Cedars-Sinai Chatbox
Creating more efficiency in the Cedars-Sinai emergency room
THE PROBLEM
There are often long wait times to see a doctor in the emergency room, which could be deadly in certain situations.
History-taking is crucial to diagnosing patients correctly. It is said that 80% of conditions can be diagnosed by history alone. No matter what is going on with the patient, the first step toward making a diagnosis is always obtaining a history. Otherwise, it is not possible to decide which tests and medications to use.
The issue is that process of history-taking takes a lot of time because doctors often have to ask the patient a question, wait for the patient to answer them, and then record the information down.
HIGH LEVEL TIMELINE
10 weeks
MAKE OF THE TEAM
Kaitlyn Tamares, Courtney Pon, Anita Heng An Liao, Fiora Chan
KEY GOAL
Create a website that would help Cedars-Sinai's emergency room process
MY ROLE
I was the lead designer and creator of the Cedars-Sinai chatbox.
My team consisted of four members. The other three members were knowledgeable in biology and public health. They would research why history-taking is important, what aspects of it can be changed to be more efficient, what kind of questions should they ask, and much more health-related questions.
I was the only one with knowledge in computer science and informatics. Therefore, I was the sole creator of the chatbox. I would use the information that my fellow members found and use it to design a chatbox for Cedars-Sinai emergency room. I used low-fidelity prototyping and Adobe XD for designing. I created a website and free source chatbox to create our product.
UNDERSTANDING WHAT PHYSICIANS NEED
History-taking includes a patient's personal information and can also help in identifying what kind of sickness a patient may have.
History-taking is important in identifying a patient's needs. History-taking is key in making a diagnosis for the patient. There are often a long wait, even at the emergency room. Since there are so many patients that physicians and doctors have to see, they may rush to see each patient, so that they can help everyone waiting. Sometimes, physicians do not succeed in obtaining the right information from patients due to the lack of time.
We specifically talked to Doctor Carl Berdahl at Cedars-Sinai Medical Center. He asked us to create a well-designed tool that could facilitate communication between patients and doctors to improve diagnostic accuracy. He wanted to tool to be able to ask patients questions and save that information for physicians to read later. He also asked the tool to be able to translate to different languages, specifically the six most common languages spoken at their specific hospital.
UNDERSTANDING THE USER
There are two users of our application: the patient and the doctor.
The patients will use the application to fill out their personal information and answer diagnostic questions. The patients will vary from age, social backgrounds, wealth, and more. The patients may be in pain and thus might not want to look at a device for long to answer many questions. The patients will vary by a lot and thus it is important to create a simple interface that will be simple and easy to use for all patients.
On the other hand, we have the doctors. The doctors are going to be the ones looking at their personal information and diagnostic in order to better help the patient. They will be looking at the data collected. We want the data to have a quick and easy-to-read format, so that there is no time wasted and to ensure that the doctors will not confuse a patient's information with another. Mistakes made could be catastrophic.
BREAKING DOWN THE PROCESS
It all started with a conversation with Dr. Carl Berdahl. It continued to research about what makes a good history-taking for doctors and design choices to make the communication between doctors and patients more efficient.
The process began when we spoke to Dr. Carl Berdahl about a tool that could facilitate communication between patients and doctors to improve diagnostic accuracy. One of the issues is that people have to wait a long time in the emergency room. There needs to be more efficiency in the emergency room. One to do so is to ask the patient questions before they see an actual doctor. We talked about different ways to ask about the patient's pain and the demographics of patients to better accommodate them. After a few meetings, we did research about what it means to take a good history of the patient's personal information and learned about different symptoms of chest pain. We then made some design considerations such as what format the questions were going to be asked, what the ratio of text and pictures were going to be, and different ways patients could answer the questions.
THESE WERE SOME MAJOR POINTS IN OUR DESIGN CONSIDERATION
Languages
According to the demographics in the Cedars-Sinai Medical Center in Los Angeles, the most common languages spoken were:
English
فارسی (Farsi)
Español (Spanish)
中文 (Chinese)
한국어 (Korean)
Tagalog (Filipino)
Personal Information
We could speed up the process of emergency rooms if we ask patients about their personal information while they are waiting rather when they are meeting up with the doctor. Personal information questions include name, gender, age, weight, height, and which insurance they have.
History-Taking
History-taking is crucial to diagnosing patients correctly, so we want to me able to understand what the patient's pain feels like and how it changes in the most accurate way. We focused on the onset, provocation/palliation, quality, region/radiation, severity, and timing of the pain.
Variety of Users
Other than just languages, we wanted most, if not all, users to be able to use our chatbox. We decided to ask questions in the simplest way possible. Users would answer by text, multiple choice, and sliding a bar. We chose the most simple option for each answer.
Chatbox Functionality
We chose a conversational style rather than a multiple choice style. A study found that patients would rather speak to a doctor face to face than to just answer questions in a form. We opted for a chatbox in order to give users the feel of communicating with a real doctor.
Collecting Information
We wanted the doctors to be able to view this information easily. The doctors can view the information in an excel format. Doctors can also view trends of all the patients, which can be beneficial in preventing common issues for future patients.
THESE ARE MY LOW-FIDELITY MOCKUPS
HERE ARE TWO OF THE CHATBOXES IN ENGLISH AND CHINESE
THERE ARE SEVERAL WAYS TO CAPTURE INFORMATION FOR DOCTORS TO VIEW
There are two ways to view the forms via collect.chat. The first is viewed through the collect.chat website. The doctor is able to view all the patient's answers and is also able to delete some of the responses that may be errors or possibly even delete responses of patients who have already been checked out to respect the patient's privacy.
The second way to view the information is through a CSV file. The doctor may be able to save the CSV file at the end of the data to keep track of how many patients have gone to the doctors that day and to look for common problems in pain in one chart for analysis.
There are also ways to analyze the information captured via collect.chat. The first is a bar chart. It takes all the answers that are in multiple choice format and separates them into a bar chart. It gives the total people that answered the question and the percentage of each option chosen.
The second way to show this information shows the metrics of the data collected. This chart shows the interactions on selected dates. This can help with showing how effective the chat box is. It shows the total amount of collections and attempts on different days.
TITLE OF THE CALLOUT BLOCK
LESSONS LEARNED
Communicating with the client before designing creates better designs. There are many design and primary end-users considerations to be made in projects.
This was the first time where someone asked us to build something for them. I learned that communicating with the user (Dr. Carl Berdahl) is very important as I want to be able to create a tool which will be most beneficial for his cause. I learned that asking him questions before we start designing and building is important, especially in short-term projects such as this one.
There were many design and primary end-user considerations, since there were a variety of users. There were the patients and the doctors. The patients can vary by age, race, wealth, education, and more. There are many design considerations in designing for such a wide range of users. Should we use pictures or pure text? Should we use multiple choice or a conversational style? In terms of primary end-user considerations, we wanted to be able to have support for patients and physicians. Although we only created a chatbox for chest pain, we could take this further for pediatrics, orthopedics, and more. We also wanted to make sure the back-end information was readable and structured for the physicians to read. We found that it is important to put design considerations for all the different groups of primary users.
An individual response can also be emailed to the doctor. The format includes the exact question and answer in the chat box. This can be helpful if there are different doctors on a given day. The patient information can be sent to a respective doctor in order to speed up the process and can also respect the privacy of the patient because their information will only be sent to the doctor they are seeing.