Better insight, analyses needed to overcome diabetes ‘data smog’
SAN DIEGO — Patients who want to better manage their diabetes with the latest “smart” devices often find themselves overwhelmed by large amounts of data. Many of the numbers are never analyzed in a meaningful way that can improve outcomes, according to a speaker at the American Association of Diabetes Educators annual meeting.
“There are 25 billion blood glucose data points collected annually ... and that’s not including information from continuous glucose monitors, food information, medication information and HbA1c,” Erik Otto, BSc, president and founder of InSpark Technologies, said during a presentation. “Most tools present all this data back to the [user] with little or no added insight. Most people, whether a patient or clinician, simply don’t have time to analyze this data. Most people, also, don’t have the clinical acumen to make sense of this data.”
The result, Otto said, is what he calls “diabetes smog” — patients struggling to understand how to best manage their disease with only a confusing and seemingly endless stream of numbers.
Several studies show that few patients and caregivers are retrospectively reviewing their data, Otto said. According to a recent study of adults and caregivers of children with type 1 diabetes, published in Diabetes Technology & Therapeutics, just 31% of adults and 56% of caregivers reported ever downloading data from one or more devices, whereas 12% of adults and 27% of caregivers were considered routine reviewers of data, reviewing their data most of the time they downloaded it. Both routine reviewers and the children of routine reviewers had lower mean HbA1c levels when compared with participants who did not review their data.
“And these are type 1 patients who are very intensively managed,” said Otto, who was diagnosed with type 1 diabetes 20 years ago. “These numbers are likely to be much lower in other populations. People are just not using the software, and there has to be a reason for that. I think it’s because they’re not providing enough utility.”
There are two big hurdles that keep patients from making sense of their diabetes data, Otto said: data from various devices stored in different places (access) and effective analytical tools are lacking to make sense of the information.
“Many of the analyses used today are basic statistics that are not particularly actionable or insightful for the management of diabetes,” Otto told Endocrine Today before the presentation.
To address the problem, software tools designed to analyze diabetes data need to accomplish several things, Otto said. The tools should be “glanceable,” meaning any insight could be gleaned in just a few seconds, similar to information gleaned from a weather app. Apps should push-based, providing alerts for patients about diabetes management insights at the right time, or control systems, actively managing the person’s diabetes for them.
“We really need to do better to provide some kind of an incentive to collect this data,” Otto said. “I really think the issue is not so much the connectivity ... it’s that [patients] are not getting the benefit if they do collect the data.”
New technology coming onto the market provides hope, Otto said, with “smog-busting enablers” like pattern recognition, predictive algorithms, machine learning and control systems engineering pioneering new ways to harness diabetes device data in a useful way. Still, he said, much more needs to be done.
“Any insights provided should be predictive, actionable, timely and unobtrusive into the life of the user,” Otto told Endocrine Today. “I also hope that there is more attention to mobile health solutions that are scientifically vetted, and that these solutions are properly valued and reimbursed in the marketplace.” – by Regina Schaffer
Reference: Otto E. F08. Help Clients Overcome Diabetes Data Smog: Perspectives from a CEO with Diabetes. Presented at: AADE 2016; Aug. 12-15, 2016; San Diego.
Disclosure: Otto is an employee and equity holder in InSpark Technologies.