurrently, which gives them very little time to review patients’ medical histories.
Persona - Dr. Ian Sellers, MD
Based on a study in 2014
Insight 2: Treating patients with a wide variety of problems increases difficulties.
Emergency physicians are medical generalists who are trained to take care of patients of all kinds. However, under the extremely intense and fast-paced environment in the ED, it is very easy for them to make mistakes when making medical decisions, which can lead to harmful consequences.
Therefore, it’s important that these physicians can communicate with specialists and have other resources at hand whenever they need them.
A triage nurse is usually the first person that a patient interacts with in an ED. Triage nurses are usually very experienced and will determine the acuity level of a patient’s illness on a level of 1-4. This is crucial information that helps physicians determine which patient to care for first. A mistake made by triage nurse can harm the outcome of treatment and disrupt physician’s whole workflow. Therefore, triage nurses will also need to review patient’s history and make fast decisions based on information consumed.
Insight 3: Nurses play an important role in determining acuity level and taking early clinical actions.
Insight 4: EHR systems are widely adopted and deeply integrated into the ED workflow.
99%* of all hospitals in the US have adopted electronic health record systems (EHRs). These systems are all-encompassing and allow emergency providers to register patients, review medical history, send lab test orders, write discharge instructions and save medical record, etc.
Because of EHRs’ large number of functions, however, many professionals complain about these systems’ steep learning curve, tedious user flow and the number of clicks to get to things.
Epic (left) and Cerner (right), the most popular EHR systems today.
REFRAMING THE PROBLEM
Based on research insights, I further broken down the design challenge into three "how might we" statements.
HMW allow emergency providers to consume patient’s medical history faster?
HMW help physicians make accurate decisions when diagnosing patients?
HMW improve the current EHR systems to be more efficient and usable?
I brainstormed potential solutions for these design problems by conducting the “reverse assumptions” activity.
Here is what the activity looked like:
First, I wrote down my assumptions/knowledge about the current state of the emergency department.
I challenged these norms by writing down the opposite of each assumption.
Then, I brainstormed ideas to help accomplish each of these reverse states.
After I generated a list of opportunities/ideas, I prioritized the ones that are most relevant to the design prompt and combined those to create my final solution concept.
Emergency physicians don’t have time to review patient’s full medical history
Emergency physicians CAN review patient’s full medical history
Extracting the most important medical history information and presenting in digestible way
The design will enable emergency providers to accurately and efficiently make medical decisions by providing summarized patient information and recommended clinical actions.
AssistED is an improvement to the current EHR systems. Its form is not limited — it can be a stand-alone EHR product, or ideally, a versatile plug-in application to any EHR systems out there. Some of the key features are:
I interview the ER medical professional again using these mid-fidelity wireframes as a tool to make sure I put accurate and logical information into the design.
The information provided by AssistED largely depend on the patient's medical record. Presumably, with AI and machine learning technology, AssistED will provide ER professionals with the most relevant information regarding current visit, such as order status and significant findings in test results, and medical history, which may include excerpts from past doctors notes, medication history, etc. It will also recommend actions to ER professionals every step of the way based on patient data, similar cases, treatment database and other information.
My goal is not to redesign existing features in an EHR system, but to imagine a new, effective solution to improve the current experience for emergency providers.
For this prototype, since I want to demonstrate how AssistED could be an add-on to existing EHR systems and work with such systems to streamline workflow, I mocked up my design after a popular EHR software and recreated colors and layouts.
Therefore, one constraint that I want to speak out about is that my high-fidelity prototype is an oversimplification of an EHR system. The final deliverable serves more as a proof of concept for AssistED, focusing on the right sidebar of the screen once you click on the floating button.
I relied on Youtube training videos for Cerner EHR platform to build a high-fidelity prototype with accurate information.
Through my research, I got an insight that:
ED professionals prioritize patients by acuity levels and refer to patients by room numbers.
Therefore, in my prototype, I designed a patient list in the left-side menu to help emergency providers better navigate their list of patients. The list is sorted in chronological order by time of admission, and color coded to reflect acuity levels (Level 1 being most acute, and Level 4 being least acute).
A user will see exam room number and acuity level on each button and can hover over it to view more information about the patient.
NEXT STEP + REFLECTION
User testing with more emergency physicians and providers
My solution is based on interview information gathered around one specific ED’s structures and experience. Although many things are universal between hospitals, I would like to recognize that there are differences as well. By getting feedback from more professionals in the field, I will be able to keep iterating on my design.
I appreciated the opportunity to have done this design challenge. I passed the interviews with this company but did not sign an offer due to personal reasons. After the project, I reflected on on my project planning, design process, and the way I handled ambiguities to see if there are areas of improvement. With a time-intensive project like this, I certainly faced a lot of challenges in the process, one being that I went too deep into researching EHR systems and had to remind myself the true focus of the project. This project is a valuable learning experience for me and will help me grow as a designer.