1. A Comparison of the Original vs. Modified Continuous Glucose Monitoring System (CGMS™) Sensor During Hypoglycemia in the Diabetes Research in Children Network (DirecNet) Accuracy Study Michael Tansey 1, Roy Beck 2, Bruce Buckingham 3, Nelly Mauras 4, Rosanna Fiallo-Scharer 5, Dongyuan Xing 2, William Tamborlane 6, Katrina Ruedy 2, and the Diabetes Research in Children Network (DirecNet) Study Group. 1 Department of Pediatrics, University of Iowa, Carver College of Medicine, Iowa City, IA; 2 Jaeb Center for Health Research, Tampa, FL; 3 Division of Pediatric Endocrinology and Diabetes, Stanford University, Stanford, CA; 4 Nemours Children’s Clinic, Jacksonville, FL; 5 Barbara Davis Center for Childhood Diabetes, University of Colorado, Denver, CO; 6 Department of Pediatrics, Yale University School of Medicine, New Haven, CT.

2. The CGMS has been used to follow glucose trends in patients with type1 diabetes, however the accuracy of this device to document hypoglycemia has not been completely evaluated. Recently, the manufacturer of CGMS(Medtronic MiniMed, Northridge, CA) developed a modified glucose sensor using a new fabrication process. To study the accuracy of these sensors, the DirecNet study group examined 91 patients, 3-17 years of age with type 1 diabetes. CGMS glucose values were compared with simultaneous reference venous glucose values, sampled every 30-60 minutes for 24 hours and every 5 minutes for up to 90 minutes during an insulin-induced hypoglycemia test. When all sensor values were analyzed, the 25 CGMSs using modified sensors were substantially more accurate than the 115 CGMSs using original sensors in median error (11% versus 19%, P 100 mg/dL. Only 8 of 26 overnight hypoglycemic episodes detected by original sensors were RG confirmed, whereas all 3 detected by modified sensors were RG confirmed. Conclusion: Recent modifications to the CGMS sensor have resulted in substantially better accuracy. However, even the modified sensors differ substantially from RG glucose values in the hypoglycemic range. ABSTRACT

3. Background Continuous Glucose Monitoring System (CGMS™) was the first glucose sensor to be approved by the FDA CGMS is capable of measuring interstitial fluid glucose every 5 minutes for up to 72 hours In November, 2002 Medtronic MiniMed changed the manufacturing process of the sensor that had been in place since 1999 by modifying the sensor fabrication process To evaluate the accuracy of the original and modified CGMS sensors in children and adolescents with type 1 diabetes mellitus, by comparing sensor values to reference glucose levels Objective

4. Methods Data pooled from three studies: Inpatient Accuracy Study –91 patients admitted to a CRC for 24-26 hours to evaluate accuracy of the CGMS –One or two CGMS sensors inserted in each patient –Blood samples collected every 60 min during the day, every 30 min overnight, and every 5 minutes for up to 90 minutes during an IV insulin induced hypoglycemia test Outpatient Pilot Study –15 patient pilot study designed to evaluate the feasibility of electronic data collection from a computer used in the patients’ homes –CGMS used by 10 patients prior to 3-month visit for 48-72 hours –On 2 days of CGMS use, 8-point blood glucose testing performed with Ultra Outpatient Randomized Trial (RCT) –Ongoing study of 200 patients –All patients provided with home computer and UltraSmart meter –CGMS used by all patients for 48-72 hours at Baseline, 3 and 6 months –On 2 days of CGMS use, 8-point blood glucose testing performed with UltraSmart –Data available from Baseline on 103 patients

5. Demographics

6. Data Analysis CGMS glucose measurements were matched to reference measurements within 2.5 min of each other –Inpatient Study reference values: IV blood samples measured in a central laboratory –Pilot Study and RCT reference values: home glucose meter CGMS time was adjusted by 2.5 min to account for averaging of glucose levels made over the previous 5 min When comparing CGMS monitors simultaneously, sensor values were matched within 2.5 min of each other

7. Sensory Accuracy for Reference Glucose Values <60 mg/dL * Absolute value of sensor glucose minus reference glucose.

8. Sensor Accuracy Inpatient Study Pilot and RCT Studies Original Sensors N=4643 Paired Values Modified Sensors N=999 Paired Values Modified Sensors N=1262 Paired Values Relative Absolute Difference Median (25 th, 75 th percentiles) 19% (9%, 34%) 11% (5%, 23%) 12% (5%, 23%) ISO Criteria Met According to Reference Glucose Level ≤70 mg/dL37%44%67% 71-120 mg/dL44%64%65% 121-180 mg/dL52%71% 181-240 mg/dL60%77%75% >240 mg/dL59%80%77% All Values54%72%

9. Effect of Glucose Concentration on Accuracy Assessed by Median RAD% Original CGMS (Inpatient) Modified CGMS (Inpatient) Modified CGMS (Pilot / RCT)

10. Conclusions The modified CGMS sensor offers improved accuracy compared to the original sensor at all glucose ranges that were measured There is decreased accuracy of both sensors in the hypoglycemic ranges when compared to the reference glucose

11. Barbara Davis Center –H. Peter Chase –Rosanna Fiallo-Scharer –Jennifer Fisher –Barbara Tallant University of Iowa –Eva Tsalikian –Michael Tansey –Linda Larson –Julie Coffey –Amy Sheehan Nemours Children’s Clinic –Tim Wysocki –Nelly Mauras –Keisha Bird –Kelly Lofton Stanford University –Bruce Buckingham –Darrell Wilson –Jennifer Block –Paula Clinton Yale University –William Tamborlane –Stuart Weinzimer –Elizabeth Boland –Kristen Sikes –Amy Steffen Jaeb Center for Health Research –Roy Beck –Katrina Ruedy –Craig Kollman –Dongyuan Xing –Cynthia Silvester