Type 2 diabetes is viewed as one of the biggest drivers of healthcare costs partly because of the complications that can arise from the condition. Diabetes costs in 2012, for example, reached $245 billion. One complication — hypoglycemia — can cause diabetic coma if it remains untreated. WellDoc is working on a way to predict hypoglycemia without patients needing to have continuous glucose monitoring. The plan is to add the mhealth tool to its prescription mobile app platform Bluestar, which the company is preparing to roll out.
It presented positive findings from a study of the hypoglycemia prediction tool at the Diabetes Technology meeting in San Francisco this week, according to a company statement.
WellDoc tested the mhealth prediction tool using data from the International Diabetes Center and its own databases.
It used seven days of spot blood glucose monitoring data based on one blood glucose test per day. The tool alerts users of hypoglycemic events that are likely to happen within the next 24 hours. The model generated accurate predictions for hypoglycemic events in 91 percent of cases, according to WellDoc medical director Mansur Shomali, who is also an endocrinologist at MedStar Union Memorial Hospital in Baltimore.
Currently, predicting hypoglycemia requires the use of continuous glucose monitoring, a technology rarely used by or reimbursed for people with type 2 diabetes, according to the statement.
Several groups are investigating ways to harness big data to develop predictive analytics tools that identify people with a chronic condition like diabetes , even before they are diagnosed. Companies are also developing predictive analytics tools to spot early warning signs of complications in COPD patients and people with heart disease. If intervention can begin before the patient’s condition worsens, it could reduce hospitalization costs.
