clinical trial safety tracing

overview

How do you track and trace tens of thousands of patients in the wild? Janssen Pharmaceuticals needed our team to help build a solution for tracing patients who visited hospitals so they could be quickly assessed for risk of infection due to E. Coli. We combined our Linkt mobile app and Admin Portal with additional features to help manage this unique use case and deliver on a tight schedule and changing requirements.

Company: Datacubed Health, 2019

Role: Chief Technology Officer

Team: UI Designer

Problem

Janssen Pharmaceuticals needed to conduct one of the worlds largest vaccine safety studies for E. Coli. But the rate of incidence was estimated to only produce side effects in .06% of participants. To successfully collect enough data, they would need to run a multi year study with over 35,000 participants.

Not only would they need to collect data from the participants, but they would need to keep them engaged over a 3 year period, and attempt to automatically detect when a patient visited the hospital.

Challenges

How would we be able to effectively engage with the population that was typically age 50+ and had a lower utilization of smartphones?
Could we effectively track hospitalizations in a way which was acceptable to the users?
How could we effectively get the information back to the clinicians in time for them to catch the potential infections or side effects?

Solution

The Linkt Platform had three components to enable success for this project, and was selected as the winner among over a dozen companies as part of a worldwide innovation challenge and RFP process.

A fun, gamified user interface which could provide educational content and mini games to encourage engagement during “healthy” periods.
HIPAA-Compliant location tracking technology which allowed for alerts when a participant visited a geofenced hospital or clinic.
A case management system which would allow clinic or hospital workers to trace and confirm infections and reported adverse events would need to be built to complete the solution. Our original proposal was not selected as part of the first RFP (see below).

I represented the team to Janssen leadership throughout the project, created and documented product requirements, UX workflows, and supported project management for the global project rollout in 3 countries.

Individual screens and UI components were built by the Senior UI designer as part of our existing administrator tools.

A fun, gamified user interface which could provide educational content and mini games to encourage engagement during “healthy” periods. [App screenshot]

HIPAA-Compliant location tracking technology which allowed for alerts when a participant visited a geofenced hospital or clinic.

A case management system which would allow clinic or hospital workers to trace and confirm infections and reported adverse events would need to be built to complete the solution.

Engagement Components

Above: Examples of the engagement content including brain games, a digitized patient information packet, weekly articles.

Reacting to major changes

A runner-up company was originally selected to handle the case management side of the project. Our team was awarded the work after their solution was deemed too complex and needed to quickly build and implement the case management tool which would send SMS messages whenever a participant entered a known hospital. The entire system was designed and built in approximately 2 months in order to hit the strict regulatory timelines. I focused on architecting the requirements, leading the team to build a lean solution, leveraging off the shelf technology where feasible, and making sure we stayed on target with deadlines and milestones.

Case Management Solution

Working with the specific clinical requirements, I created a workflow which would allow us to capture these specific hospitalizations, automatically determine whether or not they should trigger a notification and a creation of the “case” - a set protocol that the clinician would need to review to determine if the hospitalization was a result of an E-Coli infection or other possible adverse event.

To start the process, a participant would receive an alert whenever they entered a geofenced hospital that was participating in the study. To avoid the risk of false positive, clinical sites were specifically chosen in more rural areas in the US, Italy and Japan in order to limit the number of total geofences.

The alert would trigger the conversational chatbot interface with our avatar Humi who would ask a series of yes/no questions to determine if a case should be created.

This data would also inform the trial investigator whether or not the tool was working effectively.

Once a participant reported a potential infection, a case would be created and an SMS alert sent to the clinicians informing them to the steps they must take, as well as information about the report and hospital visited. This would allow the clinician to request lab results, confirm the infection, and make case notes. The automated workflow would also trigger a 28 day follow-up survey to the participant.

This process automated a process that typically involved a clinical case worker calling by phone EVERY regional hospital and asking for records on a set list of patients. This process is typical in clinical work and is one of the reasons that large scale trials fail to produce results, as it is incredibly labor intensive, time consuming and prone to human error.

Outcomes

This research trial began rolling out in late 2019. It represented the largest deal in Datacubed’s history and is well timed for a pivotal shift in healthcare technology as COVID-19 further disrupted traditional on-premise clinical trials. The Linkt Platform was offered for free to all educational institutions and healthcare customers for work supporting COVID-19.

The initial project rollout will support 6,000 users over 3 countries, with the 35,000 person trial intending to begin in early 2021.