←←All cases

Data-driven diabetes management system for better care

Client

A leading innovative medical device company

Industry

Healthcare, Biotech

Services

Product Development, Data Architecture, Data Management

Tech

ReactJS, Node.js, Python, Docker, Azure services, Databricks, DeltaLake

Challenge

People with Type 1 diabetes who need to take insulin to keep their blood sugar under control require quick access to data on their glucose levels. This data is critical for adjusting their lifestyles, a fundamental aspect of diabetes care. However, they have to wait for three months until their next scheduled check with a healthcare provider to take measurements and understand if they are meeting glycemic targets. A leading innovative medical device company wanted to solve this problem by leveraging data from its automated insulin delivery system. Combining an on-body insulin pump, controller app, continuous glucose monitor (CGM), and algorithms to automate insulin delivery based on CGM sensor values, the system automatically collects important data every few minutes round the clock. The company was understanding that this user data could be put to work to address unmet needs and improve diabetes care. So they wanted to build a data-driven diabetes management system to deliver near-real time reports and validated actionable insights to healthcare practitioners. They applied for help to ITRex, an experienced partner in data analytics solutions.

Our task was to:

●

Design and build a diabetes management data analytics and reporting system, both the front end and back end

●

Perform integration with data sources for data ingestion and other necessary integrations

●

Run comprehensive testing, including unit tests and automation tests with coverage from 70% to 100%

Solution

A HIPAA-compliant data-driven diabetes management system that collects raw data from the insulin pump and continuous glucose monitor using an ETL process to transform it into reports and valuable insights assisting healthcare providers in clinical decision-making. It has the following key functionality:

●

Report generation using data collected by the insulin pump and CGM every few minutes and dozens of formulas to provide reports in a standardized format complying with the industry’s strict regulations. Reports show daily patterns and weekly summaries of important metrics, such as average glucose, glucose variability, and time-in range (the time the person spends within the target glucose range), with daily glucose readings blended into one graph. Time frames can be customized by clinicians

●

Portal for clinicians integrated with the users’ electronic health records (EHR). It provides access to reports and has patient management functionality allowing clinicians to filter and update cases, print them out, and download them

●

In-app insights for users generated based on the reports. They help users stay in the target glucose range by tracking patterns. Such insights include events when the user’s sugar was below or above the target range, and graphs showing the glucose timeline

diabetes management system for medical device company

Impact

●

Generating automated reports with critical data and actionable insights, the data-driven diabetes management platform allows clinicians to make informed decisions about diagnosis, treatment selection, or treatment adjustments relying on unbiased CGM data gathered under conditions of daily living

●

Users are enabled to easily track their glucose levels in real time between appointments without fingersticks, with the system giving them in-app access to reports showing how well they are managing their glucose levels

●

In a matter of months, users have made significant improvements in their diabetes outcomes, seeing increased time in range and less hypoglycemia/hyperglycemia, according to research findings