Elevating Medical Decision-Making
A mobile-first app, and content management system for Zenxmed, a physician-first content delivery platform.
A world where Emergency Physicians could use a mobile-first, open-access, cloud-based application designed to support them and their patients in making informed diagnostic and treatment decisions. That was the value prop for Zenxmed. Provide a platform that simplified access to medical information, fostering learning and collaboration.
Keeping Pace
Zenxmed was founded by physicians who recognized the significant challenges in the healthcare sector, including rising costs and the overwhelming amount of constantly evolving medical information. Many practicing physicians struggled to keep up with best practices, further hindered by the limited support for their needs within the industry.
The founders envisioned a resource that would streamline access to essential information, allowing healthcare practitioners to make informed decisions in real-time, right at the point of care. This led to a clear mission for Zenxmed: to develop a user-centric platform that enhances decision-making and fosters collaboration among healthcare professionals.
User-Centered & Iterative Design
To bring the flagship product, OpenXMed, to life, a flexible user-centred design process was implemented. This included the usual suspects; user interviews with 2 dozen medical physicians who worked across emergency and family practice care, iterative design sprints, and continuous testing. The primary goal was to understand the real-world challenges, preferences, and behaviours of physicians in order to develop a platform that seamlessly integrated with their day to day activities.
We also conducted a survey with just over 800 target users such as physicians, medical residents, and students, in an effort to map their journeys and identify areas where OpenXMed could provide the most support. Two main hypotheses were formulated:
Hypothesis 1: Physicians who can swiftly navigate complex clinical guidelines will have greater confidence in their decisions, reducing the need for assumptions during the point of care.
Hypothesis 2: Implementing a contributor model instead of traditional authorship will encourage physicians to engage actively, fostering a collaborative environment that prioritizes accountability and shared medical knowledge.
An interesting side note to all this research was how quickly I had to develop my understanding of the domain, the literature, and its audience. I’m not planning to take a swing at medical school, but it was neat to actually learn a lot of the industry and esoteric terms they use. Another side side not. Physicians are generally understaffed, and are notoriously busy, but when you do get to sit them down, they are remarkably easy to talk to and they themselves are quick studies when it comes to ux research.
An Intuitive Mobile Experience
The aim with the app was to ensure that physicians and students could navigate through massive amounts of content quickly, in a logical order based on decisions they would make when with a patient. The solution: a decision tree format. A format us engineering and design folks are very familiar with!
This format allowed layering of additional information and references in the content, reducing the cognitive load for physicians at point-of-care.
Another key part of the platform was letting physicians comment on evidence, which helped medical students connect with peers and learn from practicing clinicians. User testing suggested this engagement could improve the algorithms over time, so an analytics strategy was defined to measure it.
Because most clinical decision tools lack meaningful peer interaction, robust profiles were designed that highlighted specialty, interests, and medical training to reinforce credibility. Onboarding collected only what was needed to verify users, making core identity fields required and secondary details optional, and clearly explaining why the information was requested.
Managing Content
Once the mobile, point-of-care experience was defined, a backend was needed to manage and deliver content.
This was approached in two phases:
Phase 1: Creating algorithms without a CMS using draw.io and a standardized contributor format, targeting 100 algorithms in six months, while engineering started the first CMS iteration.
The CMS workflow was intentionally simple:
- Build a topic list
- Assign each topic to a contributor and a medical lead
- Capture the required clinical details
- Provide a visual builder for algorithms
- Review, then publish to the app
Phase 2: The CMS MVP enabled migration of the phase one library into the platform, building algorithms directly in the system, and formalizing the review process with an in-product notes and feedback loop.
The key flows were also important, and the ability to make updates quickly were crutial to the success of the mobile app.
Evaluating Success
Usability: The decision-tree interface of OpenXMed provided physicians with an intuitive tool to access medical information efficiently, significantly reducing decision-making time at the point of care.
Collaboration: The shift to a contributor model promoted an environment of peer engagement, motivating users to participate actively in knowledge-sharing, which proved valuable for both experienced and aspiring medical professionals.
Scalability: The two-phased CMS approach enabled Zenxmed to rapidly expand its content library and support ongoing contributions, establishing a strong foundation for future growth.
