MONITOR
Handheld Meter Features:
Displays minute by minute readings
Rate arrow displays glucose trend
User-defined high and low alarms
Projected alarms gives advance notice of glucose excursions
On-board FreeStyle Meter for calibration eliminates transcription errors
GLUCOSE SENSER
Amperometric 3-electrode based on Wired Enzyme Technology™
3 Day Lifetime
Readings every 15 minutes
TRANSMITTER/SENSER
Wireless Transmission
Meter Information Display
On board Glucose Data with Memory
Poster: [392-P] The TheraSense, Inc. Continuous Glucose Monitor: Preliminary
Clinical Results from a Subcutaneous Sensor with a Wireless Connection
to a Hand-Held Display/Alarm
BRIAN CHO, BENJAMIN FELDMAN, PHU LE, FEI MAO, CHRISTOPHER THOMAS. Alameda,
CA.
Sunday, June 15, 2003, 6:00 PM, Poster Presentation: Clinical Therapeutics/New
Technology - Glucose Monitoring and Sensing (6:00 PM - 7:00 PM)
The TheraSense continuous glucose monitor consists of a subcutaneous
electrochemical sensor utilizing Wired Enzyme™ technology, implanted
for three days, and wirelessly coupled to a hand-held display/alarm.
The subcutaneous sensor demonstrates good linearity over an extended
glucose range, as well as excellent stability (drift of ca. 0.1%/hour)
under simulated and actual in-vivo conditions.
User studies with this system are currently underway. These trials
are an extended home-use evaluation, with subjects wearing a series
of consecutive 3-day sensors over a 21-day period. Sensor placement
is alternated between arm and abdomen.
Sensors are self-inserted, and the system is self-calibrated via arm-stick
blood glucose measurements, using an onboard (integrated into the hand-held
display/alarm) FreeStyle® blood glucose meter.
Real-time glucose values (designed to be available at a one-minute
frequency) are masked for the clinical trial, but daily alarms for both
hypo- and hyperglycemia are provided. Performance is evaluated through
the use of frequent finger-stick blood glucose measurements, using the
onboard blood glucose meter.
Results from the first 17 subjects (approximately 350 sensor days)
have been studied using a pseudo-Clarke error grid analysis (with finger-stick
blood glucose values as the reference). This analysis gave 74.2% of
readings in Clarke region A, and 97.4% in either Clarke region A or
B. Topics of discussion will include statistical accuracy, analysis
of alarm efficacy, and the clinical experience of trial subjects.
