Maryland-based health technology company WellDoc says it has developed a mathematical model that eventually will allow people with type 2 diabetes to predict the onset of a hypoglycemic episode.
According to WellDoc, its algorithms can take “sparse data”–namely as few as seven days’ worth of blood glucose monitoring data at the rate of one blood glucose test per day–and predict the likelihood of a hypoglycemic event within the next 24 hours. The company says that in initial testing of the model, it correctly predicted hypoglycemic events using sparse data 91 percent of the time.
The company adds that the model correctly predicted hypoglycemic incidents 42 percent more often than expert endocrinologists, with low rates of false positives or negatives.
WellDoc researchers tested the technology using tens of thousands of blood glucose data points gathered from both the International Diabetes Center and WellDoc’s proprietary databases. The research showed that by using about one test per day, the hypoglycemia prediction model worked any day of the week and across patient populations regardless of their medication regimen.
WellDoc says that after more development and testing, it plans to add the prediction technology to its BlueStarTM mobile prescription therapy device, including the capability to advise users in real time their best options for avoiding an impending hypoglycemic episode.
The company’s plan would dramatically change how type 2s anticipate and deal with hypoglycemia. The best current technology, continuous glucose monitors, are expensive and not often used by type 2s. Typically, type 2s test their blood glucose levels once a day, which can create a huge data gap when it comes to spotting a trend toward hypoglycemia.
The WellDoc algorithms’ ability to accurately predict hypoglycemic events based on “sparse data” would add a powerful new tool to type 2s’ diabetes management kit.
WellDoc is presenting the initial results of its hypoglycemia prediction model today at the 2013 Diabetes Technology Society Meeting in San Francisco.