et al Health

Overview

et al health is a search engine to help people with rare diseases find a qualified specialist. Our algorithm recommends doctors based on their academic research activity. I led visual and front end design and conducted user interviews and competitive analysis. et al Health placed 3rd in our Big Ideas at Berkeley category. We also won the Berkeley I School Chen award, which recognizes innovative Master’s thesis work.

Identifying A Need

flow chart of et al health data
Diagram of information flow in et al Health.

Currently the NIH recommends finding a rare disease doctor by searching PubMed for doctors who actively research the disease and contacting them. A data scientist in our master’s program saw an opportunity to simplify this process for patients, some of whom may have never encountered academic research before.

I joined his team, excited by the challenge of translating scientific journals into an acessible form, and the opportunity to help the rare disease community, which includes 30 million Americans.

Testing Our Assumptions

We began to conduct user research to test our hypothesis that this would help patients find providers, and to understand how it would fit into their search process. et al Health, along with many rare disease patient advocacy groups, operates on the assumption that rare diseases share enough commonalities to host services on common platforms.

We have tested this assumption from the beginning by focusing on Castleman, a little known disease, and ALS, which has more publicity and cases. 20 interviews with doctors and patients later, we began to understand the complexity of their search behavior and establish personas, which are summarized below.

Identifying Our Users

Since users bring a wide range of skills and knowledge to our site, it was important to categorize and refer back to three main groups during the design process.

The Patient

The Doctor

The Parent

A person directly suffering from a rare disease. May not have a direct support network. Feels alone, because their condition isn’t well-documented or supported by the healthcare system. Diagnosed a patient with a rare disease that they had never heard of. They care about their patient but it’s out of their responsibilities to find them a provider. A family member or close friend has a condition, and they are assisting in finding treatment. They spend significant amounts of their free time doing research on the condition.
patient portrait doctor portrait parent portrait
“No one can relate; they haven't even heard about it. The diagnosis took months and finding a doctor turned out to be a struggle as well. ” “I don’t have time to give my patients the care they deserve. If I want to help, it's in my spare time.” “I didn't stop asking questions or trying to learn about things I didn’t understand. I had to do it for my son.”
  • Tired and/or otherwise incapacitated
  • Has geographic concerns - worried about traveling far for treatment
  • Knows a little about medicine
  • Has too many patients to care for
  • Averse to new technology/time wastes
  • Medical expert; but not up to date in the latest/obscure fields
  • No medical background but doing lots of medical research
  • Tenacious; not afraid to challenge doctors and make demands
  • Bouncing between care providers

I illustrated the proto personas and contributed to their development

Interview Key Findings

Wireframes

We approached UI design by first quickly hand sketching multiple divergent site layouts. We identified the strengths of each and I began to create wireframes in Sketch. These went through several iterations, and I made an InVision prototype out of the pages to conduct our first round of usability testing.

Below are two pages from the InVision prototype

User Testing Key Findings

After conducting six usability tests, we learned the following:

Visual Design

Our team's belief that design can be both beautiful and accessible drove the visual design process. We also faced the constraints of dynamic content- the layout had to accommodate varied assets and empty states. As the visual design lead I approached these problems by working with live data whenever possible, using Google Material design as a starting point, and holding frequent design reviews.

Results

et al Health is a responsive web app with live data for 4 diseases, including Castleman and ALS. et al Health placed 3rd in our Big Ideas at Berkeley category, and won the I School Chen award.