A new University of Virginia study on the efficacy of a smartphone-integrated artificial pancreas in children showed good signs. Mark DeBoer, MD, of UVA's Center for Diabetes Technologies, presented the work at ENDO 2017, the annual meeting of the Endocrine Society, this week in Orlando, Florida,
Artificial pancreas systems typically function on algorithms designed for use in adults, and as such can be difficult to calibrate for children. Kids aged 5 through 8 with type 1 diabetes were featured in the study, which put them (and their families) in an offseason ski resort outside of Charlottesville, Virginia.
“The onus is completely on the individual, the family in this case, or the physician recommended certain insulin doses. The way the artificial pancreas works is a much more complex process,” he said, comparing traditional treatment with the version at hand, “it takes data that is collected from a continuous glucose monitor…so it knows where the glucose is, and it where the glucose is going, whether it’s rising or falling.”
It has certain parameters for things like the basal rate and carbohydrate ratio, as well as the overall daily dose requirement. To adjust for use in children, the team set parameters in advance and created a lock screen in the integrated mobile app to prevent any meddling intervention from the young patients, which was universally effective in the study.
The study followed 12 children with a mean age of 7, equally male and female, all equipped with a FitBit and an artificial pancreas and accompanied by a parent, for 72 hours on that offseason resort. It featured structured meals and activities.
They used a system that sensed the blood sugars in the children and sent the information to the accompanying smartphone, and also to the researchers.
“We compared this to a period that either preceded this camp-type situation or followed this camp-type situation at home…to be able to compare to what things were like on the system,” DeBoer said. Families were instructed to eat their normal diet and engage their normal activity level with the children when off the study.
Based on the Fitbit’s step count, the children did exercise more during the study period than at home (there was also a mechanical bull that was a hit with them, as per DeBoer). Still, the team adjusted for these variations and found successes in the artificial pancreas.
Even given the change in activity level, adjusted, the children spent 73% of the time in the target blood sugar range on the artificial pancreas, as opposed to 47% of the time in their usual home lives without. They also spent 25.8% of their time in high blood sugar ranges on the system, as opposed to 51.5% in usual life without it. Events of low blood sugar were about equal between study time and home time.
Mean glucose level was lower: “152 versus 190, and that’s the kind of thing that’s clinically significant,” DeBoer said, “This was able to be done by giving insulin at the right time, not too much at some time, not too little at some time.”
“Artificial pancreas use in young children was safe in this setting, without dangerous lows or dangerous highs,” he concluded, “Clearly, we will need to further develop this and test artificial pancreas use in a home setting, as we’re doing right now in adults and adolescents. Artificial pancreas use resulted in lower mean blood sugars and an increase in time in the target range…certainly, further testing is necessary to see how necessary a lockout screen like this is, and whether there are artificial pancreas features that could be used for benefit in young children.”