HOME -> TACKLING BIAS AGAINST WOMEN IN HEALTHCARE

Tackling Bias Against Women in Healthcare

TL;DR

Kas combines a symptom journal with AI to generate questions to help users have more productive conversations with their physicians about gynecological concerns.

Created for capstone and the 2025 SSDC.

Prompt & Research

Following the 2025 SSDC prompt of addressing bias in various industries (and bonus points for implementing AI), my group decided to tackle the bias women face when seeking medical care, specifically that of dismissal and medical gaslighting.

During our research phase, my main role was to interview industry experts, where I had the pleasure of talking with Emily Dwass, the author of Diagnosis Female: How Medical Bias Endangers Women's Health, Emily Tousaw, a family doctor in Stratford, and Lisa Wilde, the executive director of The Emily Murphy Centre.

Problem Statement

Our group identified the issue that many women face dismissal and misdiagnosis with gynecological issues due to long-standing bias in a male-centred medical field, and that many patients take doctors' advice at face value.

Introducing Kas

To address the problem, we created Kas, an app-based service that promotes patient advocacy by generating questions to ask your doctor based on symptoms you've entered.

Kas has 2 main functions:

First, a journal that helps women keep track of their gynecological symptoms.

Second, Kas uses the data you've entered to generate questions to ask your doctor during your upcoming appointment.

Prototypes & User Testing

Service Design Blueprint

Although Kas is an app, it’s part of a larger service. To better understand how Kas might fit into the real world, we developed a service blueprint to better visualize a user's journey.

Wireframes & Low-Fi Prototype

Afterwards, as a team, we sketched crude wireframes on a whiteboard, allowing us to quickly refine our ideas. We polished Kas into its (now) key features: logging symptoms, generating questions, and providing insights.

User Testing

We conducted user testing using the RITE method with 6 users over 4 days.

From these tests, we learned 3 key details: 1) that our UI was difficult to understand and it took too long to enter symptoms, 2) our features that incorporated AI were too wordy and users didn’t understand their purpose, and 3) a careful balance of friendly but medical language would be key to communicating clearly with our users.

Hi-Fi Prototype

We then translated our prototype into a higher fidelity version, keeping in mind our recent findings.

Our new prototype boasts a simpler UI with a greater emphasis on a strong visual hierarchy, fewer words overall, new custom icons, and a new onboarding process to help users better understand Kas' features.