Babson Diagnostics and ClosedLoop Team Selected as 1st Place Winner of ‘Help with Hemolysis’ Data Analytics Challenge

Austin, Texas — Babson Diagnostics, a science-first healthcare technology company, and ClosedLoop, the leading healthcare data science platform, today announced their team was selected as the first-place winner of the ‘Help with Hemolysis’ Data Analytics Challenge. As winners, the Babson and ClosedLoop team will present their approach for using data science to solve the real-world laboratory challenge of hemolysis during a scientific session titled “Blood & Bytes: Reducing sample quality errors in clinical laboratories using data science” at the 2023 AACC Annual Scientific Meeting & Clinical Lab Expo in Anaheim on July 23-27, 2023.

Babson Diagnostics and ClosedLoop have worked together closely since 2021, when the two companies collaborated to develop artificial intelligence solutions, which help detect and prevent sample quality errors that can occur when working with capillary blood samples. So it was a natural fit for the companies to team up to tackle the ‘Help with Hemolysis’ challenge.

The second annual data analytics competition, co-hosted by the Washington University in St. Louis Section of Pathology Informatics and the AACC Data Analytics Steering Committee, invited participants to test their analytics skills by developing an algorithm for retraining phlebotomists to reduce sample errors from hemolysis. The winning team was led by Babson’s Founder, Chairman and Chief Operating Officer Eric Olson, his son Ethan Olson, and ClosedLoop’s Chief Technology Officer and Co-Founder Dave DeCaprio – and placed first out of 17 teams from across the United States.

 

“Hemolysis is one of the biggest sample quality problems that clinical laboratories face whether they are working with venous or capillary blood samples.”

Eric Olson, Founder, Chief Operating Officer, Chairman

 

“Combining Babson’s expertise in hemolysis and sample quality with ClosedLoop’s experience in artificial intelligence and machine learning enabled us to work the problem biologically and computationally at the same time," Eric Olson said. "We appreciate the opportunity to collaborate with the teams at Washington University in St. Louis and the AACC to tackle a challenge with real impacts on people’s lives.” 

In vitro hemolysis is the breaking open of red blood cells during blood specimen collection – and is the most common preanalytical laboratory error and the most common cause of sample rejection – leading to delays in clinical results, increased costs, and decreased patient satisfaction. Educational interventions, which have been shown to reduce hemolysis rates, are often resource-intensive.

This challenge invited participants to rank all blood specimen collectors at a major tertiary care center based on how much money would be saved over the following year by re-educating that collector on hemolysis prevention techniques. The goal of the solution would be to maximize the reduction in hemolysis costs by targeting all available educational resources in the most effective way possible.

“In vitro hemolysis can affect up to 3.3% of all routine samples, leading to sample rejections that delay critical results and increase patient and institutional costs,” said Dave DeCaprio. “We felt uniquely positioned to solve this challenge with our partners at Babson Diagnostics. We combined Babson Diagnostics’ expertise in maximizing the clinical utility of blood testing with ClosedLoop’s AI/ML platform to build a predictive model that identified precisely which individuals should receive education.”

The winning data models developed by this competition will directly inform institutions to spend their educational resources more efficiently to maximally prevent in vitro hemolysis to ultimately improve patient care and decrease costs.