The team created HRO-ML Health, a tool that provides health care organizations a way to assess their performance.
The Innovation Hub at Research Park is proud to support the team behind HRO-ML Health, a tool that will help health care organizations assess their performance regarding the standards of High Reliability Organizations (HRO) through machine learning (ML). The project was recently accepted into the National Science Foundation Innovation Corps Program (NSF I-Corps) at Texas Tech University.
HRO-ML Health was created by Timothy Matis, an associate professor in the Department of Industrial, Manufacturing & Systems Engineering (IMSE) in the Edward E. Whitacre Jr. College of Engineering and the director for the Center for Excellence in High Reliability Organizations & Processes (CEHROP). The team includes Abhishake Kundu, an IMSE graduate student, and Bjorn Sorenson, a former graduate student in the Healthcare Administration Program at the Texas Tech University Health Science Center. Matis serves as the CEO and principal investigator, Kundu as the chief technology officer and technical lead, and Sorenson as the chief marketing officer and entrepreneurial lead.
This startup provides a near miss- and safety incident-reporting platform designed for ease of use and uses a machine-learning algorithm to score those reports against a scientifically backed rubric to assess strengths and weaknesses in the organization's practice.
HROs are organizations that operate with exceptional safety records in industries in which a system failure would be catastrophic to both the organization and society. It was coined by researcher Karlene Roberts and later expounded upon by Karl Weick and Kathleen Sutcliffe in "Managing the Unexpected," a book empirically summarizing the management practices of these HRO organizations into five hallmarks of practice.
The five management practices that constitute the hallmarks of an HRO organization are:
- Preoccupation with failure: the organization will be on the lookout for anomalies in routine operations, no matter how small.
- Reluctance to simplify interpretations: the organization will treat those anomalies that are discovered as being symptomatic of an array of larger problems and will not rush to the quick and easy explanation.
- Sensitivity to operations: the organization will keep operational needs paramount when making business decisions.
- Commitment to resilience: the organization will prepare their plans for reacting to a broad range of operational disruptions.
- Deference to expertise: the organization will flip the managerial hierarchy in times of operational disruption to let those with the most knowledge of the disrupted operation take charge in restoring the organization to normal operations.
Since the publication of the book, several organizations have sought to mimic or adopt the HRO practices in their operations to avoid catastrophic failures. In 2011, The Joint Commission for health care accreditation championed the adoption of the HRO hallmarks to improve patient safety.
Throughout the course of his career, Matis has worked in health care operations engineering, organizational safety and probability and statistics. In 2011, he founded CEHROP and, through his work, developed a deeper understanding of the science behind the hallmarks of an HRO.
"I noticed there was a large gap in measuring HRO performance," Matis said. "That was the missing link in being able to help the HRO theory be adopted into practice."
Matis said there currently is no other tool available for health care organizations to measure their success with HRO hallmarks.
"This is where we come in," Matis said. "Our software provides a measurement that a health care institution may use to make targeted improvements to align their practice with those of an HRO organization."
Matis pointed out that while many individuals and products seek to tell organizations what they should do in a general prescriptive sense, there is nothing in existence able to measure HRO performance and provide a data-driven approach to target resources in an organization's journey.
"Bringing my knowledge and experience together, HRO-ML was started as a means to provide a tangible product to do just that and make a significant improvement in health care and patient safety," Matis said. "The health care system is complicated, and I would say that both nobody and everybody is to blame for this. I've never met a health care provider or administrator who doesn't want the best care possible for patients. The product we have developed supports both the provider and administrator in reaching these goals."
The journey to acceptance in the NSF-I Corps program began when Matis spoke with Bryan Norman, department chair of IMSE. Norman pointed Matis to the NSF-I Corps program at the Innovation Hub.
"When I approached Abhishake and Bjorn about the idea, both of them latched on without much persuasion," Matis said. "After a few initial Zoom meetings and a trip to Fort Worth to meet face–to-face, we hammered things out and competed in the regional I-Corps competition. As time has gone on, the HRO-ML team has bonded, and we formed an industrial advisory council consisting of health care executives and providers from across the nation, to help guide us in our startup.
"After receiving the recommendation from the regional program for nationals and successfully completing the interviews with the program officers at NSF, we are slated for the second spring cohort. We look forward to further developing our business model in the national program of NSF I-Corps, then seeking funding through the NSF Small Business Innovation Research and Small Technology Transfer Research programs and private investors to continue developing and testing the software."
Matis said that after regionals, the team expanded from developing just the machine-learning part of the product to additionally developing a user interface that responds to comments from health care providers.
"We are most looking forward to receiving instruction on how to develop the business model surrounding our product," Matis said. "In nationals, we will further develop the business-model canvas. We consider that to be invaluable information that we wouldn't know how to get otherwise."