A sensor-augmented insulin delivery system is a highly accurate method for self-monitoring of blood glucose (SMBG) in type 1 diabetics, researchers announced at the 67th Scientific Sessions of the American Diabetes Association (ADA).

The Paradigm REAL-Time Continuous Glucose Monitoring (RT-CGM) system, manufactured by Medtronic, provides real-time continuous readings of glucose levels with high and low glucose alarms, and allows the patient to view glucose trends in real-time and retrospectively.

John J. Mastrototaro, PhD, vice president of sensors and implantable products for Medtronic Diabetes, in Northridge, California presented the results of an open-label trial that assessed sensor accuracy and reliability in 72 type 1 diabetics using RT-CGM as an adjunct to SMBG.

“The results suggest that Medtronic Real-Time CGM may be the most accurate CGM device for home use,” Dr. Mastrototaro pointed out. “This is important because strong accuracy results are required for ultimate fingerstick replacement for CGM,” Dr. Mastrototaro observed. “Unlike SMBG, which only provides a snapshot of blood glucose at a single point in time, RT-CGM provides more of a real-world representation of actual 24-hour glucose fluctuations.”

Fingerstick replacement is the end goal for continuous glucose monitoring, Dr. Mastrototaro noted. Currently approved devices require a confirmatory fingerstick (SMBG) reading before the patient can alter his therapy. Stronger accuracy results may support a case for changing this labeling.

In their trial, the investigators analyzed over 60,000 paired sensor-meter readings.

Patients enrolled in the trial ranged from 12 to 80 with baseline hemoglobin A1C greater than 7.5% and were already using an insulin pump to treat their diabetes.

Subjects wore the sensor for three days consecutively after which it was replaced with a new sensor, and they were instructed to wear the sensor for at least six days a week during the six-month study period.

In addition to the sensor, subjects used the Paradigm 722 Insulin Pump.

Glucose values calculated from sensor signals compared to capillary blood glucose concentrations from intermittent fingerstick testing (a mean of 4 per day) and were analyzed for accuracy.

Results showed that only 1% of the 7,115 sensors failed initial start-up after the two-hour initialization period.

In 12,439 paired sensor readings in the adolescent population, the median error observed with the Paradigm REAL-Time was 12.4%. In 47,611 paired sensor readings in the adult population, the Paradigm REAL-Time median error was 10.5%.

The study was sponsored by Medtronic.

Written by: Jill Stein
Jill Stein is a Paris-based freelance medical writer.
Jillstein03 at cs.com