A JournAl of the AmericAn DiAbetes AssociAtion

Clinical Diabetes/Therapeutics30-LBAccuracy and Acceptability of the 6-Day Enlite Continuous SubcutaneousGlucose SensorTIMOTHY S. BAILEY, RONALD L. BRAZG, MARK P. CHRISTIANSEN, ANDREWJ. AHMANN, ROBERT R. HENRY, SATISH GARG, ELAINE WATKINS, FRANCINER. KAUFMAN, Escondido, CA, Renton, WA, Walnut Creek, CA, Portland, OR, SanDiego, CA, Aurora, CO, Chula Vista, CA, Northridge, CAThe Enlite subcutaneous glucose sensor (Medtronic) was previouslyevaluated in adults and children and shown to be accurate for 6 days,durable, and acceptable to patients and parents/caregivers. A pivotal 6-daytrial in adults was conducted at 7 US investigational centers. Sensors wereself-inserted and taped, and were calibrated 3-4 times/day. Accuracy wasevaluated vs frequently-sampled YSI plasma glucose values. Hypoglycemiaand hyperglycemia were induced on days 1 (immediately after initialcalibration), day 3, and day 6. Patient satisfaction with Enlite was alsoevaluated with a 7-point Likert-type questionnaire. Adults with type 1(N=65) or type 2 (N=25) diabetes (mean age 44, range 18-71) wore either1 or 2 sensors on their abdomens for 6 days. The Table shows mean andmedian absolute relative differences (ARD) between sensor and YSI glucosevalues by day and by glucose range. Mean responses of patient satisfactionsurveys were 5.9/7 for “ease of use,” 6/7 for “comfort,” 5.9/7 for “ease ofinsertion,” and 5.8/7 for “would recommend.” There were no device-relatedadverse events. We conclude that the Enlite sensor is accurate, durable,comfortable, safe, and easy to use. Improvements in continuous glucosesensing should expedite development of semi-automated insulin deliveryfeatures in modern pumps.Table. Enlite sensor accuracy by mean and median absolute relative differencesBy Day Overall Day 1 Day 3 Day 6Mean (Median) ARD 13.6% (10.1%) 15.9% (12.2%) 11.8% (9.1%) 13.2% (9.6%)By Glucose Range Overall 75-180 mg/dL >180 mg/dLMean (Median) ARD 13.6% (10.1%) 10.8* (8.5)* 12.7% (9.4%) 12.0% (9.0%)* mean absolute differencesSupported by: Medtronic, Inc.31-LB3 Continuous Glucose Monitoring Systems: A Comparison with 6Sensors per Subject in ParallelGUIDO FRECKMANN, MANUELA LINK, STEFAN PLEUS, EVA ZSCHORNACK,CORNELIA HAUG, HANS-MARTIN KLÖTZER, ELISABETH RAMSTETTER, WIL-FRIED SCHMIDT, MICHAEL SCHOEMAKER, Ulm, Germany, Weinheim, Germany,Mannheim, GermanyThis study aimed to compare 3 commercially available needle-typecontinuous glucose monitoring (CGM) systems under daily life conditions.Twelve people with type 1 diabetes (age: 47.7 ± 6.9 years (mean ±standard deviation (SD)), HbA1c 8.2 ± 1.4 %, duration of diabetes: 23.7 ± 9.9years) wore 6 sensors in parallel, two sensors each of the three followingCGM systems: Abbott Diabetes Care FreeStyle Navigator , DexCom Seven ® Plus 3rd Generation and Medtronic MiniMed Guardian ® REAL-Timewith Enlite Sensor. Each system was used for one sensor lifetime, whichwas 5 days for the FreeStyle Navigator , 6 days for the Guardian ® REAL-Time or 7 days for the Seven ® Plus. CGM measurements were paired tocapillary blood glucose (bG) measurements (approx. 30 per day and subject).Performance was characterized by mean absolute relative differences(MARD) between CGM and paired bG measurements and percentageabsolute relative difference (pARD), calculated from the absolute relativedifferences between paired CGM measurements of the two sensors of eachsystem.Capillary bG measurements ranged from 35 to 446 mg/dL. MARD andpARD are reported in the table (mean ± SD [range]).All systems could be worn for nearly 100 % of their estimated lifetime(FreeStyle Navigator : 100.6 %, Seven ® Plus: 98.3 % and Guardian ® REAL-Time: 99.0 %) before the sensors had to be removed.The 3 CGM system tested showed good overall performance, with theFreeStyle Navigator achieving lower sensor-to-bG differences and sensorto-sensordifferences than the DexCom Seven ® Plus 3rd generation andthe Medtronic MiniMed Guardian ® REAL-Time with Enlite Sensor.ADA-Funded ResearchCGM systemCGM system characteristicsMARD(in %, n = 24)FreeStyle Navigator 12.4 ± 3.6[8.4 - 19.0]Seven ® Plus 16.7 ± 3.8[12.7 - 29.2]Guardian ® REAL-Time 16.4 ± 6.9[7.5 - 23.9]Supported by: Roche Diagnostics GmbH, GermanypARD(in %, n = 12)10.1 ± 4.1[5.2 - 17.6]15.4 ± 4.2[9.2 - 23.5]18.1 ± 6.5[9.2 - 31.2]Average datareportingpercentage(in %)97.3 ± 5.295.9 ± 4.698.9 ± 0.832-LBRandomized Outpatient In-Home Pilot Trial of Predictive NocturnalPump Shut-OffBRUCE A. BUCKINGHAM, FRASER CAMERON, JOHN LUM, DAVID MAAHS,PAULA CLINTON, BREANNE HARRIS, JAMIE REALSON, WAYNE B. BEQUETTE,DARRELL M. WILSON, ROY W. BECK, PETER H. CHASE, Stanford, CA, Troy, NY,Tampa, FL, Aurora, COWe have developed a nocturnal hypoglycemia prediction algorithm tosuspend insulin pump delivery to avoid hypoglycemia that utilizes Kalmanfiltered glucose sensor data. This study was designed to test how the systemand algorithm functions with home use and is among the first studies in theUS of in home partial closed-loop control.Each subject wirelessly linked a Medtronic Paradigm ® insulin pump andSof-Sensor ® to a small bedside laptop computer nightly. The algorithmstops insulin infusion when the sensor glucose is predicted to become low.Once the glucose is past the nadir, basal insulin is restarted. The maximumsingle suspension time is 120 minutes, and the maximum suspension timeeach night is 180 minutes. Once the sensor glucose is past the nadir, basalinsulin is restarted. A randomization occurs each night on initialization ofthe system, so that the algorithm is active on 67% of the nights and 33% ofnights are control nights (algorithym off).Five subjects (ages 18-32 yrs.; A1c’s 6.4-7.7%) were studied for 101nights with the algorithm active on 65 nights. When active a pump shutoffoccurred on 77% of the nights with median a shutoff duration of 60minutes.Glycemic Changes with Nocturnal Pump SuspensionAlgorithm Status Algorithm Off Algorithm OnMean Bedtime Glucose 167 +/- 53 mg/dl 154 +/- 47 mg/dlMean Fasting Glcuose 123 +/- 48 mg/dl 157 +/- 52 mg/dl% Fasting