Abstracts / Can J Diabetes 36 (2012) S24eS76
Trial Name
Lead Institution City Investigator (PI)
Artificial Pancreas Projectrelated trials CGM-based semi-automated Perkins basal assessment management pilot Nocturnal hypoglycemia Hramiak pump shut-off RCT of Sensor-Augmented Lawson Pump vs Standard Pump Initiation (CGM TIME) CGM in T1D pregnancy Feig (CONCEPTT) Trials in Diabetes Complications
Additional sites
Table Accuracy of Enlite and Sof-sensor subcutaneous glucose sensors at different glucose concentration ranges. N ¼ the number of paired sensor/YSI data points. Lower numbers indicate better accuracy. Reference Range
UHN
Toronto
Western/ Lawson CHEO
London
Mt Sinai Hospital
Toronto 7 affiliate sites
Ottawa
40-80 mg/dL N Mean (SD) Median >80-120 mg/dL N Mean (SD) Median >120-240 mg/dL N Mean (SD) Median >240-400 mg/dL N Mean (SD) Median All 40-400 mg/dL N Mean (SD) Median
4 affiliate sites
Health Research Institute, the Hospital for Sick Children, and the Children's Hospital of Eastern Ontario, University of Ottawa, serve as regional hubs for Clinical Partner sites and Affiliate Sites. A total of 17 institutions in Southern Ontario are participating in CCTN activities. CCTN is implementing nine clinical trials and two technology development projects: 164 Administrative Coding of Type 1 and Type 2 Diabetes: Assessment of Validity SIMONA C. BURS*, DOREEN RABI, WILLIAM GHALI, ALUN EDWARDS, PETER SARGIOUS Calgary, AB Background: Accurate diagnostic coding of diabetes diagnosis is critical in health services research using administrative data. Type 1 (T1DM) and type 2 diabetes (T2DM) are not always easily clinically distinguished though do have distinct ICD-10 diagnostic codes. It is not known whether diagnostic coding in practice accurately identifies the type of diabetes. Purpose: to assess the quality of administrative coding of diabetes among patients discharged from hospital. Methods: All patients admitted to hospital with myocardial infarction and diabetes in 2006 were identified using the APPROACH database. A chart review of identified patients was conducted. Clinical data reflecting the presence of T1DM vs. T2DM was extracted (age at diagnosis of diabetes, BMI, mode of treatment at admission) along with data provoking uncertainty (e.g. use of different diabetes classification on the same chart). If all discriminatory clinical data was present, the coded diagnosis was considered “certain”; if incomplete - the coded diagnosis was considered “uncertain”. Results: 141 charts met selection criteria, 38 (27%) were considered to have an “uncertain” diagnosis. Of the uncertain cases, 47.4% had insufficient documentation of clinical data to ascribe with any certainty whether these individuals had T1DM or T2DM. Inconsistent nomenclature (31.6%) and inconsistent diabetes classification (19.4%) were also more common among the “uncertain” cases. Conclusions: This study identifies significant concerns with the quality of diabetes coding, largely stemming from physician documentation. Standardized diabetes data forms on hospital charts might mitigate the uncertainty around phenotypic definition, for more accurate diagnostic discrimination and outcome evaluation. 165 Comparison of Sensor Performance with the Enlite Sensor versus Sof-sensor in Adults with Diabetes JOHN J. SHIN*, SUIYING HUANG, SCOTT W. LEE, JOHN B. WELSH, BOB JANOWSKI, FRANCINE R. KAUFMAN Northridge, CA The Enlite subcutaneous glucose sensor was designed to improve upon the accuracy of the predicate Sof-Sensor released by
S49
Enlite
Sof-sensor
633 14.0 (11.2) 11.8
342 21.1 (17.0) 18.1
1355 13.2 (13.4) 9.9
762 20.0 (17.8) 15.2
2869 12.5 (12.1) 9.1
2268 13.9 (13.3) 9.9
1000 13.2 (11.1) 10.2
437 14.1 (14.2) 9.1
5857 13.7 (13.2) 10
3809 16.9 (17.2) 11.8
*
For the 40-80 mg/dL range, Mean (SD) and Median are expressed as mean absolute difference in mg/dL. For other glucose ranges, Mean (SD) and Median are expressed as ARD in %.
Medtronic MiniMed first in 1999. Both Enlite and Sof-Sensor were previously studied in adults using the Guardian REAL-Time glucose monitor. Analysis of sensor accuracy was performed pairing sensor glucose (SG) and plasma glucose [Yellow Springs Instrument (YSI)] values taken from adult patients with type 1 and type 2 diabetes. The Enlite study was a prospective single-sample correlationaldesign study conducted in 99 subjects (31% female, mean age 42.2 years, 18% type 2 diabetes). Enlite sensors (2 per subject) were placed in the abdomen and buttocks for 6 days, subjects calibrated sensors 4 times per day, and participated in a single-day 10 hour frequent (q15 min) blood sampling protocol to compare SG and YSI glucose values. The Sof-sensor study was a single-sample correlational design study in 16 subjects (mean age 39.6, 37.5% female, 100% type 1 diabetes); subjects wore 2 sensors in the abdomen for 3 days as inpatients. One sensor was calibrated 2 times per day and the other sensor was calibrated 4 times per day. Subjects had blood samples drawn every 30 min to compare SG and YSI glucose levels. Both the Sof-sensor and Enlite demonstrate accurate sensor performance. The Enlite sensor demonstrated improved sensor accuracy at all glucose ranges (Table). Its use should help to expedite development of semi-automated insulin delivery for people with diabetes.
166 Impact of Cyanidin-3-Glucoside on Glycated LDL-induced Oxidative Stress, Mitochondrial Dysfunction and NADPH Oxidase in Cultured Vascular Endothelial Cells XUEPING XIE, RUOZHI ZHAO, GARRY SHEN Winnipeg, MB Increased levels of glycated low density lipoprotein (glyLDL) are commonly elevated in diabetic patients. Our previous studies demonstrated that glyLDL increased the production of reactive oxygen species (ROS), activated NADPH oxidase (NOX) and suppressed mitochondrial electron transport chain (mETC) enzyme activities in vascular endothelial cells (EC). Cyanidin-3-glucoside (C3G), a type of anthocyanin abundant in dark berries or red wine, reduced oxidative stress and insulin resistance in diabetic animals. The present study examined the effect of C3G on glyLDL-induced ROS production, NOX activation and mETC enzyme activity in porcine aortic EC (PAEC). Co-treatment of C3G inhibited glyLDLinduced ROS increase in EC. GlyLDL increased NOX activity in EC and