BG Level Research Articles

Continuous glucose monitoring using machine learning models and IoT device data: A meta-analysis.

Machine learning offers diverse options for effectively managing blood glucose levels in diabetes patients. Selecting the right ML algorithm is critical given the array of available choices. Integrating data from IoT devices presents promising opportunities to enhance real-time blood glucose management models. This meta-analysis aims to evaluate the effectiveness of machine learning models utilizing IoT device data for predicting blood glucose levels. We systematically searched electronic databases for studies published between 2019 and 2023. We excluded studies lacking ML model derivation or performance metrics. The Quality Assessment of Diagnostic Accuracy Studies tool assessed study quality. Our primary outcomes compared ML models for BG level prediction across different prediction horizons (PHs). We analyzed ten eligible studies across prediction horizons of 15, 30, 45, and 60 minutes. ML models exhibited mean absolute RMSE values of 15.02 (SD 1.45), 21.488 (SD 2.92), 30.094 (SD 3.245), and 35.89 (SD 6.4) mg/dL, respectively. Random Forest demonstrated superior performance across these PHs. We observed significant heterogeneity across all subgroups, indicating diverse sources of variability. As the PH lengthened, the RMSE for blood glucose prediction by the ML model increased, with Random Forest showing the highest relative performance among the ML models.

White House proposed FY 2025 budget would increase MH BG $35 million, keep SUPTRS BG level

On March 11, the White House released the proposed budget for fiscal year (FY) 2025, which begins Oct. 1, 2024. The proposal includes a 10% set‐aside in the Substance Use Prevention, Treatment, and Recovery Services (SUPTRS) Block Grant, which is funded at just over $2 billion, level with FY 2023, the most recent year for which final numbers are available. The Community Mental Health Services Block Grant would receive $1.043 billion in the proposed budget, a $35 million increase compared to FY 2023. The budget request (the White House asks Congress to fund programs) includes a proposal in which the Substance Abuse and Mental Health Services Administration would use HIV instead of AIDS to calculate the set‐aside in the SUPTRS BG. The State Opioid Response (SOR) grant would receive $1.595 billion, an increase of more than $20 million compared to FY 2023. Among other programs included in the White House request: Strategic Prevention Framework–Partnerships for Success at $135.5 million, level compared to FY 2023; Medication‐Assisted Treatment for Prescription Drug and Opioid Addiction at $111 million, level compared to FY 2023; Drug Courts at $74 million, level compared to FY 2023; Screening, Brief Intervention, and Referral to Treatment grants at $33.8 million, level compared to FY 2023; and Pregnant and Postpartum Women care at $43,931,000, an additional $39 million compared to FY 2023).

Phytochemical investigation and evaluation of antihyperglycemic activity of Praecitrullus fistulosus fruits

Diabetes mellitus [antihyperglycemic] is the endocrine metabolic disorder characterized by raised glucose level in the sanguine liquid of body i.e. blood serum due to disturbance of carbohydrate metabolism. In Diabetes mellitus extent of glucose is maintained by well know endocrine pancreatic secreted hormone fuel called insulin which help in metabolizing the glucose level by muscular utilization and other organ pathological activity. When the levels of insulin fall below the body desire it can to hyperglycemic condition. Antihyperglycemic therapeutic plant has an ability to restore abnormal pancreatic condition by increase of insulin, enhancement of body carbohydrate utilization. There are more than 400 reported plant species used an antihyperglycemic and antihyperlipidemic. Praecitrullus fistulosus is reported to contain polyphenols, flavonoids, ascorbic acid, tannin, alkaloid, saponin, phytosterol, diterpines, thiamin, and carotene that possess antidiabetic, antioxidative and anthelmintic activity. On investigation and understanding the cause of diabetes mellitus, had opened the large view for the treatment. Phytomedicine has considerably taken the prominent position for the treatment of many metabolic disorders. Biochemical studies concluded that Praecitrullus fistulosus extract shows alteration in blood glucose level and also Praecitrullus fistulosus improved the lipid profile and showed positive aspect on different parameter in diabetic control rats and it was noted that Praecitrullus fistulosus drug extract was effective and helped in controlling the BG level in Albino Wistar Rat Praecitrullus fistulosus is reported to contain polyphenols, flavonoids, ascorbic acid, tannin, alkaloid, saponin, Phytosterol, diterpines, thiamin, and carotene that possess antidiabetic, antioxidative properties. The aim of research is to evaluate and investigate the anti-hyperglycemic activity of Praecitrullus fistulosus fruit extract on Streptozotocin (STZ)-Nicotinamide induced diabetic rat.
 Keywords: Antihyperglycemic, Praecitrullus fistulosus, Phytochemicals, Phytomedicine

High prevalence of pneumocystis pneumonia in interstitial lung disease: a retrospective study.

Interstitial lung disease (ILD) is a new risk category for pneumocystis pneumonia (PCP) with a high mortality rate. The definite diagnostic criteria of PCP in ILD patients have not been established until now. The aims of this study were to identify potential risk factors of PCP in patients with ILD, and to evaluate the performance of metagenomic next-generation sequencing (mNGS), CD4 + T cell count, (1-3)-β-D-Glucan (BG) and lactate dehydrogenase (LDH) in the diagnosis of PCP in ILD patients. This is a retrospective, single-center, case-control study. ILD patients who underwent mNGS from December 2018 to December 2022 were included in the study. Based on the diagnosis criteria of PCP, these patients were divided into PCP-ILD and non-PCP-ILD groups. The potential risk factors for PCP occurrence in ILD patients were analysed via logistic regression. The diagnostic efficacy of mNGS was compared with serological biomarkers. 92 patients with ILD were enrolled, 31 of which had a definite PCP and were assigned to the PCP-ILD group while 61 were to the non-PCP-ILD group. The infection rate of PJ in ILD patients was 33.7% (31/92). The history of glucocorticoid therapy, CD4 + T cell count, BG level and traction bronchiectasis on HRCT were associated with PCP occurrence in ILD patients. LDH level did not reach statistical significance in the logistic regression analysis. mNGS was confirmed as the most accurate test for PCP diagnosis in ILD patients. ILD is a new risk group of PCP with high PCP prevalence. Clinicians should pay close attention to the occurrence of PCP in ILD patients who possess the risk factors of previous glucocorticoid therapy, decreased CD4 + T cell count, increased BG level and absence of traction bronchiectasis on HRCT. mNGS showed the most excellent performance for PCP diagnosis in ILD patients. Peripheral blood CD4 + T cell count and BG level are alternative diagnostic methods for PCP in ILD patients. However, the diagnostic value of serum LDH level was limited in ILD patients.

THU353 Late Onset Of Immune Checkpoint Inhibitor-Related Diabetes After 3 Years

Abstract Disclosure: S. Avula: None. A. Ahmed: None. M. Salim: None. Z.J. Anderson: None. A. Kumar: None. A. Estrada: None. Background: Immune checkpoint inhibitor-induced diabetes mellitus (ICI-DM) is characterized by acute onset of dramatic hyperglycaemia with severe insulin deficiency following exposure to PD-1/PD-L1 inhibitors. The frequency of type 1 diabetes related to ICIs is estimated to be ∼3.5%. Endogenous insulin secretion is depleted within a few weeks of the clinical onset of diabetes. Moreover, the positive rate for islet-related autoantibodies varies from 5% to 50%. Case: 51-year-old male with a PMH of stage 4 adenocarcinoma of the lung with metastasis to brain, bone, and kidneys was sent to the ED for a POC BG above 700 mg/dl. He was diagnosed with stage 4 lung adenocarcinoma three years prior to the current presentation. He received stereotactic radiotherapy for the brain metastases, Carboplatin/pemetrexed/pembrolizumab x4 cycles, then continued maintenance therapy of pemetrexed/pembrolizumab with the last dose given 3 weeks prior. He had recent evidence of progression, so chemotherapy adjustment was planned, and he started dexamethasone 4mg on the day prior to presentation. Upon ED presentation, he reported new onset polyuria and polydipsia for 1 week accompanied by 10% weight loss over the last 3 months. His vitals were stable. Physical examination was unremarkable. Serum glucose level was found to be 726 mg/dl, Bicarbonate level 28 mEq/L (22-30), creatinine 1.99 mg/dl (baseline 1.2), venous blood PH was 7.29, Bicarbonate 24 mEq/L (24-28), anion gap 15 mEq/L (8-16), B-hydroxybutyrate 2.91 mmol/L (0.02-0.27), sodium 127 mEq/L(135-148) and HbA1c 6.3%. He received normal saline boluses and NPH insulin 15 units SQ, after which his repeat BG level was 426 mg/dl. Given prediabetic Hba1c with recent dexamethasone therapy he was thought to have steroid-induced hyperglycaemia. He was discharged on NPH 15 units daily with Novolog sliding scale. Upon clinic follow-up 1 week later, his BGs did not improve significantly despite stopping steroids. His C-peptide level was low at 0.22 ng/ml (1.10-4.40) with BG 443mg/dl, and GAD -65 antibodies were negative. He continued to have hyperglycaemia while on NPH /Novolog regimen, thus he was switched to Lantus and Novolog regimen with better glycaemic control. He continues to have significant insulin requirements 3 months later. Conclusion: Although the mean interval between the first ICI administration and T1DM development was found to be 201 ± 187 days, the development of type 1 DM may occur with longer intervals from initiating the ICI therapy. In our case, the patient developed new-onset diabetes with low C-peptide after 1,095 days from receiving the first dose of pembrolizumab. On the previous case reports, the presence of GAD-65 antibodies is typically associated with shortened intervals between receiving the ICI therapy and the development of DM. Presentation: Thursday, June 15, 2023

A Challenge To Treat Hypoglycemia For The Diabetic Elderly Patient In Daily Clinical Practice

Background: Hypoglycemia is an acute complication that often occurs in daily clinical practice. The incidence of hypoglycemia is increased, one of which is by age factors and the use of insulin. 
 Case Presentation: A 66-years old Indonesian woman with obesity (BMI=35.2 kg/m2) came to Emergency Department (ED) with decreased of consciousness due to a skipped meal after a prandial insulin injection with errors in calculating insulin dose. She had Diabetes Mellitus type 2 since 10 years ago and got insulin prescription since last year. The random blood sugar (RBG) was 27 mg/dl. Patient was administered intravenous D40% 75ml and the symptoms was improved. Patient was diagnosed this patient with severe hypoglycemia. During hospitalization, patient’s blood sugar had fluctuation in 2 days consecutive mornings and began to stable in day 3 on ward. When the patient discharged, we prescribed 24U basal insulin given before bed if the RBG level >200mg/dl. 
 Discussion: In older population, the neuroglycopenic symptoms, such as lethargy, dizziness, and delirium appear in higher BG level than the autonomic symptoms. The disproportion of demand and supply of blood glucose in the brain causes the neuroglycopenic symptoms. In addition, those signs are also likely to be misinterpreted as signs of stroke, dementia, or vision problems. This cycle of misidiagnosed and the tendency of neuroglypenic symptoms lead the older population into hypoglycemia unawareness. 
 Conclusion: Hypoglycemia unawareness is one of the warning signs of the severity of Diabetes Mellitus type 2 caused by multiple factors which has high prevalence in the elderly. Therefore, particular management is needed in this population by selecting the suitable glucose-lowering agent considering individualized patient preferences rather than focusing on the glycemic target.

Evaluation of Blood Sugar, Blood Pressure and Foot Health in Individuals Not Diagnosed with Diabetes, Hypertension

Objective: Diabetes Mellitus constitutes a significant health concern given its incidence and associated complications. Furthermore, obesity, being a chronic disease, represents a risk factor for hypertension, cardiovascular diseases, and diabetes. In light of these considerations, our study aims to perform a public screening for conditions that are closely related, including hypertension, obesity, diabetes, and foot health. Methods: The study was conducted over a three-day period from September 12 to September 15, 2017, between the hours of 09:00-17:00. The sample comprised of volunteers who passed by the established stands and were not diagnosed with diabetes or hypertension. Blood glucose levels (BG) were measured using a glucometer, while weight and height were measured using a scale. Waist circumference was measured using a tape measure, and foot examinations were conducted using a podoscope. In addition, a questionnaire was administered to gather sociodemographic information and health-related data from participants. Measurements were taken of random capillary BG, blood pressure (BP), height, weight, and waist circumference. A general foot examination was performed, and any pressing foot issues were evaluated with a podoscope. Results: In our study, the mean BG level of 428 participants was 126 g/dl, with 7% of them recording BG levels above 200 g/dl. During foot examinations, 29% of participants reported experiencing pain, while 26% had issues with false nail clipping. Pes planus was detected in 14% of individuals during podoscopic examination, and it was observed that those who used sports shoes and those who had previous foot examinations had fewer foot health problems. Additionally, individuals with high BMI values exhibited a higher prevalence of claudication findings, ingrown nails, and dryness and cracks in the skin. Participants with higher systolic BP were found to have more instances of ingrown nails and other nail problems, including interdigital maceration. Conclusion: The selection of appropriate footwear and regular foot examinations are crucial in the prevention of foot health problems. Our study found a significant association between foot health problems and high BMI and systolic blood pressure levels.

1075-P: Therapeutic Inertia of Glycemic Management in Hospital Setting

Among hospitalized patients, glycemic variability, hyperglycemia and hypoglycemia are associated with increased mortality and other adverse outcomes. Therapeutic inertia, defined as failure to intensify or de-intensify when treatment goal is not met, is found as one of the major obstacles in achieving appropriate glycemic control in the outpatient setting. However, there are very scare data on therapeutic inertia among hospitalized patients with hyperglycemia. Our study looked at hyperglycemia (BG>200) and hypoglycemia (BG<70) rate in hospitalized patients with diabetes admitted in June of 2021 as well as the rate of intensification and de-intensification of insulin within 8 hours of abnormal BG level as defined above. Our study showed that there were treatment adjustment in only 20% of the day with abnormal BG level. Disclosure N.Aung: None. M.E.Lape: None. M.J.Bapana: None. K.Win: None. E.Maung: None. C.Hanani: None. S.Jhaveri: None. A.Amos: None. A.Covaci: None. A.Marquess: None.

Neutrophil-to-leukocyte ratio and admission glycemia as predictors of short-term death in very old elderlies with lobar intracerebral hemorrhage.

The incidence of spontaneous intracerebral hemorrhage (SICH) is highest in very old elderlies (≥75 years). The increasing use of antithrombotic drugs is shifting the epidemiology of SICH towards predominance of lobar subtype, suggesting an incremented propensity of bleeding associated with underlying cerebral amyloid angiopathy. With population aging and antithrombotic use, a parallel raise of proportion of lobar SICH is occurring. Improvement of prognostication in this specific age group and SICH type is needed. Routine blood biomarkers can contribute to prediction of short-term mortality after SICH. Our aim was to investigate the contribution of routine blood biomarkers for short-term mortality (30-days) in elderly patients with lobar SICH. Retrospective analysis of consecutive 130 patients with ≥ 75 years and lobar SICH. The outcome was 30-day mortality. Logistic regression analysis was used to investigate whether admission routine biomarkers can be used as predictors. The case fatality was 40.8%. Admission glycaemia level, neutrophil to lymphocyte ratio and mean platelet volume were significantly different between groups (p = 0.001, p = 0.024, p = 0.038, respectively). There was no significant difference in all other routine biomarkers. On multivariate analysis, admission higher mean BG level (odds ratio [OR]: 1.010, 95% confidence interval [CI]: 1.001-1.019, p = 0.026) and neutrophil to lymphocyte ratio (OR: 1.070, 95% CI: 1.008-1.136, p = 0.027) emerged as predictors. In very old patients with lobar SICH, higher BG level and neutrophil to lymphocyte ratio are associated with increased risk of short-term death.

Effect of Noninvasive Positive Pressure Ventilation on Prognosis and Blood Gas Level in COPD Patients Complicated with Respiratory Failure.

Chronic obstructive pulmonary disease (COPD) is a respiratory disease caused by chronic bronchitis, which seriously threatens the life safety of patients. Noninvasive positive pressure ventilation (NIPPV) has great advantages in its treatment. Here, we explore the effect of NIPPV on prognosis and blood gas level in COPD patients complicated with respiratory failure (RF). A case control study was retrospectively analyzed, where 36 COPD patients with RF were regarded as the regular group to carry on the routine treatment, and 42 patients were assigned to the research group to carry out the routine treatment plus NIPPV. The monofactorial analysis showed that the overall response rate, forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC), and FEV1/FVC in the research group were higher than those in the regular group, while partial pressure of arterial carbondioxide (PaCO2), posttreatment endotracheal intubation (EI), length of stay (LOS), tumor necrosis factor (TNF-α), interleukin (IL)-6, IL-1β, Acute Physiology and Chronic Health Evaluation (APACHE) II score, and modified Medical Research Council (mMRC) scores in the research group were lower than those in the regular group. These results indicated that NIPPV can improve the curative effect of emergency medicine patients with RF, improve BG level and PF, reduce inflammation, and facilitate patient's recovery.

Retrospective, observational study to see the effect of evogliptin on continuous glucose monitoring (CGM) in T2DM Indian patients: A real-world experience

This study aimed to assess effectiveness of Evogliptin 5 mg through continues glucose monitoring (CGM) in patients with T2DM in retrospective observational real world settings. Overall 6 patients who received Evogliptin as routine clinical practice in management of T2DM were analyzed retrospectively from single center. Data collected from past medical records. FreeStyle Librepro 1.0.6 was used for CGM. CGM was done 15 days prior to adding Evogliptin and repeated immediately after that for next 15 days. Mean BG level, Percentage time in target range (80-140mg/dl), Percentage time above target and Percentage time below target were assessed prior and after adding Evogliptin in existing treatment regimen. Significant reduction in Mean blood glucose level seen after adding Evogliptin in existing treatment regimen from 215 mg/dl to 138 mg/dl (-77 mg/dl P=0.006). Significant improvement seen in Percentage time in target range (80-140mg/dl) from 17% to 44% (27% P value 0.007) and in Percentage time above target from 81% to 43% (- 38%, P valve 0.003). 13.5 % of the patients seen below target. Evogliptin was found to be effective when added to the patients who were uncontrolled on other oral anti-diabetic medications. It effectively showed improvement in continues glucose monitoring (CGM) parameters like Mean blood glucose, more number of patients were in Time in Target range i.e (80-140mg/dl) after adding Evogliptin to existing anti-diabetic medications & well tolerated. Small sample size and retrospective study

100 years of insulin-enormous gratitude for my extra 65 years!

My name is Peter Davies, and with this year being the one hundredth anniversary of insulin, it is a tremendous honour to be invited to write this account of my life with Diabetes.

A Negative (1,3)-β-D-Glucan Result Alone Is Not Sufficient to Rule Out a Diagnosis of Pneumocystis Pneumonia in Patients With Hematological Malignancies.

Background: Serum (1,3)-β-D-glucan (BG) testing is increasingly being used in the diagnostic armamentarium for invasive fungal diseases. Given its high sensitivity, some studies suggest that a negative BG result contributes to rule out a diagnosis of Pneumocystis pneumonia (PCP). However, recent reports described a suboptimal sensitivity in HIV-negative immunocompromised patients. In this study, we evaluated the performance of BG assay for PCP diagnosis in HIV-negative patients with diverse PCP risk factors. We also assessed the correlation between Pneumocystis jirovecii load in pulmonary samples and serum BG levels.Methods: We retrospectively included HIV-negative patients with microscopically proven PCP and for whom a BG result was available. We also enrolled patients colonized by Pneumocystis as control group. Colonized patients were matched with PCP patients based on their underlying condition that exposed to PCP. Pulmonary fungal loads were determined by an in-house real-time PCR, and BG levels were measured by using the Fungitell® kit (Associates of Cape Cod, Inc.).Results: Thirty-nine patients were included in each of the two groups. Thirty-four of 39 PCP patients and one of 39 colonized patient had a positive BG test, resulting in a sensitivity of 0.87 (95% CI: 0.73–0.94), a specificity of 0.97 (95% CI: 0.87–0.99), a positive predictive value of 0.97 (95% CI: 0.85–0.99), and a negative predictive value of 0.88 (95% CI: 0.75–0.95) for BG assay. Nonetheless, median BG level differed according to the underlying condition. Among the PCP group, the lowest median level of 211 pg/ml was observed in patients with hematological malignancy (HM) and differed significantly from that observed either in solid organ transplants (3,473 pg/ml) or in patients with autoimmune or inflammatory disorder (3,480 pg/ml). Indeed, the sensitivity of BG assay was estimated at 0.64 (95% CI: 0.35–0.85) in HM patients and was lower than the one observed in the whole PCP group. Furthermore, BG level and fungal burden correlated poorly among all PCP patients.Conclusion: BG is not a reliable biomarker for ruling out PCP in HIV-negative patients with HM. Interpretation of a negative BG result should take into account, but not be limited to, the underlying condition predisposing to PCP.

LSTMs and Deep Residual Networks for Carbohydrate and Bolus Recommendations in Type 1 Diabetes Management.

To avoid serious diabetic complications, people with type 1 diabetes must keep their blood glucose levels (BGLs) as close to normal as possible. Insulin dosages and carbohydrate consumption are important considerations in managing BGLs. Since the 1960s, models have been developed to forecast blood glucose levels based on the history of BGLs, insulin dosages, carbohydrate intake, and other physiological and lifestyle factors. Such predictions can be used to alert people of impending unsafe BGLs or to control insulin flow in an artificial pancreas. In past work, we have introduced an LSTM-based approach to blood glucose level prediction aimed at “what-if” scenarios, in which people could enter foods they might eat or insulin amounts they might take and then see the effect on future BGLs. In this work, we invert the “what-if” scenario and introduce a similar architecture based on chaining two LSTMs that can be trained to make either insulin or carbohydrate recommendations aimed at reaching a desired BG level in the future. Leveraging a recent state-of-the-art model for time series forecasting, we then derive a novel architecture for the same recommendation task, in which the two LSTM chain is used as a repeating block inside a deep residual architecture. Experimental evaluations using real patient data from the OhioT1DM dataset show that the new integrated architecture compares favorably with the previous LSTM-based approach, substantially outperforming the baselines. The promising results suggest that this novel approach could potentially be of practical use to people with type 1 diabetes for self-management of BGLs.

HbA1c Is Not Potent Glucose Control Assessment Tool in Type 2 Diabetes Patients Treated With Insulin

Abstract Glycated haemoglobin (HbA1c) is used for defining of glucose control in diabetic patients nevertheless its insufficiency to present overall control in some specific cases. Continuous Glucose Monitoring (CGM) is usually used for adjustment of insulin doses but the derived data are helpful for exact glucose control. We assess the potency of HbAc for defining of real glucose control in subgroup of type 2 diabetes patients treated with different insulin regimens. We studied 54 diabetic patients (33 men, 21 women; age 60.23±5.99 years, disease duration 12.64±5.02 years) - 33 with type 2 diabetes on pre-mixed insulin, 21 with type 2 on multiple insulin injection (MII). Patients performed multiple daily blood glucose measurements of fasting and prandial blood glucose for three months period. HbA1c was measured and CGM by using iProTM for seven days was performed at the end of this period. In pre-mixed insulin treated group and in intensified regimen group, moderate positive correlation was found between HbA1c and mean blood glucose derived from CGM (7.64±1.40% and 7.69±1.23%, respectively 7.64±1.48mmol/l and 7.60±1.30mmol/l), with r1=0.642 (p<0.01) and r2=0.570 (p<0.05). Even lower was correlation between HbA1c and time-in-range (r1=0.431 and r2 =0.401). There were no correlations between HbA1c and percentage of time spent below the target and number of hypoglycemic episodes in each group. Same trend of correlations was found comparing HbA1c and mean BG level in eight-point profile. Based on HbA1c assessment 36.36% of patients on premixed insulin, 19.05% of type 2 patients on MII were with good control. After estimation of results from SMBG these percentage were respectively 28.14% and 12.11%. CGM defined 27.27% of patients on premixed insulin, 13.80% of type 2 patients on MII as well controlled. We conclude that in insulin treated type 2 patients HbA1c gives relative information about overall control with no precise presenting of glucose fluctuations and out-of-range values of blood glucose with no information about hypoglycemic episodes. Nevertheless, short observed period CGM data could give much information that is comparable to three months blood glucose measurement and could replace the use of HbA1c for assessment of overall control. Reference: (1) Chehregosha H, et al. Diabetes Ther. 2019; 10, 853–863 (2) Beyond A1c Writing Group. Diab Care. 2018; 41: e92-e94 (3) Hirsch I et al. Diabetes Tech Ther 2017, 19 (3): S38-S48

Short-term prediction of future continuous glucose monitoring readings in type 1 diabetes: Development and validation of a neural network regression model

Background and objectiveCGM systems are still subject to a time-delay, which especially during rapid changes causes clinically significant difference between the CGM and the actual BG level. This study had the aim of exploring the potential of developing and validating a model for prediction of future CGM measurements in order to overcome the time-delay. MethodsAn artificial neural network regression (NN) approach were used to predict CGM values with a lead-time of 15 min. The NN were trained and internally validated on 23 million minutes of CGM and externally validated on 2 million minutes of CGM. The validation included data from 278 type 1 diabetes patients using three different CGM sensors. The NN performance were compared with three alternative methods, linear extrapolation, spline extrapolation and last observation carried forward. ResultsThe internal validation yielded a RMSE of 9.1 mg/dL, a MARD of 4.2 % and 99.9 % of predictions were in the A + B zone of the consensus error grid. The external validation yielded a RMSE of 5.9–11.3 mg/dL, a MARD of 3.2–5.4 % and 99.9–100 % of predictions were in the A + B zone of the consensus error grid. The NN performed better on all parameters compared to the two alternative methods. ConclusionsWe proposed and validated a NN glucose prediction model that is potential simple to use and implement. The model only needs input from a CGM system in order to facilitate glucose prediction with a lead time of 15 min. The approach yielded good results for both internal and external validation.

A prospective randomized trial comparing computerized columnar insulin dosing chart (the Atlanta protocol) versus the joint British diabetes societies for inpatient care protocol in management of hyperglycemia in patients with acute coronary syndrome admitted to cardiac care unit in Alexandria, Egypt

BackgroundHyperglycemia in acute coronary syndrome (ACS) is linked to raised morbidity and mortality. Insulin administration using insulin infusion protocols (IIP) is the preferred strategy to control hyperglycemia in critically ill patients. To date, no specific IIP has been identified as the most efficient for achieving glycemic control. Aimto compare glycemic achievements (safety) (primary objective), and coronary and other clinical outcomes (efficacy) (secondary objective) by hyperglycemia management in Cardiac Care Unit (CCU) using computerized Atlanta Protocol (Group (I)) versus paper-based Joint British Diabetes Societies (JBDS) For Inpatient Care Protocol (Group (II)). Patients and methodsThe study was done on 100 ACS patients admitted to Alexandria Main University hospital CCU with RBG >180 mg/dL. They were randomized into the 2 groups in a 1:1 ratio. CBG was measured hourly for 72 hours and was managed by IV insulin infusion. ResultsGroup (I) showed statistically significant less mean time for target BG achievement (3.52 ± 1.53hours), lower incidence of Level 1 hypoglycemia (2%) than Group (II) (4.76 ± 2.33 hours, 22%, p = 0.013, 0.002 respectively) and statistically significant less mean number of episodes above the glycemic target after its achievement than Group (II) (p < 0.001). Regarding Level 2 hypoglycemia the difference was not significant statistically. ConclusionBoth protocols successfully maintained target BG level with low incidence of clinically significant hypoglycemia, however, the computerized Atlanta protocol achieved better glycemic outcomes. We recommend the use of the computerized Atlanta protocol in CCU rather than JBDS for Inpatient Care Protocol whenever this is feasible.

ADO09, a co-formulation of the amylin analogue pramlintide and the insulin analogue A21G, lowers postprandial blood glucose versus insulin lispro in type 1 diabetes.

To compare the safety, pharmacokinetics and pharmacodynamics of ADO09 with insulin lispro (Lispro) and separate subcutaneous injections of human insulin and pramlintide (Ins&Pram) in 24 subjects with type 1 diabetes. At three dosing visits, participants received single doses of ADO09, Ins&Pram or Lispro immediately before eating a standardized mixed meal together with 1 g of acetaminophen, which was used as a surrogate marker to evaluate the kinetics of gastric emptying. Premeal blood glucose was adjusted to 126 mg/dL ± 10% by means of insulin and glucose infusions. The insulin dose was 7.5 U and the pramlintide dose was 45 μg. Blood glucose, glucagon and acetaminophen concentrations were assessed as pharmacodynamic endpoints; insulin and pramlintide concentrations were analysed as pharmacokinetic endpoints, and safety and tolerability were assessed. Compared with Lispro, ADO09 reduced postprandial blood glucose (ppBG) excursions by more than 95% in the first hour postmeal (mean ± SD ∆AUC BG 0-1 h: 1.4 ± 9.9 mg*h/dL vs. 43.5 ± 15.3 mg*h/dL; p < .0001). Maximum ppBG was significantly improved with ADO09 (∆BGmax 87.0 ± 35.5 mg/dL) versus both Lispro (109.2 ± 31.1 mg/dL; p = .0133) and Ins&Pram (109.4 ± 44.3 mg/dL; p = .0357). Gastric emptying with ADO09 was similar to Ins&Pram and significantly slower than with Lispro. All treatments were well tolerated and both adverse events and hypoglycaemic events were rare during the meal test procedure. ADO09 was well tolerated and markedly reduced ppBG compared with Lispro. ADO09 formulation was generally similar to the separate administration of insulin and pramlintide, except for a better BG level in the 4-8 h interval postmeal. These positive results warrant further investigations with ADO09.

914-P: Predictors of ED and Inpatient Admissions after Hypo- and Hyperglycemic Events: Leveraging Remote Monitoring Data of People with Diabetes

Background: Predictive models to identify people with diabetes mellitus (DM) at high-risk for future ED visits and inpatient admissions (IA) are an area of clinical interest. The models do not include self-monitoring blood glucose (SMBG) levels. Livongo, the leader in Applied Health Signals, includes cellular-enabled BG meter that allows instantaneous uploading of SMBG values into the cloud with millions of values in people with DM. Methods: Leveraging data collected via the Livongo program with medical and pharmacy claims, machine learning techniques were used to identify the 25 most predictive variables of severe hypo- and hyperglycemia (BG ≤54 mg/dL and ≥400 mg/dL, respectively) resulting in an ED or IA encounter within three days. Four models were constructed for DM (type 1 and type 2) and encounter (ED and IA). Participants had to be enrolled for at least 12 months with continuous medical benefits eligibility for 24 months. Variable categories included in modeling were demographics, comorbidities based on ICD-10 codes, prior HCU, new and current medications, 30-days SMBG patterns with mean BG levels, and months on program. Area under the curve (AUC) was used to assess model performance. Results: There were 7,633 people selected. They had a mean age of 54 years, 48% were female, and 11% had type 1 DM. In this group, 924 and 1,518 severe hypo- and hyperglycemic with ED or IA encounters occurred. Random forest models had the highest AUC with values greater than 98% and sensitivity and specificity above 93% and 99%, respectively. HCU variables were the most predictive variables in all 4 models. Mean 7-day BG level, 30-day count of BG checks, and before-breakfast checking were also highly predictive. Conclusions: SMBG variables are independent predictors of hypo- and hyperglycemia with ED and IA encounters. Real-time BG remote monitoring programs have the capability to identify people at high-risk of costly HCU and develop interventions to improve care. Disclosure W. Lu: Employee; Self; Livongo Health. R. James: Advisory Panel; Self; AstraZeneca. Employee; Self; Livongo Health. S.L. Painter: Employee; Self; Livongo Health. B. Shah: Employee; Self; Livongo Health.

1925-P: Beneficial Effects of a Novel Long-Acting Glucagon Analog, HM15136, on Obesity and Obesity-Related Metabolic Disorders in Animal Models

Obesity is a major risk factor for a number of metabolic disorders such as hyperlipidemia, glucose intolerance, and T2DM. Recent researches indicated a stimulatory effect of glucagon (GCG) on energy expenditure and appetite inhibition, suggesting the utilization of GCG as a pharmacological therapy for obesity and related metabolic disorders. In line with this, we previously observed that chronic treatment of HM15136, a novel long-acting glucagon analog, conferred efficient body weight loss (BWL) in DIO mice. Here, we further investigated the potential benefits of HM15136 in obesity-related metabolic disorders. In DIO mice, time-course change in fat mass, liver triglyceride (TG), and fasting blood glucose (FBG) was determined. Chronic treatment of HM15136 was associated with progressive BWL (-37.5% vs. vehicle (Veh) at day17) and fat mass reduction with little effect on lean mass, confirming its potent anti-obesity effect. Interestingly, despite marginal BWL at initial drug treatment period, significant reduction in blood TG (-56.5% vs. Veh at day2) and subsequent liver TG (-41.5% vs. Veh at day7) was achieved, and maintained thereafter. Moreover, the TG lowering both in blood and liver was followed by sustainable FBG lowering, suggesting that rapid depletion of lipid, the main sources for gluconeogenesis, might contribute to FBG control by HM15136. To evaluate the potential effect on post-prandial glucose control, oral glucose tolerance test (oGTT) was performed, and reduced BG level during oGTT was observed as BWL occurred. Time-dependent reduction in HOMA-IR evidenced the insulin sensitivity improvement. To our knowledge, this is the first study demonstrating the potential benefits of chronic GCG engagement in the management of BW, blood lipid, and even BG. Based on these results, HM15136 may be a novel option for the management of obesity and related metabolic disorders. Human study is ongoing for the clinical relevance of these findings. Disclosure J. Lee: Employee; Self; Hanmi Pharmaceutical. S. Lee: Employee; Self; Hanmi Pharmaceutical. J. Choi: Employee; Self; Hanmi Pharmaceutical. E. Park: Employee; Self; Hanmi Pharmaceutical. J. Lee: Employee; Self; Hanmi Pharmaceutical. S. Lee: Employee; Spouse/Partner; Hanmi Pharmaceutical. S. Bae: Employee; Self; Hanmi Pharmaceutical. D. Kim: Board Member; Self; Hanmi Pharmaceutical. I. Choi: Board Member; Self; Hanmi Pharmaceutical.