Continuous glucose monitoring: evidence and consensus statement for clinical use

Explainable Machine Learning for Real-Time Hypoglycemia and Hyperglycemia Prediction and Personalized Control Recommendations

BACKGROUND

The occurrences of acute complications arising from hypoglycemia and hyperglycemia peak as young adults with type 1 diabetes (T1D) take control of their own care. Continuous glucose monitoring (CGM) devices provide real-time glucose readings enabling users to manage their control proactively. Machine learning algorithms can use CGM data to make ahead-of-time risk predictions and provide insight into an individual's longer term control.

METHODS

We introduce explainable machine learning to make predictions of hypoglycemia (<70 mg/dL) and hyperglycemia (>270 mg/dL) up to 60 minutes ahead of time. We train our models using CGM data from 153 people living with T1D in the CITY (CGM Intervention in Teens and Young Adults With Type 1 Diabetes)survey totaling more than 28 000 days of usage, which we summarize into (short-term, medium-term, and long-term) glucose control features along with demographic information. We use machine learning explanations (SHAP [SHapley Additive exPlanations]) to identify which features have been most important in predicting risk per user.

RESULTS

Machine learning models (XGBoost) show excellent performance at predicting hypoglycemia (area under the receiver operating curve [AUROC]: 0.998, average precision: 0.953) and hyperglycemia (AUROC: 0.989, average precision: 0.931) in comparison with a baseline heuristic and logistic regression model.

CONCLUSIONS

Maximizing model performance for glucose risk prediction and management is crucial to reduce the burden of alarm fatigue on CGM users. Machine learning enables more precise and timely predictions in comparison with baseline models. SHAP helps identify what about a CGM user's glucose control has led to predictions of risk which can be used to reduce their long-term risk of complications.

Insight into continuous glucose monitoring: from medical basics to commercialized devices

According to the latest statistics, more than 537 million people around the world struggle with diabetes and its adverse consequences. As well as acute risks of hypo- or hyper- glycemia, long-term vascular complications may occur, including coronary heart disease or stroke, as well as diabetic nephropathy leading to end-stage disease, neuropathy or retinopathy. Therefore, there is an urgent need to improve diabetes management to reduce the risk of complications but also to improve patient's quality life. The impact of continuous glucose monitoring (CGM) is well recognized, in this regard. The current review aims at introducing the basic principles of glucose sensing, including electrochemical and optical detection, summarizing CGM technology, its requirements, advantages, and disadvantages. The role of CGM systems in the clinical diagnostics/personal testing, difficulties in their utilization, and recommendations are also discussed. In the end, challenges and prospects in future CGM systems are discussed and non-invasive, wearable glucose biosensors are introduced. Though the scope of this review is CGMs and provides information about medical issues and analytical principles, consideration of broader use will be critical in future if the right systems are to be selected for effective diabetes management.

Does introduction of continuous glucose monitoring at diagnosis of type 1 diabetes increase uptake in children and adolescents?

OBJECTIVE

To assess whether introduction of continuous glucose monitoring (CGM) at diagnosis of type 1 diabetes (T1D), leads to greater uptake and continuation at 12 and 24 months, in a population-based pediatric diabetes clinic.

RESEARCH DESIGN AND METHODS

All T1D children and adolescents diagnosed in the 12 months following full government subsidization of CGM were offered CGM from diagnosis at Women's and Children's Hospital, SA (Cohort 1). Uptake and continuation of CGM was compared to those diagnosed in the preceding year, who were started on CGM after diagnosis, but otherwise had identical diabetes management (Cohort 2). Demographic and clinical data were collected prospectively. The primary outcome variable was CGM wear >75% of the time at 12 and 24 months.

RESULTS

In Cohort 1, 84% were started on CGM at diagnosis. 88% had commenced CGM by 12 months and 90% by 24 months. In Cohort 2, CGM was started on average 10 months after diagnosis (range 1-25 months), with 81% started on CGM within 24 months of subsidization. At 24 months, 78% of Cohort 1 and 66% of Cohort 2 were wearing CGM >75% of the time (p = 0.26), higher than the WCH Clinic as a whole (58%). There was no difference in HbA1c between cohorts.

CONCLUSION

Starting CGM at diagnosis of T1D is feasible and well received by families, with high uptake across all ages. Although CGM continuation (wearing CGM >75% of the time) was slightly higher in Cohort 1 than Cohort 2, this did not reach statistical significance.

Glucose Concentrations from Continuous Glucose Monitoring Devices Compared to Those from Blood Plasma during an Oral Glucose Tolerance Test in Healthy Young Adults

Continuous glucose monitoring devices measure glucose in interstitial fluid. The devices are effective when used by patients with type 1 and 2 diabetes but are increasingly being used by researchers who are interested in the effects of various behaviours of glucose concentrations in healthy participants. Despite their more frequent application in this setting, the devices have not yet been validated for use under such conditions. A total of 124 healthy participants were recruited to a ten-day laboratory study. Each participant underwent four oral glucose tolerance tests, and a total of 3315 out of a possible 4960 paired samples were included in the final analysis. Bland-Altman plots and mean absolute relative differences were used to determine the agreement between the two methods. Bland-Altman analyses revealed that the continuous glucose monitoring devices had proportional bias (R = 0.028, p < 0.001) and a mean bias of -0.048 mmol/L, and device measurements were more variable as glucose concentrations increased. Ninety-nine per cent of paired values were in Zones A and B of the Parkes Error Grid plot, and there was an overall mean absolute relative difference of 16.2% (±15.8%). There was variability in the continuous glucose monitoring devices, and this variability was higher when glucose concentrations were higher. If researchers were to use continuous glucose monitoring devices to measure glucose concentrations during an oral glucose tolerance test in healthy participants, this variability would need to be considered.

Safety and feasibility of a factory-calibrated continuous glucose monitoring system in term and near-term infants at risk of hypoglycemia

Objective

Hypoglycemia increases the risk of adverse neurological outcomes in neonates. Adequate glucose monitoring requires repetitive and painful blood sampling. We aimed to evaluate the feasibility and accuracy of a continuous glucose monitoring system (CGMS) using factory-calibrated sensors to improve glucose monitoring and decrease the frequency of blood samples in neonates.

Material and Methods

A methodological study was conducted to investigate a correlation of CGMS values with blood glucose measurements.

Results

Factory-calibrated CGMS sensors were placed on 21 infants at risk of hypoglycemia after delivery. CGMS values were compared with blood glucose concentrations. Thirty-seven pairs of CGMS and blood glucose values were obtained. There was a good correlation between CGMS and blood glucose values (R=0.67, p<0.01) with a mean difference (2 standard deviations) of 9.78 (-24.68 to 44.25) mg/dL. The mean differences at <3 hours and ≥3 hours after sensor placement were 17.35 (-4.54 to 39.21) mg/dL and 0.88 (-37.62 to 39.38) mg/dL, respectively. CGMS values were significantly higher than blood glucose concentration at <3 hours after sensor placement (p<0.01), whereas no significant differences in glucose values were observed between the CGMS and blood glucose values at ≥3 hours after sensor placement (p=0.852).

Conclusion

The factory-calibrated CGMS was a safe and feasible modality for glucose monitoring. However, it has a tendency to overestimate the blood glucose concentrations. Therefore, this system should be used cautiously for neonates at risk of hypoglycemia, especially within 3 hours after sensor placement.

Ramadan fasting in insulin-treated patients is associated with potentially unfavourable changes in glucose metrics: A flash glucose monitoring (FGM) study

AIM(S)

Ramadan fasting (RF) can represent various challenges to glycaemic control especially in insulin-treated patients with diabetes. We aimed to assess the effect of RF on several glucose metrics using flash glucose monitoring (FGM).

METHODS

Complete FGM data for 29-30 days before, during and after Ramadan were available for 40 patients with type 1 (n = 13) and type 2 diabetes (n = 27) on insulin (with or without oral hypoglycaemic) treatment. Indicators of mean glucose, glucose variability (GV) and time in different glycaemic ranges were analysed.

RESULTS

RF was associated with increase in time in hyperglycaemia (38.5 ± 18.2 vs 48.7 ± 20.7%; P < 0.001) and decrease in time in hypoglycaemia (3.2 ± 2.8 vs 2.1 ± 2.1%; P = 0.003), and time in target range (56.3 ± 17.2 vs 47.9 ± 19.7%, P < 0.001). There were no significant differences in markers of GV with RF; however, RF was associated with a significant reduction in GV during the day but not night time with an increase in the ensuing non-fasting period.

CONCLUSIONS

In insulin-treated patients, RF is associated with an increase in time in hyperglycaemia, a reduced time in target range and nocturnal increase in GV, indicating a need for more refined management algorithms.

Effectiveness of real-time continuous glucose monitoring to improve glycaemic control and pregnancy outcome in patients with gestational diabetes mellitus: a study protocol for a randomised controlled trial

INTRODUCTION

Real-time continuous glucose monitoring (rt-CGM) informs users about current interstitial glucose levels and allows early detection of glycaemic excursions and timely adaptation by behavioural change or pharmacological intervention. Randomised controlled studies adequately powered to evaluate the impact of long-term application of rt-CGM systems on the reduction of adverse obstetric outcomes in women with gestational diabetes (GDM) are missing. We aim to assess differences in the proportion of large for gestational age newborns in women using rt-CGM as compared with women with self-monitored blood glucose (primary outcome). Rates of neonatal hypoglycaemia, caesarean section and shoulder dystocia are secondary outcomes. A comparison of glucose metabolism and quality of life during and after pregnancy completes the scope of this study.

METHODS AND ANALYSIS

Open-label multicentre randomised controlled trial with two parallel groups including 372 female patients with a recent diagnosis of GDM (between 24+0 until 31+6 weeks of gestation): 186 with rt-CGM (Dexcom G6) and 186 with self-monitored blood glucose (SMBG). Women with GDM will be consecutively recruited and randomised to rt-CGM or control (SMBG) group after a run-in period of 6-8 days. The third visit will be scheduled 8-10 days later and then every 2 weeks. At every visit, glucose measurements will be evaluated and all patients will be treated according to the standard care. The control group will receive a blinded CGM for 10 days between the second and third visit and between week 36+0 and 38+6. Cord blood will be sampled immediately after delivery. 48 hours after delivery neonatal biometry and maternal glycosylated haemoglobin A1c (HbA1c) will be assessed, and between weeks 8 and 16 after delivery all patients receive a re-examination of glucose metabolism including blinded CGM for 8-10 days.

ETHICS AND DISSEMINATION

This study received ethical approval from the main ethic committee in Vienna. Data will be presented at international conferences and published in peer-reviewed journals.

TRIAL REGISTRATION NUMBER

NCT03981328; Pre-results.

Efficacy and Safety of Glimepiride With or Without Linagliptin Treatment in Patients With HNF1A Diabetes (Maturity-Onset Diabetes of the Young Type 3): A Randomized, Double-Blinded, Placebo-Controlled, Crossover Trial (GLIMLINA)

OBJECTIVE

Sulfonylureas are first-line treatment of hepatocyte nuclear factor 1-α (HNF1A) diabetes (maturity-onset diabetes of the young type 3), but many patients do not achieve optimal glycemic control without episodes of hypoglycemia. We investigated the combination of the sulfonylurea glimepiride and the dipeptidyl peptidase 4 inhibitor linagliptin versus glimepiride monotherapy with respect to glycemic variability, glycemic control, and risk of hypoglycemia.

RESEARCH DESIGN AND METHODS

In a randomized, double-blinded, crossover trial, patients with HNF1A diabetes (n = 19; mean ± SD age 43 ± 14 years, BMI 24.8 ± 2.8 kg/m², and glycated hemoglobin [HbA_1c] 7.4 ± 0.2% [57.1 ± 7.3 mmol/mol]) were randomly assigned to treatment with glimepiride + linagliptin 5 mg (16 weeks), washout (4 weeks), and glimepiride + placebo (16 weeks) (or vice versa). Glimepiride was titrated targeting a fasting plasma glucose of 4.5-6.0 mmol/L without hypoglycemia. Treatments were evaluated by continuous glucose monitoring (CGM), HbA_1c, and meal test.

RESULTS

Compared with glimepiride + placebo, glimepiride + linagliptin did not significantly improve the primary end point, mean amplitude of glycemic excursions (MAGE) (mean difference -0.7 mmol/L, P = 0.1540), but displayed significant reductions in coefficient of variation on CGM (-3.6%, P = 0.0401), HbA_1c (-0.5%, P = 0.0048), and glimepiride dose (-0.7 mg/day, P = 0.0099). β-cell glucose sensitivity (assessed as C-peptide-to-glucose ratio) during meal test improved with glimepiride + linagliptin. Incidences of hypoglycemia were similar with both treatments.

CONCLUSIONS

Linagliptin as add-on treatment to glimepiride improved glycemic variability and control without increasing risk of hypoglycemia in patients with HNF1A diabetes.

A machine-learning approach to predict postprandial hypoglycemia

BACKGROUND

For an effective artificial pancreas (AP) system and an improved therapeutic intervention with continuous glucose monitoring (CGM), predicting the occurrence of hypoglycemia accurately is very important. While there have been many studies reporting successful algorithms for predicting nocturnal hypoglycemia, predicting postprandial hypoglycemia still remains a challenge due to extreme glucose fluctuations that occur around mealtimes. The goal of this study is to evaluate the feasibility of easy-to-use, computationally efficient machine-learning algorithm to predict postprandial hypoglycemia with a unique feature set.

METHODS

We use retrospective CGM datasets of 104 people who had experienced at least one hypoglycemia alert value during a three-day CGM session. The algorithms were developed based on four machine learning models with a unique data-driven feature set: a random forest (RF), a support vector machine using a linear function or a radial basis function, a K-nearest neighbor, and a logistic regression. With 5-fold cross-subject validation, the average performance of each model was calculated to compare and contrast their individual performance. The area under a receiver operating characteristic curve (AUC) and the F1 score were used as the main criterion for evaluating the performance.

RESULTS

In predicting a hypoglycemia alert value with a 30-min prediction horizon, the RF model showed the best performance with the average AUC of 0.966, the average sensitivity of 89.6%, the average specificity of 91.3%, and the average F1 score of 0.543. In addition, the RF showed the better predictive performance for postprandial hypoglycemic events than other models.

CONCLUSION

In conclusion, we showed that machine-learning algorithms have potential in predicting postprandial hypoglycemia, and the RF model could be a better candidate for the further development of postprandial hypoglycemia prediction algorithm to advance the CGM technology and the AP technology further.

Open Source Software for the Real-Time Control, Processing, and Visualization of High-Volume Electrochemical Data

Electrochemical sensors are major players in the race for improved molecular diagnostics due to their convenience, temporal resolution, manufacturing scalability, and their ability to support real-time measurements. This is evident in the ever-increasing number of health-related electrochemical sensing platforms, ranging from single-measurement point-of-care devices to wearable devices supporting immediate and continuous monitoring. In support of the need for such systems to rapidly process large data volumes, we describe here an open-source, easily customizable, multiplatform compatible program for the real-time control, processing, and visualization of electrochemical data. The software’s architecture is modular and fully documented, allowing the easy customization of the code to support the processing of voltammetric (e.g., square-wave and cyclic) and chronoamperometric data. The program, which we have called Software for the Analysis and Continuous Monitoring of Electrochemical Systems (SACMES), also includes a graphical interface allowing the user to easily change analysis parameters (e.g., signal/noise processing, baseline correction) in real-time. To demonstrate the versatility of SACMES we use it here to analyze the real-time data output by (1) the electrochemical, aptamer-based measurement of a specific small-molecule target, (2) a monoclonal antibody-detecting DNA-scaffold sensor, and (3) the determination of the folding thermodynamics of an electrode-attached, redox-reporter-modified protein.

Sydney Diabetes centre's experience of the Australian Government's roll out of subsidised continuous glucose monitoring for children with type 1 diabetes mellitus

AIM

To determine patient/carer expectations of continuous glucose monitoring (CGM) and short-term satisfaction, to assess the efficacy of CGM in improving: fear of hypoglycaemia and glycaemic control (HbA_1c , ketosis, hypoglycaemia) and to determine time requirements of diabetes clinic staff in commencing and administering CGM.

METHODS

We assessed CGM-naïve patients starting on CGM at a Sydney Diabetes Centre following the introduction of a nationwide government subsidy for CGM. A standardised questionnaire was administered collecting demographic and glycaemic information in addition to Likert scale assessment of expectations and satisfaction. Clinic staff reported time dedicated to CGM education, commencement and follow-up.

RESULTS

A total of 55 patients or parents/carers completed baseline questionnaires, with 37 completing a 3-month follow-up questionnaire. There were high expectations of CGM prior to commencement and high satisfaction ratings on follow-up. CGM improved fear of hypoglycaemia, and total daily insulin dose increased after commencement of CGM. There was a trend towards lower HbA_1c that was not statistically significant and no statistically significant reduction in ketosis or hypoglycaemia. Comments were mostly positive, with some concern raised regarding technical issues and a lack of subsidy after 21 years of age. Staff time requirements were substantial, with an estimated average of 7.7 h per patient per year.

CONCLUSIONS

Patients and families have high expectations of CGM, and satisfaction levels are high in the short term. Total insulin delivery increased after CGM commencement. Time requirements by staff are substantial but are worthwhile if families' overall satisfaction levels are high.

Implementation of discharge recommendations in type 1 diabetes depends on specialist nurse follow-up

AIM

This paper presents an implementation study following previous research including a randomised controlled trial (RCT) of hospital-based home care (HBHC), referring to specialist care in a home-based setting. The aim was to evaluate whether the effects sustained when rolled out into wider practice.

METHODS

In 2013-2014, 42 children newly diagnosed type 1 diabetes were included in the study at a university hospital in Sweden and followed for two years. Measurements of child safety, effects of services, resource use and service quality were included. Descriptive statistics were used to present the results and then discussed in relation to the same intervention of HBHC previously evaluated in an RCT.

RESULTS

Shorter in-hospital stay was partially implemented but increased support after discharge by the diabetes nurse was not. The results indicated that the implemented HBHC was equally effective in terms of child outcomes two years from diagnosis but less effective in terms of parents' outcome. The results furthermore indicated that the quality of services decreased.

CONCLUSION

The suggested overall conclusion was that the implemented HBHC services were safe but had become less effective, at least in relation to the HBHC provided under controlled circumstances.

Current Diabetes Technology: Striving for the Artificial Pancreas

Diabetes technology has continually evolved over the years to improve quality of life and ease of care for affected patients. Frequent blood glucose (BG) checks and multiple daily insulin injections have become standard of care in Type 1 diabetes (T1DM) management. Continuous glucose monitors (CGM) allow patients to observe and discern trends in their glycemic control. These devices improve quality of life for parents and caregivers with preset alerts for hypoglycemia. Insulin pumps have continued to improve and innovate since their emergence into the market. Hybrid closed-loop systems have harnessed the data gathered with CGM use to aid in basal insulin dosing and hypoglycemia prevention. As technology continues to progress, patients will likely have to enter less and less information into their pump system manually. In the future, we will likely see a system that requires no manual patient input and allows users to eat throughout the day without counting carbohydrates or entering in any blood sugars. As technology continues to advance, endocrinologists and diabetes providers need to stay current to better guide their patients in optimal use of emerging management tools.

Patient and disease characteristics of adult patients with type 1 diabetes in Germany: an analysis of the DPV and DIVE databases

BACKGROUND

An understanding of the current status of patients with type 1 diabetes mellitus (T1DM) can help to provide appropriate treatment.

METHODS

This was a retrospective analysis of the DIabetes Versorgungs-Evaluation (DIVE) and the Diabetes-Patienten-Verlaufsdokumentation (DPV) databases for Germany.

RESULTS

The analysis included 56,250 people with T1DM (54.2% male), a median age of 36.8 years, and a median diabetes duration of 12.4 years. 15.3% were obese (body mass index ≥ 30kg/m²). Long-acting insulin analogs were used by 53.3%, short-acting analogs by 72.1%, and oral antidiabetic drugs by 4.7%. Patients had a median glycosylated hemoglobin (HbA1c) of 7.8%. There was a drop in HbA1c and an increase in the rate of hypertension, oral antidiabetic drug use, and in the rate of severe hypoglycemia (all p < 0.01) with age. Flash glucose monitoring (FGM) showed the best glucose values with fewer complications compared to other monitoring systems. HbA1c and FBG were lower in patients using a pump versus multiple daily injections (MDIs; 7.7 versus 7.9% and 7.8 versus 8.7 mmol/l; all adjusted p < 0.01). Patients had a lower risk of at least one severe hypoglycemic or DKA episode during the most recent treatment year with pump treatment compared to MDI (9.4% versus 10.5% and 4.7% versus 6.1%, both adjusted p < 0.01).

CONCLUSION

The data demonstrated less-than-optimal glycemic control in the young, an increasing metabolic pattern in T1DM with increasing age, a benefit of FGM to improve HbA1c control and adverse effects, as well as benefits of pump treatment over MDIs.

Going beyond HbA1c to understand the benefits of advanced diabetes therapies

The gold standard for monitoring overall glycemia is HbA1c. However, HbA1c has several important limitations, giving more weight to the prior 2 to 3 months rather than short-term glycemic control. In addition, the level of the HbA1c does not reflect the important interpersonal differences in its relationship with mean glucose, and HbA1c is affected by many common clinical conditions (anemia, uremia) that can interfere with the accuracy of its measurement in the laboratory. The development and refinement of continuous glucose monitoring (CGM), a glucose- and patient-centric technology, over the past two decades have permitted the creation of new single and composite metrics, such as the percentage of time in range and the glucose pentagon, respectively, which provide clinically relevant insights into short-term glycemic control. In addition, CGM creates new outcome metrics for clinical management and investigational studies (percentage of time in hypoglycemia, percentage of time in target range) that can accurately and meaningfully report the effects of an intervention, whether that is a drug, a device, or a psychosocial program, and CGM provides the key input to drive algorithm-based insulin delivery. Finally, CGM linked with artificial intelligence permits real-time feedback to patients about modifiable patterns of glycemic excursions.

Are glucose profiles well-controlled within the targets recommended by the International diabetes Federation in type 2 diabetes? A meta-analysis of results from continuous glucose monitoring based studies

AIMS

To assess continuous glucose monitoring (CGM) derived intra-day glucose profiles using global guideline for type 2 diabetes recommended by the International Diabetes Federation (IDF).

METHODS

The Cochrane Library, MEDLINE, PubMed, CINAHL and Science Direct were searched to identify observational studies reporting intra-day glucose profiles using CGM in people with type 2 diabetes on any anti-diabetes agents. Overall and subgroup analyses were conducted to summarise mean differences between reported glucose profiles (fasting glucose, pre-meal glucose, postprandial glucose and post-meal glucose spike/excursion) and the IDF targets.

RESULTS

Twelve observational studies totalling 731 people were included. Pooled fasting glucose (0.81 mmol/L, 95% CI, 0.53-1.09 mmol/L), postprandial glucose after breakfast (1.63 mmol/L, 95% CI, 0.79-2.48 mmol/L) and post-breakfast glucose spike (1.05 mmol/L, 95% CI, 0.13-1.96 mmol/L) were significantly higher than the IDF targets. Pre-lunch glucose, pre-dinner glucose and postprandial glucose after lunch and dinner were above the IDF targets but not significantly. Subgroup analysis showed significantly higher fasting glucose and postprandial glucose after breakfast in all groups: HbA1c <7% and ≥7% (53 mmol/mol) and duration of diabetes <10 years and ≥10 years.

CONCLUSIONS

Independent of HbA1c, fasting glucose and postprandial glucose after breakfast are not well-controlled in type 2 diabetes.

Glimepiride monotherapy versus combination of glimepiride and linagliptin therapy in patients with HNF1A-diabetes: a protocol for a randomised, double-blinded, placebo-controlled trial

INTRODUCTION

Hepatocyte nuclear factor 1α (HNF1A)-diabetes is the most common monogenetic subtype of diabetes. Strict glycaemic control is crucial for a good prognosis for patients with HNF1A-diabetes. Sulfonylurea (SU) is used as a first-line therapy in HNF1A-diabetes. However, SU therapy may be problematic as it confers a high risk of hypoglycaemia. We hypothesise that low dose of SU in combination with a dipeptidyl peptidase 4 inhibitor provides a safer and more efficacious treatment in patients with HNF1A-diabetes compared with SU as monotherapy.

METHODS AND ANALYSIS

In a randomised, double-blinded, crossover study, patients with HNF1A-diabetes will randomly be assigned to 16 weeks of treatment with glimepiride+linagliptin, 4 weeks of washout and 16 weeks of treatment with glimepiride+placebo (or vice versa). Treatment will be evaluated with continuous glucose monitoring and combined meal and bicycle tests conducted at baseline and at the end of each of the two treatment periods. The primary end point is the absolute difference in the mean amplitude of glycaemic excursions between the two treatments (glimepiride+linagliptin vs glimepiride+placebo) at the end of each treatment period.

ETHICS AND DISSEMINATION

The study protocol is approved by the Danish Medicines Agency, The Scientific-Ethical Committee of the Capital Region of Denmark (H-17014518) and the Danish Data Protection Agency. The trial will be carried out and monitored in compliance with Good Clinical Practice guidelines and in accordance with the latest version of the Declaration of Helsinki. Positive, negative and inconclusive results will be published at scientific conferences and as one or more scientific manuscripts in peer-reviewed journals with authorship in accordance with the International Committee of Medical Journal Editors' recommendations.

TRIAL REGISTRATION NUMBER

2017-000204-15.

[Technological Innovations in Diabetes Therapy]

Technological Innovations in Diabetes Therapy Abstract. In the last few years a whole array of technical innovations has dramatically increased treatment options for patients with diabetes mellitus. Capillary blood glucose measurements are increasingly replaced by continuous glucose monitoring. More and more insulin pump systems are linked up to continuous glucose monitoring, which thereby become ever more self-regulating. Novel ultra-long and ultra-short acting insulins have become available. There will soon be oral alternatives for several anti-diabetic treatments, which hitherto needed to be injected.

Closed-loop insulin delivery: current status of diabetes technologies and future prospects

INTRODUCTION

Type 1 diabetes is characterised by destruction of pancreatic beta cells, leading to insulin deficiency and hyperglycaemia. The mainstay of treatment remains lifelong insulin therapy as a sustainable cure has as yet proven elusive. The burden of daily management of type 1 diabetes has contributed to suboptimal outcomes for people living with the condition. Innovative technological approaches have been shown to improve glycaemic and patient-related outcomes.

AREAS COVERED

We discuss recent advances in technologies in type 1 diabetes including closed-loop systems, also known as the 'artificial pancreas. Its various components, technical aspects and limitations are reviewed. We also discuss its advent into clinical practice, and other systems in development. Evidence from clinical studies are summarised.

EXPERT COMMENTARY

The recent approval of a hybrid closed-loop system for clinical use highlights the significant progress made in this field. Results from clinical studies have shown safety and glycaemic benefit, however challenges remain around improving performance and acceptability. More data is required to establish long-term clinical efficacy and cost-effectiveness, to fulfil the expectations of people with type 1 diabetes.

A prolonged run-in period of standard subcutaneous microdialysis ameliorates quality of interstitial glucose signal in patients after major cardiac surgery

We evaluated a standard subcutaneous microdialysis technique for glucose monitoring in two critically ill patient populations and tested whether a prolonged run-in period improves the quality of the interstitial glucose signal. 20 surgical patients after major cardiac surgery (APACHE II score: 10.1 ± 3.2) and 10 medical patients with severe sepsis (APACHE II score: 31.1 ± 4.3) were included in this investigation. A microdialysis catheter was inserted in the subcutaneous adipose tissue of the abdominal region. Interstitial fluid and arterial blood were sampled in hourly intervals to analyse glucose concentrations. Subcutaneous adipose tissue glucose was prospectively calibrated to reference arterial blood either at hour 1 or at hour 6. Median absolute relative difference of glucose (MARD), calibrated at hour 6 (6.2 (2.6; 12.4) %) versus hour 1 (9.9 (4.2; 17.9) %) after catheter insertion indicated a significant improvement in signal quality in patients after major cardiac surgery (p < 0.001). Prolonged run-in period revealed no significant improvement in patients with severe sepsis, but the number of extreme deviations from the blood plasma values could be reduced. Improved concurrence of glucose readings via a 6-hour run-in period could only be achieved in patients after major cardiac surgery.

Continuous Glucose Monitoring After Gastric Bypass to Evaluate the Glucose Variability After a Low-Carbohydrate Diet and to Determine Hypoglycemia

BACKGROUND

Roux-en-Y gastric bypass (RYGB) alters glucose metabolism and can cause postprandial hypoglycemia. Continuous glucose monitoring (CGM) has been proposed as an evaluation tool in hypoglycemic RYGB individuals. The objective of this study is to investigate the use of CGM in clinical decision-making including diagnosing hypoglycemia and evaluating treatment effects. Furthermore, we aim to assess its accuracy in RYGB-operated individuals.

METHODS

Thirteen RYGB individuals with symptomatic hypoglycemia and 13 asymptomatic RYGB individuals underwent CGM for 5 days. During this period, a mixed-meal test with concomitant plasma glucose (PG) measurements was performed. Furthermore, the RYGB individuals followed a low-carbohydrate diet (LCD) for 1 day and maintained their ordinary diet (OD) for the rest of the period.

RESULTS

LCD reduced the CGM-determined glycemic variability of the mean interstitial fluid glucose (IFG) significantly compared to OD (p < 0.0001). Receiver operating characteristic analysis confirmed that low blood glucose index (e.g., the frequency and amplitude of hypoglycemic events) is the most reliable parameter related to the development of symptomatic hypoglycemia, with a sensitivity of 0.91 (confidence interval [CI] 0.59; 1.00) and a specificity of 0.77 (CI 0.46; 0.95). However, CGM, measuring the IFG in the subcutaneous adipose tissue, overestimated the minimum glucose levels by 1.1 ± 0.9 mmol/l compared with PG.

CONCLUSIONS

CGM was a good method for demonstrating increased glycemic variability among RYGB individuals and for displaying dietary effects on reducing this glycemic variability, including hypoglycemic events. In RYGB individuals, CGM-measured IFG overestimated the real glucose value by about 1 mmol/l in the hypoglycemic range. This should be taken into consideration if CGM is used to diagnose hypoglycemia after RYGB.

The Comprehensive Glucose Pentagon: A Glucose-Centric Composite Metric for Assessing Glycemic Control in Persons With Diabetes

BACKGROUND

Composite metrics have the potential to provide more complete and clinically useful information about glycemic control than traditional individual metrics such as hemoglobin A1C, %/time/area under curve of hypoglycemia and hyperglycemia.

METHODS

Using five key metrics that are derived from continuous glucose monitoring, we developed a new, multicomponent composite metric, the Comprehensive Glucose Pentagon (CGP) that demonstrates glycemic control both numerically and visually. Two of its axes are composite metrics-the intensity of hypoglycemia and intensity of hyperglycemia. This approach eliminates the use of the surrogate marker, hemoglobin A1C (A1C), and replaces it with glucose-centric metrics.

RESULTS

We reanalyzed the data from two randomized control trials, the STAR 3 and ASPIRE In-Home studies using the CGP. It provided new insights into the effect of sensor-augmented pumping (SAP) in the STAR 3 trial and sensor-integrated pumping with low-glucose threshold suspend (SIP+TS) in the ASPIRE In-Home trial.

CONCLUSIONS

The CGP has the potential to enable health care providers, investigators and patients to better understand the components of glycemic control and the effect of various interventions on the individual elements of that control. This can be done on a daily, weekly, or monthly basis. It also allows direct comparison of the effects on different interventions among clinical trials which is not possible using A1C alone. This new composite metric approach requires validation to determine if it provides a better predictor of long-term outcomes than A1C and/or better predictor of severe hypoglycemia than the low blood glucose index (LBGI).

Continuous glucose monitoring: A review of the technology and clinical use

Continuous glucose monitoring (CGM) is an increasingly adopted technology for insulin-requiring patients that provides insights into glycemic fluctuations. CGM can assist patients in managing their diabetes with lifestyle and medication adjustments. This article provides an overview of the technical and clinical features of CGM based on a review of articles in PubMed on CGM from 1999 through January 31, 2017. A detailed description is presented of three professional (retrospective), three personal (real-time) continuous glucose monitors, and three sensor integrated pumps (consisting of a sensor and pump that communicate with each other to determine an optimal insulin dose and adjust the delivery of insulin) that are currently available in United States. We have reviewed outpatient CGM outcomes, focusing on hemoglobin A1c (A1C), hypoglycemia, and quality of life. Issues affecting accuracy, detection of glycemic variability, strategies for optimal use, as well as cybersecurity and future directions for sensor design and use are discussed. In conclusion, CGM is an important tool for monitoring diabetes that has been shown to improve outcomes in patients with type 1 diabetes mellitus. Given currently available data and technological developments, we believe that with appropriate patient education, CGM can also be considered for other patient populations.

Assessing the Therapeutic Utility of Professional Continuous Glucose Monitoring in Type 2 Diabetes Across Various Therapies: A Retrospective Evaluation

BACKGROUND

There have been few large studies that have analyzed the effect of professional (masked) continuous glucose monitoring (P-CGM) on glycemic control in patients with type 2 diabetes (T2DM) who were on a broad spectrum of baseline therapies.

METHODS

We performed a retrospective, blinded evaluation of glycemic control in 296 T2DM adults for 6 months following a 6- to 7-day study of their glycemic profile using masked P-CGM. At baseline, 91% of the patients were on some form of insulin treatment with oral hypoglycemic agents (OHA), while 7% were on one or more OHAs without insulin, and the remaining 2% were on GLP-1RAs. On the basis of the masked CGM profile, patients were counselled on diet and exercise change(s) in their baseline diabetes therapy by our professionally trained diabetes team. They also continued to receive regular treatment advice and dose titrations through our Diabetes Tele-Management System (DTMS^®). The baseline changes in hemoglobin A1C (A1C) observed in these patients after 6 months of undergoing P-CGM was compared to a matched control group.

RESULTS

P-CGM revealed that the predominant pattern of hyperglycemia was postprandial while previously unknown hypoglycemia was found in 38% of the patients; over half of the cases of hypoglycemia were nocturnal. The mean A1C of the P-CGM group dropped from 7.5 ± 1.4% at baseline vs. 7.0 ± 0.9% at 6 months (p < 0.0001). The frequency of performing self-monitoring of blood glucose (SMBG) was also found to be significantly increased in these patients from the baseline. Meanwhile, no significant improvement in A1C was noted in the control group during the same time frame (7.7 ± 1.1% at baseline vs. 7.4 ± 1.1% at 6 months; p = 0.0663) and frequency of SMBG remained almost unchanged.

CONCLUSIONS

P-CGM can provide actionable data and motivate patients for diabetes self-care practices, resulting in an improvement in glycemic control over a wide range of baseline therapies.

Continuous Glucose Monitoring: A Review of Recent Studies Demonstrating Improved Glycemic Outcomes

Continuous Glucose Monitoring (CGM) has been demonstrated to be clinically valuable, reducing risks of hypoglycemia and hyperglycemia, glycemic variability (GV), and improving patient quality of life for a wide range of patient populations and clinical indications. Use of CGM can help reduce HbA1c and mean glucose. One CGM device, with accuracy (%MARD) of approximately 10%, has recently been approved for self-adjustment of insulin dosages (nonadjuvant use) and approved for reimbursement for therapeutic use in the United States. CGM had previously been used off-label for that purpose. CGM has been demonstrated to be clinically useful in both type 1 and type 2 diabetes for patients receiving a wide variety of treatment regimens. CGM is beneficial for people using either multiple daily injections (MDI) or continuous subcutaneous insulin infusion (CSII). CGM is used both in retrospective (professional, masked) and real-time (personal, unmasked) modes: both approaches can be beneficial. When CGM is used to suspend insulin infusion when hypoglycemia is detected until glucose returns to a safe level (low-glucose suspend), there are benefits beyond sensor-augmented pump (SAP), with greater reduction in the risk of hypoglycemia. Predictive low-glucose suspend provides greater benefits in this regard. Closed-loop control with insulin provides further improvement in quality of glycemic control. A hybrid closed-loop system has recently been approved by the U.S. FDA. Closed-loop control using both insulin and glucagon can reduce risk of hypoglycemia even more. CGM facilitates rigorous evaluation of new forms of therapy, characterizing pharmacodynamics, assessing frequency and severity of hypo- and hyperglycemia, and characterizing several aspects of GV.

Head-to-head comparison between flash and continuous glucose monitoring systems in outpatients with type 1 diabetes

PURPOSE

Continuous glucose monitoring (CGM) is being increasingly used in clinical practice. The flash glucose monitoring (FGM) and CGM are different systems of interstitial glucose recording. We aimed to determine the agreement between the factory-calibrated FGM FreeStyle Libre (FSL) and the gold-standard CGM Dexcom G4 Platinum (DG4P).

METHODS

We analyzed data from n = 8 outpatients with type 1 diabetes, who wore the FSL and DG4P for up to 14 days during their habitual life. We aligned FSL and DG4P recordings to obtain paired glucose measures. We calculated correlation coefficients, mean absolute relative difference (MARD), percentages in Clarke error grid areas, time spent in hyperglycaemia, target glycaemia, or hypoglycaemia, as well as glucose variability with both sensors. Comparison with self-monitoring of blood glucose (SMBG) was also performed.

RESULTS

Patients varied in terms of age, diabetes duration, and HbA1c (from 5.9 to 9.6 %). In the pooled analysis of 10,020 paired values, there was a good correlation between FSL and DG4P (r ² = 0.76; MARD = 18.1 ± 14.8 %) with wide variability among patients. The MARD was significantly higher during days 11-14 than in days 1-10, and during hypoglycaemia (19 %), than in normoglycaemia (16 %) or hyperglycaemia (13 %). Average glucose profiles and MARD versus SMBG were similar between the two sensors. Time spent in normo-, hyper-, or hypoglycaemia, and indexes of glucose variability was similarly estimated by the two sensors.

CONCLUSIONS

In outpatients with type 1 diabetes, we found good agreement between the FSL and DG4P. No significant difference was detected in the estimation of clinical diagnostic parameters.

Glycemic management in critically ill patients

Hyperglycemia associated with critical illness, also called “stress hyperglycemia” or “stress diabetes”, is a consequence of many pathophysiologic hormonal responses including increased catecholamines, cortisol, glucagon, and growth hormone. Alterations in multiple biochemical pathways result in increased hepatic and peripheral insulin resistance with an uncontrolled activation of gluconeogenesis and glycogenolysis. Hyperglycemia has a negative impact on the function of the immune system, on the host response to illness or injury, and on infectious and overall outcomes. The degree of glucose elevation is associated with increased disease severity. Large randomized controlled trials including the Van den Berghe study, the NICE-SUGAR trial, VISEP and GLUCONTROL have shown that the control of glucose levels in critically ill patients has implications on outcome and that both hyperglycemia and hypoglycemia are detrimental and should be avoided. Glucose variability has also been shown to be detrimental. Aggressive glucose control strategies have changed due to the concerns of hypoglycemia and therefore intermediate target glucose control strategies are most often adopted. Different patient populations may vary with regards to optimal glucose targets, timing and approach for glucose control, and with regards to the prognostic significance of glucose excursions and variability. Medical, surgical, and trauma patients may benefit at different rates from glucose control and the approach may need to be adapted to various medical settings and to correspond to the workflow of health providers. Effect modifiers for the success of insulin therapy for hyperglycemia include the methods of nutritional supplementation and exogenous glucose administration. Further research is required to improve insulin protocols for glucose control, to further define glucose targets, and to enhance the accuracy of glucose measuring technologies.

Accuracy of Continuous Glucose Monitoring (CGM) during Continuous and High-Intensity Interval Exercise in Patients with Type 1 Diabetes Mellitus

Continuous exercise (CON) and high-intensity interval exercise (HIIE) can be safely performed with type 1 diabetes mellitus (T1DM). Additionally, continuous glucose monitoring (CGM) systems may serve as a tool to reduce the risk of exercise-induced hypoglycemia. It is unclear if CGM is accurate during CON and HIIE at different mean workloads. Seven T1DM patients performed CON and HIIE at 5% below (L) and above (M) the first lactate turn point (LTP₁), and 5% below the second lactate turn point (LTP₂) (H) on a cycle ergometer. Glucose was measured via CGM and in capillary blood (BG). Differences were found in comparison of CGM vs. BG in three out of the six tests (p < 0.05). In CON, bias and levels of agreement for L, M, and H were found at: 0.85 (−3.44, 5.15) mmol·L⁻¹, −0.45 (−3.95, 3.05) mmol·L⁻¹, −0.31 (−8.83, 8.20) mmol·L⁻¹ and at 1.17 (−2.06, 4.40) mmol·L⁻¹, 0.11 (−5.79, 6.01) mmol·L⁻¹, 1.48 (−2.60, 5.57) mmol·L⁻¹ in HIIE for the same intensities. Clinically-acceptable results (except for CON H) were found. CGM estimated BG to be clinically acceptable, except for CON H. Additionally, using CGM may increase avoidance of exercise-induced hypoglycemia, but usual BG control should be performed during intense exercise.

Survey on the use of insulin pumps in Italy: comparison between pediatric and adult age groups (IMITA study)

AIMS

The aim of the study was to evaluate and compare continuous subcutaneous insulin infusion (CSII) use in pediatric and adult age groups.

METHODS

Data were collected with a questionnaire sent by e-mail to CSII-experienced Diabetes Centers. The questionnaire assessed: (1) number of CSII-treated patients; (2) patient demographic data and characteristics; (3) structure and organization of Diabetes Centers providing CSII therapy; (4) pump characteristics (conventional pump, sensor-augmented pump); and (5) CSII dropouts.

RESULTS

A total of 217 out of 1093 Italian centers participated: 51 pediatric (23.5 %) and 166 (76.5 %) adult centers (AP). Compared to a survey performed in 2005, there was a significant increase in the number of pediatric units when compared to adult units (112 vs 37 %, respectively, p < 0.05). Pediatric age is characterized by a greater concern for quality of life and injections, and a higher dropout rate (10.6 vs 8.9 %) mainly related to pump wearability and site reactions. A complete diabetes-care team is associated with a superior use of technology (fewer dropouts, increased CGM and advanced bolus use) which is, however, still used in a small percentage of patients.

CONCLUSIONS

In Italy, the number of CSII-treated pediatric patients (PP) is growing more significantly when compared to adults. Only 60 % of all patients are using advanced functions and 20 % are using CGMs continuously. This confirms the great interest in diabetes technology that is growing in pediatric diabetologists. However, much improvement is warranted in the organization and specialized training of pediatric, adult and transitional facilities.

[New technologies in diabetology. How far are we from a closed loop?]

Today, assistive technologies are highly important in the treatment of diabetes, especially in the therapy of type 1 diabetes. The use of insulin pumps, for example, has become an established form of treatment. Modern insulin pumps offer various functions, such as different basal rate profiles, split delivery of the meal bolus, and integrated bolus calculators. Some pumps are additionally connected to a blood glucose meter or a continuous glucose-monitoring sensor. Several minimally invasive needle-type sensor systems for continuous tissue glucose monitoring are already available. Recent developments aim at increasing the functionality of insulin pumps and the improvements of sensors for continuous glucose monitoring. In addition, many research groups are working on closing the loop between these two components and thus developing an artificial pancreas, which automatically regulates insulin delivery. The first steps have already been taken and bolus calculators or sensor-augmented insulin pumps with suspension of insulin delivery are now available. Many experimental models show promising results. Prior to the implementation of a fully automated system for everyday use, however, partially automated systems that require user input are to be expected. This article aims at giving an overview of the current state of development in the field of diabetes technology.

AMERICAN ASSOCIATION OF CLINICAL ENDOCRINOLOGISTS AND AMERICAN COLLEGE OF ENDOCRINOLOGY 2016 OUTPATIENT GLUCOSE MONITORING CONSENSUS STATEMENT

This document represents the official position of the American Association of Clinical Endocrinologists and American College of Endocrinology. Where there were no randomized controlled trials or specific U.S. FDA labeling for issues in clinical practice, the participating clinical experts utilized their judgment and experience. Every effort was made to achieve consensus among the committee members. Position statements are meant to provide guidance, but they are not to be considered prescriptive for any individual patient and cannot replace the judgment of a clinician.

A new horizon for glucose monitoring

Regular self-monitoring of blood glucose is crucial for proper insulin dosing and gives a reliable foundation for reasonable glycaemic control. According to recent data, recommended values for glycated haemoglobin A1c as set by the professional associations remain out of the reach for a large proportion of the paediatric population. In the last decades, the treatment of type 1 diabetes has changed significantly as new devices gain a role in routine clinical care. Real-time glucose levels can be monitored with continuous glucose monitoring (CGM), which provides a broad spectrum of information on glucose trends on a moment-to-moment basis. This information can be useful for patients' decision making and clinicians' understanding of patients' conduct. However, several barriers, including the current price, impede a broader use of CGM in most regions of the world. This review summarizes data from randomized, controlled trials that included a paediatric population, and it provides some evidence-based visions for the possible broader utilization of CGM, also for incorporation into insulin delivery devices that enable a closed-loop insulin delivery.

Technology to Reduce Hypoglycemia

Hypoglycemia is a major barrier toward achieving glycemic targets and is associated with significant morbidity (both psychological and physical) and mortality. This article reviews technological strategies, from simple to more advanced technologies, which may help prevent or mitigate exposure to hypoglycemia. More efficient insulin delivery systems, bolus advisor calculators, data downloads providing information on glucose trends, continuous glucose monitoring with alarms warning of hypoglycemia, predictive algorithms, and finally closed loop insulin delivery systems are reviewed. The building blocks to correct use and interpretation of this range of available technology require patient education and appropriate patient selection.

The benefits, limitations, and cost-effectiveness of advanced technologies in the management of patients with diabetes mellitus

Hypoglycemia mitigation is critical for appropriately managing patients with diabetes. Advanced technologies are becoming more prevalent in diabetes management, but their benefits have been primarily judged on the basis of hemoglobin A1c. A critical appraisal of the effectiveness and limitations of advanced technologies in reducing both A1c and hypoglycemia rates has not been previously performed. The cost of hypoglycemia was estimated using literature rates of hypoglycemia events resulting in hospitalizations. A literature search was conducted on the effect on A1c and hypoglycemia of advanced technologies. The cost-effectiveness of continuous subcutaneous insulin infusion (CSII) and real-time continuous glucose monitors (RT-CGM) was reviewed. Severe hypoglycemia in insulin-using patients with diabetes costs $4.9-$12.7 billion. CSII reduces A1c in some but not all studies. CSII improves hypoglycemia in patients with high baseline rates. Bolus calculators improve A1c and improve the fear of hypoglycemia but not hypoglycemia rates. RT-CGM alone and when combined with CSII improve A1c with a neutral effect on hypoglycemia rates. Low-glucose threshold suspend systems reduce hypoglycemia with a neutral effect on A1c, and low-glucose predictive suspend systems reduce hypoglycemia with a small increase in plasma glucose levels. In short-term studies, artificial pancreas systems reduce both hypoglycemia rates and plasma glucose levels. CSII and RT-CGM are cost-effective technologies, but their wide adoption is limited by cost, psychosocial, and educational factors. Most currently available technologies improve A1c with a neutral or improved rate of hypoglycemia. Advanced technologies appear to be cost-effective in diabetes management, especially when including the underlying cost of hypoglycemia.

Interstitium versus Blood Equilibrium in Glucose Concentration and its Impact on Subcutaneous Continuous Glucose Monitoring Systems

The relationship between both interstitial and blood glucose remains a debated topic, on which there is still no consensus. The experimental evidence suggests that blood and interstitial fluid glucose levels are correlated by a kinetic equilibrium, which as a consequence has a time and magnitude gradient in glucose concentration between blood and interstitium. Furthermore, this equilibrium can be perturbed by several physiological effects (such as foreign body response, wound-healing effect, etc.), with a consequent reduction of interstitial fluid glucose versus blood glucose correlation. In the present study, the impact of operating in the interstitium on continuous glucose monitoring systems (CGMs) will be discussed in depth, both for the application of CGMs in the management of diabetes and in other critical areas, such as tight glycaemic control in critically ill patients.

Feasibility of Factory Calibration for Subcutaneous Glucose Sensors in Subjects With Diabetes

BACKGROUND

Continuous glucose monitoring using subcutaneously inserted sensors currently requires blood glucose tests for sensor calibration. Alternatively, sensors precalibrated during the manufacturing process may eliminate the need for fingerstick calibrations. In this study we evaluated the feasibility of sensor factory calibration in subjects with diabetes.

METHODS

A total of 33 subjects with diabetes were asked to wear 4 sensors in parallel, 2 on the arm and 2 on the abdomen. Sensors from a lot with low in vitro sensitivity coefficient of variation were used in the study. Based on frequent capillary blood glucose measurements, the average glucose sensitivity of each sensor was determined over a 5-day wear time. The in vivo sensitivities were analyzed for inter- and intrasubject variation. Mean absolute relative difference (MARD) calculation and consensus error grid analysis (EGA) were performed using a single calibration factor for all sensors, to simulate factory calibration and compared against conventional finger-stick calibration.

RESULTS

The sensitivity coefficient of variation between sensors increased from 2.9% in vitro to 6.0% in vivo. No difference in sensor response between subjects (P = .069) as well as between insertion sites (arm and abdomen) was detected (P = .104). Applying one calibration factor to all sensors in the study resulted in an MARD of 13.4%, and 83.5% of the values fell in consensus EGA zone A. Multiple fingerstick calibration resulted in an MARD of 12.7% and 84.1% in zone A.

CONCLUSIONS

Feasibility of factory calibration was demonstrated in subjects with diabetes using sensors based on "wired enzyme" technology, resulting in accuracy metrics similar to sensors calibrated with capillary blood glucose.