Worldwide, more than 422 million adults are diabetic, resulting in a death from the disease every six seconds. At least $825 billion is spent annually on diabetes-related costs globally, including treatment, complications, hospitalizations, and lost productivity. In the U.S., more than one-third of all adults have prediabetes. The diabetes and prediabetes epidemics are crying out for early intervention, prevention, and treatment.
This means every opportunity that enables care managers to identity prediabetics at high risk for progressing to diabetes is critical, especially within a time period where intervention could make a difference by delaying or halting progression of the disease. Medial EarlySign’s AI-based machine learning research focuses on precisely addressing that problem by flagging prediabetics at high risk for progressing to diabetes with a high accuracy rate, with as few false positives as possible. This risk stratification helps healthcare organizations allocate their diabetes resources to the right patients at the right time.
Medial EarlySign is developing a suite of machine learning models for diabetes designed to aid healthcare providers in identifying prediabetic patients at high risk for becoming diabetic within a designated timeframe, and for developing diabetes-related complications. By uncovering “hidden” signals within existing medical data, Medial EarlySign’s goal is to create machine learning-powered prediabetes risk predictors to identify the likelihood of which prediabetics will become diabetic within one year when opportunities to intervene can be more likely to deliver better outcomes.
EarlySign’s models are built with our clinically supported proprietary machine learning AI technology. These clinical risk predictors are designed to provide care managers at healthcare organizations opportunities to better allocate their diabetes-focused resources and plan for their patient care needs accordingly.
In a retrospective clinical data study involving 645,000 UK patients, Medial EarlySign used its artificial intelligence and machine learning-based diabetes algorithms to research EHR data, and identify which prediabetic patients were at high risk for becoming diabetic within 12 months. The results of this study substantiate the ability of our prediabetes to diabetes algorithm to enhance health organizations’ ability to proactively flag prediabetics at high risk, deliver timely care management for them, and potentially prevent or delay disease progression.
For more details and information about the study, contact Medial EarlySign.