Drug-drug interaction risk calculator

Project description

In modern healthcare, patients often rely on multiple medications to manage various illnesses, especially as the global population ages. This increases the risk of drug-drug interactions (DDIs), which can either enhance or diminish a drug’s effectiveness or even introduce dangerous side effects. Alarmingly, in the USA, 9% of people over 55 take ten or more prescription medications, and in Europe, DDIs are responsible for 197,000 deaths annually.

To address these risks, pharmaceutical developers need tools to predict potential interactions early in drug development. The Drug-Drug Interaction Risk Calculator (DDIRc), integrated into PharmaPendium, was created with this goal in mind. PharmaPendium helps users benchmark new drug candidates against approved ones to identify possible approval blockers.

Collaboration and Development

DDIRc was developed in collaboration with leading pharmaceutical companies such as Sanofi, Merck, Servier, and Boehringer Ingelheim. In return for their expertise, these partners received early access and contributed to shaping the tool. Our initial focus was creating a comprehensive user journey, mapping pain points and opportunities, which guided the scope of our first releases.

Enhanced User Workflow

The latest iteration of DDIRc introduces a more intuitive workflow. Users can now store and compare multiple simulations, enhancing their ability to evaluate drug interaction risks.

Input Form: One of the most intricate aspects of this project, the input form required extensive collaboration. Together with designer Tina Bota, we improved the form’s structure by organizing it as a wizard, guiding users through each step with clear, logical progression. We also ensured transparency by communicating the sources of calculated values and offering default predictions when necessary.
Prediction Results & Visualizations: The result pages feature visualizations tailored to various expert needs. Users can access an overview of interactions or dive deep into specific drugs through forest plots or tables. After two rounds of user research, we focused most improvements on this area.

DDI-form — Redesign changes after the study findings

UX Research

As our team lacked a dedicated UX researcher, I leveraged my experience to lead two UX studies. This involved setting up alpha and beta testing sessions, where we observed user interactions, conducted moderated sessions, and gathered feedback to refine the tool.

Alpha Testing: We began by observing users' natural interactions with the tool, followed by structured moderated sessions to gather deeper insights.
Beta Testing: This phase focused on improving essential elements like navigation, filtering, and visualization alternatives. We conducted workshops to share findings with the product and tech teams, driving further enhancements.

Wall — From Miro - important moments of the interview shared with team members. Raw data connection never lost.

results.png-2 — A slide from the result presentation

Excel-scorecard2 — Excel scorecard doesn't look very appealing while it was very well accepted by product folks. Sometimes we need to talk/present differently.

prioritization — Prioritization together with product management and dev team

All findings transferred into Jira tickets mainly by me and got some blessings of PM and tech lead.

Reflection

DDIRc is a critical component of PharmaPendium, marking a shift from mere data extraction to providing predictive insights. The complex logic embedded in the input form increased the testing scope, but this complexity also led to higher retention and user satisfaction. Customers frequently ran 40-45 simulations per project, signaling the tool's effectiveness.

Looking back, conducting remote user studies was a valuable learning experience for the entire team, fostering an appreciation for UX research and inspiring more such efforts in future projects.

Date

2020