COMPLICATIONS—HYPOGLYCEMIA COMPLICATIONS—HYPOGLYCEMIA ± 10 pg/ml; P < 0.01), and endogenous glucose production improved from baseline to 6 and 18 months (0.42 ± 0.08 to 0.47 ± 0.04 and 0.87 ± 0.20 mg/ kg/min; P < 0.05), albeit remaining less than normal (1.39 ± 0.11 mg/kg/min; Moderated Poster Discussion: Risks for and Consequences of Hypogly- P < 0.05); there was no improvement in epinephrine or the autonomic symp- cemia (Posters: 382-P to 387-P), see page 21. tom score. Real-time CGM can reduce the burden of problematic hypogly- cemia in patients with long standing T1D complicated by hypoglycemia & 382-P unawareness, which may in part be related to improvement in the endog- enous glucose production response to insulin-induced hypoglycemia that is Critical Hypoglycemia in Hospitalized Patients with Diabetes: Early POSTERS required to prevent or correct low blood glucose. Complications Outcomes of Standardized Reporting and Management Supported By: National Institutes of Health (R01 DK091331, UL1 TR000003, P30 Acute and Chronic KATHERINE A. ARAQUE, DEEPAK KADAYAKKARA, NINO GIGAURI, KASIA J. LIP- DK19525, T32 DK007314) SKA, GREGORY BULLER, CLARE A. FLANNERY, Bridgeport, CT, New Haven, CT Critical hypoglycemia (glucose <50 mg/dl) is associated with increased morbidity and mortality in hospitalized patients. We performed a root cause & 384-P analysis of critical hypoglycemia events (HGE) on two inpatient medical Risk of Cardiac Arrhythmias during Spontaneous Hypoglycemia in wards with the highest incidence of HGE. Pre-intervention, management Young People with Type 1 Diabetes and reporting of HGE were inconsistent, and time from HGE to documented PETER NOVODVORSKY, ALAN BERNJAK, ELAINE CHOW, AHMED IQBAL, LIANNE euglycemia (>100 mg/dl) was over 3 hours. To reduce HGE incidence and SELLORS, SCOTT WILLIAMS, ROBERT A. FAWDRY, RICHARD JACQUES, JEFFER- improve the time from HGE to euglycemia, we initiated several system- SON L.B. MARQUES, PAUL J. SHERIDAN, SIMON R. HELLER, Sheffi eld, United based changes. The EPIC EMR was programmed to provide a standardized Kingdom, Florianópolis, Brazil order set for basal-bolus insulin regimen, automated dextrose order for all Hypoglycemia has been strongly implicated in sudden nocturnal death in HGE (gel or IV), automated MD notifi cation, and recommendation for Endo- type 1 diabetes (T1DM) but the precise mechanism remains unclear. This crine consultation after 2 critical HGE. The Pyxis Med Station was pro- observational study evaluated spontaneous hypoglycemic episodes in 37 grammed to alert nurses to recheck blood glucose 15 minutes after dextrose. participants with T1DM (median age 34 years, 25th/75th centile 25.5 - 40.5) A card with the HGE management protocol was attached to each provider’s who underwent 4 days of simultaneous 12-lead ECG Holter and blinded con- ID badge, and educational seminars were given to all providers on the 2 tinuous interstitial glucose (IG) monitoring (CGM) while recording symptom- inpatient wards. Pre- and post-intervention events (Table 1) were analyzed atic hypoglycemic episodes. Altogether 3165 hours of CGM recordings were by 2-tailed Mann Whitney U and Chi-square tests. A marked improvement in analysed. Thirty-two participants (86.5%) experienced at least one episode the incidence of HGE and time to recovery demonstrate the positive impact of hypoglycemia (IG ɖ 3.5mmol/L, duration ɗ 20 min) and 23 participants of these system-based interventions. These interventions were expanded (62.2%) experienced at least one episode of nocturnal hypoglycemia (23.00- to the entire hospital and can be potentially applied to other hospitals to 7.00). Mean duration of nocturnal hypoglycemia (105 min ± 110 min) was lon- improve patient safety. ger compared to mean duration of daytime hypoglycaemia (59 min ± 48 min) (p = 0.011). Only 27% of nocturnal and 57% of daytime hypoglycemias were Table 1. Outcomes of HGE Management Pre- and Post Interventions. reported as symptomatic. Cardioacceleration was detected during daytime Pre-Intervention Post-Intervention p-value hypoglycemia (mean heart rate (HR) 80 ± 14bpm) vs. matched euglycemia (duration 3 months) (duration 6 months) (mean HR 77 ± 12bpm) (p = 0.021). Mean QTc was prolonged during nocturnal Co-Primary Outcome: Time from HGE to 225 ± 46 min 87 ± 26 min 0.03 hypoglycemia (405 ± 27msec) vs. euglycemia (400 ± 22msec) (p = 0.03) as euglycemia (median +/- SD) well as during daytime hypoglycemia (413 ± 29 msec) vs. euglycemia (401 Co-Primary outcome: Time from HGE to 76 ± 14 min 28 ± 10 min <0.001 ± 29 msec) (p < 0.001). We detected a higher incidence rate of bradycardia next fi ngerstick (median +/- SD) during nocturnal hypoglycemia compared to euglycemia (incidence rate ratio Secondary Outcome: Incidence of HGE (number 11 per 100 admissions 5.6 per 100 admissions <0.001 (IRR) 9.57 (95% CI 9.39 - 9.75), p < 0.001). During daytime hypoglycemia, of HGE per total ward admissions with DM) (35/320) (37/653) atrial ectopic activity was higher (IRR 2.50 (95% CI 1.35 - 4.66), p = 0.004), Recurrent HGE (glucose<50 mg/dl more than 37% (13/35) 24% (9/37) 0.09 but incidence of bradycardia was lower when compared to euglycemia (IRR once on a given patient) 0.025 (95% CI 0.002 - 0.26), p = 0.002). Our data suggest differences in car- diac electrophysiological responses during daytime vs. nocturnal hypogly- Documented Treatment of HGE 63% (22/35) 97% (36/37) <0.001 cemia. The increased risk of bradycardia during nocturnal hypoglycemia Physician Notifi cation 51.4% (18/35) 78% (29/37) <0.001 in susceptible individuals is clinically relevant and potentially amenable to therapeutic intervention. Supported By: Sheffi eld Teaching Hospitals & 383-P Effect of Real-Time Continuous Glucose Monitoring on Glucose & 385-P Counterregulation in Long-Standing Type 1 Diabetes: Preliminary Baseline Corrected QT Interval Is Associated with Prolongation of Results QTc during Severe Hypoglycemia in Patients with Type 2 Diabetes MICHAEL R. RICKELS, AMY PELECKIS, CORNELIA DALTON-BAKES, SHANNON SEON AH CHA, JAE-SEUNG YUN, HYUK-SANG KWON, JUNG-MIN LEE, YU-BAE O’BRIEN, NINA RAN, JOSEPH R. NAJI, HUONG-LAN NGUYEN, SANJIAN CHEN, AHN, SEUNG-HYUN KO, Suwon, Republic of Korea, Seoul, Republic of Korea INSUP LEE, MARK H. SCHUTTA, Philadelphia, PA We investigated an association between baseline heart rate−corrected Patients with long standing type 1 diabetes (T1D) are at increased risk QT (QTc) interval before severe hypoglycemia (SH) and prolongation of QTc for severe hypoglycemia due to defects in glucose counterregulation and interval during SH in patients with type 2 diabetes. hypoglycemia symptom recognition, in part mediated through exposure to Between January 2004 and June 2014, 208 patients with type 2 diabe- hypoglycemia. We sought to determine whether implementation of real- tes, who visited the emergency department for SH and conducted standard time continuous glucose monitoring (CGM) as a strategy for hypoglycemia 12−lead electrocardiography (ECG) within 6 months before SH (baseline avoidance could restore glucose counterregulation in response to insulin- ECG), were consecutively enrolled by retrospective medical record review. induced hypoglycemia in patients with long standing T1D complicated by The QT interval was corrected using Bazett’s formula (QTc), and the QTc hypoglycemia unawareness. Eleven subjects (age 44 ± 4, T1D duration 31 ± interval was analyzed during SH and 6 months before and after SH. QTc 4 years) completed paired hyperinsulinemic (1.0 mU/kg/min) euglycemic (90 intervals of 450 ms or longer in men and 460 ms or longer in women were mg/dl for 4 hours) and hypoglycemic (80, 65, 55 and 45 mg/dl each hour) considered abnormally prolonged. clamps at baseline and 6 months after initiation of CGM, with 9 subjects to The mean age and diabetes duration were 68.1 ± 12.1 and 14.1 ± 10.1 date having completed clamps after 18 months of CGM, and results com- years, respectively. The mean QTc intervals at baseline and SH episodes pared to a nondiabetic control group (n = 12). The T1D subjects experienced were 433 ± 33 ms and 460 ± 33 ms, respectively (P < 0.001). One hundred and improvement in hypoglycemia awareness (Clarke score 6.0 ± 0.3 to 4.6 ± 0.3 fourteen patients (54.8%) had a prolonged QTc interval during SH. There was and 2.9 ± 0.7; P < 0.01) and severity (HYPO score 1,884 ± 295 to 715 ± 135 and a signifi cant decrease in the prolonged QTc interval within 6 months after SH 354 ± 64; P < 0.01) without change in HbA1c (7.2 ± 0.2 to 7.0 ± 0.2 and 7.1 ± (QTc interval prolongation during SH vs. after recovery, 54.8% vs. 33.8%, P 0.3%). During the fi nal hour of clamped hypoglycemia, there was a trend for < 0.001). The prolonged QTc interval was signifi cantly associated with base- improvement in glucagon from baseline to 6 and 18 months post-interven- line QTc interval prolongation (odds ratio [OR] 2.92; 95% confi dence interval tion (41 ± 5 to 44 ± 5 and 50 ± 10 pg/ml; P = 0.08) still less than normal (96 [CI] 1.22−6.96; P = 0.016) after adjusting for age, sex, diabetes duration, body ADA-Supported Research & Moderated Poster Discussion For author disclosure information, see page A696. A101 COMPLICATIONS—HYPOGLYCEMIA mass index, smoking, presence of hypertension and chronic kidney disease, Table 1. Patient Characteristics and Economic Outcomes by Hypoglycemia previous SH, and use of insulin or sulfonylurea. Severity among Patients with T2DM Using Basal Insulin. In conclusion, a prolonged QTc interval at baseline was signifi cantly associated with prolongation of the QTc interval during SH in patients with type 2 diabetes. Therefore, this study suggests that we should pay attention to baseline QTc interval in patients with type 2 diabetes to avoid SH.