The lifecycle of a clinical trial matching mechanism starts with importing trial protocols into the system based on established medical oncologies. Genospace and Sarah Cannon curation teams use Clarify to establish and manage a central knowledgebase of trials and precision medical information. More specifically, the curation team is responsible to convert the protocol language (including amendments) into machine-readable inclusion/exclusion criteria with different weights. The standardized information will then be processed by the backend system and be used to match patients to the most suitable trials. As more trial protocols get entered into Genospace, along with constant amendments being added to protocols, curation team face challenges curating trials, and monitoring and validating changes quickly.

There are also limitations with the rather outdated interface which slow the curation efficiency. For example, the old Clarify layout doesn't enable recognition of output effectively. The rule creation process was also rather tedious, requiring many clicks, therefore slowing down curating efficiency. Certain functions (eg. version control) are missing, which make QA process more time-consuming.

At the same year, one of Genospace's business goals is to improve trial match specificity by 20%, which naturally requires more accurate trial curation. It was the perfect timing to start the Clarify Improvement initiative. (add some pics to demonstrate the current state)

The Exploratory Research Process

As my very first product overhaul initiative at Genospace, I started off by interviewing with every curation team member to drive the problem-solving process. I conducted 4 individual contextual inqury interviews, faciliated 1 group disucssion. During my research process, I adopted the Job-to-be-done framework to help understand the curation workflow and uncover pain points and opportunities.

The Users and Things We Learned from Them

After the initial round of interviews with the curation team and evaluation of the application current state. I generalized the key user goals and opportunities for product improvement.

Who are we designing for?

While curation scientists are the main power users of Clarify, clinical research program managers can also benefit from Clarify to finish trial feasibility assessment. However, there are some major pain points that block curation scientists from curating more efficiently, and a lack of certain feature for program managers to really adopt the tool:

While we identified that there are multiple areas of improvement that can be made for Clarify. It would be a big engineering lift to implement all features at the same time along with other company initiatives. Therefore, we broke down the features by feasibility, desirability, and viability. We also evaluated individual feature by the effort/impact scale to decide on which feature to develop first that can yield largest impact for users. (add a pic of feasibility evaluation)

After careful evaluation with the product and the engineer teams, we decided to implement the Clarify Improvement Initiative in 3 releases:

- Phase 1: Enable Clarify Versioning Control

- Phase 2: Clarify Usability Fix & Interface Upgrade

- Phase 3: Clarify Feature Expansion - Trial Feasibility

Since this product is mainly used by a small team of curation scientists, I tried a new way of working with users, bringing them into every stage of the design process. By meeting with the curation team regularly to review design and get concrete design feedback, together we prototyped, built and refined the designs to create a simpler and a more streamlined trial curating experience.

The Clarify redesign process is a rather complicated process that involves much detailed workflow mapping and interaction design. Throughout the concept development process, I benefited much from constantly reviewing the design with my user panel - the curation team and prototyping with real data and concrete examples.
It helps me take the complexity into consideration from the very beginning. Among all the different aspects of the design iteration, below are some examples of how my designs evolved along the way:

The Layout

The original layout of the Clarify curating interface was a 50/50 form where user can scroll the left side content and fill out the form on the right. While it's useful to have the curation and the editor side by side, the original layout wasted a lot of space and requires a lot of scrolling.
My initial exploration around better curation UI was rather conservative. I wanted to create a design that needs minimal engineering effort. But as I spoke with users and the engineering team, it turned out that the layout change wouldn't be a huge lift. Users can also benefit a lot from having more space when editing trial criteria and rules. I then started to pursue a full-width layout and explore how to best navigate around it.

The Rule Builder Exploration

The rule builder is one of the most chanllenging part in developing the new UI for Clarify because the rule builder is essentially a combiination of a form and a query builder that translates plain langugae into system-recognizable data. On the one hand, user needs to be able to quickly transform text into structured data, on the other hand, user also needs the flexibility to construct complex logic that feeds into the matching algorithm.

More specifically, user needs to be able to map data onto different established medical categories. User also needs to create different condition groups using the rule builder. The original rule builder was quite comprehensive, but it lacked visual clarity and can be extremely hard to read when a rule gets very complicated. I spent a lot of time exploring different styles and components to construct a powerful yet elegant rule builder. I also gained user feedback not only from curation team but also general users to make sure the interaction is truly easy to use.

As a bonus result of the exploration, the rule builder concept was adopted by other team members in other products.