dr lan - Changhua Christian Hospital CHATBOT
–
YEAR: 2018-2020
pRODUCT pLANNING: JC YEH
UX & ui DESIGN: JC YEH, Emily wang
visual design: CONNIE KANG, Emily wang
character design: CONNIE KANG
user research: jane wang, Yin Peng, JC YEH
Dr Lan is an AI-based chatbot aiming to improve an outpatient's hospital visiting experience from noticing the first symptom till he finished the treatment. As a caring companion, Dr Lan not only helps a patient making appointments or getting health exam reports, he follows up on the patient's problems and taking notes for the doctor before visiting the hospital.
Goal
Symptom checking assesses a patient by asking a series of questions about their symptoms to predicting potential diseases and suggests related departments. Our goal is to create a chatbot that uses the symptom checker AI to help the patients.
Design challenge
1. The hospital already offered a patient service app and a website. To grow the new chatbot service's user base, we need to create distinctive features and deliver an intuitive experience.
2. Interacting with a chatbot for a task with multiple processes can be time-consuming and tedious.
OUTPATIENT JOURNEY MAP
Research Method
1. Desk Research
2. Contextual Inquiry
3. InterviewParticipant
4 Patients with chronic diseases
2 Doctors
2 Nurses
4 Case managers
After getting insights from user research, we use customer journey mapping to uncover areas of opportunities. We analyze each step based on patients' perspective and align them with insights from the medical staff they've encountered in specific stages.
Patients need assistance and care at every stage of this journey. Doctors need someone to collect valuable information about their patient's symptoms and treatment before seeing them. Dr Lan can serve as a virtual case manager who fulfils both.
“The best bots, usually have a “helper” element, as if your best friend, or, even better, your mother, is helping you complete a task.”
STRATEGY
Dr Lan should become a case manager who knows what the patient needs, whether just attention or friendly reminders. He should offer his help anytime and anticipate what he can prepare for the patient even without asking. Dr Lan also curates patients' information for the doctor in advance.
Caring
Messages with great sympathy. Friendly reminders of the action that the user needs to take.
Timing
Carefully design the timing of notifications and caring messages.
Understanding
With real-name authentication, Dr Lan in sync with the hospital's database and knows what service the patient needs.
UX DESIGN
Before Hospital Visiting
During Hospital Visiting
After Hospital Visiting
UI DESIGN
FEEDBACK & DESIGN UPDATE
Learning from: Product Data
Issue: Users drop-off during symptom picking task
Due to the technical restrain of LINE Chatbot, the early design of symptom body map can only be cards in the message. It is challenging for patients who don't understand their symptom to navigate through these cards. With the new LIFF browser component from LINE, We can redesign the body map into an interactive web page - the user can jump between body parts to find his symptoms.
Learning from: Product Data
Issue: The timing to ask the user about their status for pre-visit note is too early.
Dr Lan asks the user about his status in the evening before he visits the hospital. Few months after the first launch, we noticed that users replied to this message and filled in the information needed while waiting for the doctor in the hospital.
Video from National Center for High‑Performance Computing and CCH