Flash Glucose Monitoring Research Articles

Hip sensor location for flash interstitial glucose monitoring system provides adequate clinical accuracy in non-diabetic hypoglycemic dogs.

Flash glucose monitoring systems (FGMS) are frequently used for interstitial glucose monitoring in dogs with diabetes mellitus and are typically placed between the scapulae. We aimed to evaluate the variability between glucose measurements from FGMS placed in 2 locations (between the scapulae and over the hip) in non-diabetic dogs during rapidly induced hypoglycemia. 24 apparently healthy colony dogs that were subjects in a teaching laboratory. Prospective interventional study. FGMS sensors (FreeStyle Libre 14-day system) were placed between the scapulae and over the hip of all dogs. Regular insulin was administered (0.3 u/kg IV) and subsequent hypoglycemia was corrected. Before insulin administration and every 10 minutes over 90 minutes, interstitial glucose was recorded from both locations, and blood glucose was measured with a point-of-care blood glucose monitor (AlphaTRAK 2). There was a constant bias of 5.6 mg/dL (95% limits of agreement: -26.3 to 37.5 mg/dL) between locations, but the proportional bias was not apparent. There was a correlation between FGMS locations (r = 0.731, P = < .001). Sensor site B was clinically accurate with 100% of paired samples within Parkes error grid zones A (83%) and B (17%) but did not meet the criteria for analytical accuracy. In this model of induced hypoglycemia in healthy dogs, variation between measurements from FGMS locations was unlikely to have affected the clinical outcome. Placement of FGMS over the hip may be an acceptable alternative to placement between the scapulae, but the utility in hyperglycemic dogs is unknown.

Classification of Hypoglycemic Events in Type 1 Diabetes Using Machine Learning Algorithms.

To improve the utilization of continuous- and flash glucose monitoring (CGM/FGM) data we have tested the hypothesis that a machine learning (ML)modelcan be trained to identify the most likely root causes for hypoglycemic events. CGM/FGM data were collected from 449 patients with type 1 diabetes. Of the 42,120 identified hypoglycemic events, 5041 wererandomly selected for classification by two clinicians. Three causes of hypoglycemia were deemed possible to interpret and later validate by insulin and carbohydrate recordings:(1)overestimated bolus (27%), (2)overcorrection of hyperglycemia (29%) and (3)excessive basal insulin presure (44%). The dataset wassplit into a training (n = 4026 events, 304 patients) and an internal validation dataset (n = 1015 events, 145 patients). A number of ML model architectures were applied and evaluated. A separate dataset was generated from 22 patients (13 'known' and 9 'unknown') with insulin and carbohydrate recordings. Hypoglycemic events from this dataset were also interpreted by five clinicians independently. Of the evaluated ML models, a purpose-built convolutional neural network (HypoCNN) performed best. Masking the time series, adding time features and using class weights improvedthe performance of this model, resulting in an average area under the curve (AUC) of 0.921 in the original train/test split. In the dataset validated by insulin and carbohydrate recordings (n = 435 events), i.e. 'ground truth,' our HypoCNN model achieved an AUC of 0.917. The findings support the notion that ML models can be trained to interpret CGM/FGM data. Our HypoCNN model provides a robust and accurate method to identify root causes of hypoglycemic events.

Analysis of Factors Related to Type 2 Diabetic Peripheral Neuropathy Based on Flash Glucose Monitoring System.

BACKGROUND Diabetes peripheral neuropathy is one of the common and serious microvascular complications of type 2 diabetes, which seriously affects the quality of life of patients. Because there is no effective clinical treatment to delay or reverse the progress of DPN. Therefore, early and effective control of the risk factors of DPN is of great significance to prevent the occurrence of DPN and improve the clinical prognosis.The objective of this study was to evaluate factors associated with peripheral neuropathy (DPN) in type 2 diabetes mellitus (T2DM) by flash glucose monitoring (FGM) using a retrospective clinical study. MATERIAL AND METHODS 325 T2DM patients diagnosed and treated in Chu Hsien-I Memorial Hospital of Tianjin Medical University from February 2020 to May 2021.All patients wear FGM continuously for 14 days. According to whether diabetic peripheral neuropathy (DPN) occurs, the patients were divided into DPN group (n=150) and non-DPN group (n=175). By comparing the clinical data, biochemical indicators and blood glucose fluctuation data of the two groups, the risk factors affecting DPN were analyzed. RESULTS Spearman correlation analysis showed that smoking, diabetes course, FBG, 2hPG, HbA1c, HOMA-IR, MBG, CV, SD, MAGE, MODD, TAR were positively, TIR was negatively associated with DPN. Multivariate logistic regression analysis showed that smoking (OR=4.235, 95% CI: 2.151-8.339, P=0.000), diabetes course (OR=1.103, 95% CI: 1.028-1.185, P=0.007), HOMA-IR (OR=1.366, 95% CI: 1.093-1.707, P=0.006), TIR (OR=0.915, 95% CI: 0.853-0.982, P=0.014) were related factors of DPN. CONCLUSIONS Smoking, diabetes course, HOMA-IR and TIR were related factors of type 2 diabetic peripheral neuropathy.

Flash glucose monitoring in young people with type 1 diabetes—a qualitative study of young people, parents and health professionals: ‘It makes life much easier’

ObjectivesFlash glucose monitoring for patients with T1 diabetes avoids frequent painful finger-prick testing, thus potentially improving frequency of glucose self-monitoring. Our study aimed to explore experiences of young people using...

An Old Theme and New Technology: Marked Glucocorticoid-Induced Hyperglycemia in a Hospitalized Patient Uncovered Retrospectively by Her Flash Glucose Monitoring

Abstract Background Glucocorticoid-induced hyperglycemia is a problem-facing endocrinologists and internal medicine specialists in hospital wards. Case History A 63-year-old woman with type 2 diabetes was admitted to the hospital with acute exacerbation of chronic obstructive airway disease. She was treated with a short course of intravenous hydrocortisone followed by oral prednisolone. On discharge, she attended her regular diabetes consultation. Throughout the period, she had her flash glucose monitoring (FGM) sensor in place, and she was monitoring her blood glucose regularly. As part of her diabetes clinic routine, the meter data uploaded and ambulatory glucose profiles were examined. These revealed three distinctly different blood glucose levels before, during, and after glucocorticoid therapy. Glucocorticoid therapy resulted in a marked rise in blood glucose that lasted for a further week before it returned to the pre-treatment levels. This old phenomenon has yet to be demonstrated using the new FGM technology. Conclusions The story (1) asserts the significant impact of glucocorticoid therapy on glycemic control, (2) demonstrates the prolonged impact on glycemic control following discontinuation of glucocorticoids, (3) suggests a lack of adequate monitoring and timely recognition of hyperglycemia in the hospital, and poor management glucocorticoid-induced hyperglycemia either due to failure of conventional monitoring methods or a degree of complacency to appreciate its magnitude.

Transient increase in glucose variability during Ramadan fasting in patients with insulin-treated type 2 diabetes: A preliminary study

AimWe aimed to examine the effect of Ramadan fasting on interstitial glucose control and variability before, during, and after Ramadan in type 2 diabetes patients receiving insulin therapy. MethodsParticipants received a flash glucose monitoring (FGM) system one week before Ramadan that was removed on the sixth or seventh day (pre- and early Ramadan periods) of Ramadan and a second FGM system one week before the end of Ramadan that was removed one week after the end of Ramadan (late and post-Ramadan periods). Fasting blood samples were collected during the pre-, early, and late Ramadan study visits and tested for HbA1c, serum creatinine, and plasma glucose levels. ResultsThirty-four patients were prospectively included. The standard deviation and coefficient of variation of glucose concentrations were higher in the early Ramadan period than in the pre-Ramadan period, but did not differ in the late or post-Ramadan periods. Changes in the early Ramadan period were restricted to males and patients aged <55 years. No significant changes were observed in the average glucose level, glucose management indicator, time in range, time in hyperglycemia, or time in hypoglycemia at any time point. ConclusionsRamadan fasting in patients with insulin-treated type 2 diabetes is associated with an initial increase in glucose variability that quickly returned to pre-Ramadan levels. Ramadan fasting was not associated with any significant changes in glycemic control measures.

Switching to an advanced hybrid closed-loop system in real-world practice improves hypoglycemia awareness and metabolic control in adults with type 1 diabetes, particularly in those with impaired perception of hypoglycemia symptoms

AimsTo evaluate the efficacy of an advance closed-loop (AHCL) system in restoring awareness of hypoglycemia in patients with type 1 diabetes (T1D). MethodsWe conducted a prospective study including 46 subjects with T1D flash glucose monitoring (FGM) or continuous glucose monitoring (CGM) switching to a Minimed 780G® system. Patients were classified in three groups according to the therapy used before switching to Minimed® 780G: multiple dose insulin (MDI) therapy + FGM (n = 6), continuous subcutaneous insulin infusion + FGM (n = 21), and sensor-augmented pump with predictive low-glucose suspend (n = 19). FGM/CGM data were analyzed at baseline, after 2 and 6 months on AHCL. Clarke’s score of hypoglycemia awareness was compared at baseline and 6 months recordings. We also compared the efficacy of the AHCL system in improving A1c among patients with appropriate perception of symptoms of hypoglycemia compared to those presenting with impaired awareness of hypoglycemia (IAH). ResultsParticipants had a mean age of 37 ± 15 and a diabetes duration of 20 ± 10 years. At baseline, 12 patients (27%) showed IAH as defined by a Clarke’s score ≥ 3. Patients with IAH were older and had lower estimated glomerular filtration rate (eGFR) compared with those who did not have IAH; with no differences in baseline CGM metrics or A1c. An overall decrease in A1c was observed after 6 months on AHCL system (from 6.9 ± 0.5% to 6.7 ± 0.6%, P < 0.001), regardless of prior insulin therapy. The improvement in metabolic control was greater in patients with IAH, showing a reduction in A1c from 6.9 ± 0.5 to 6.4 ± 0.4% vs 6.9 ± 0.5 to 6.8 ± 0.6% (P = 0.003), showing a parallel increase in total daily boluses of insulin and automatic bolus correction administered by the AHCL system. In patients with IAH Clarke’s score decreased from 3.6 ± 0.8 at baseline to 1.9 ± 1.6 after 6 months (P < 0.001). After 6 months on AHCL system, only 3 patients (7%) presented with a Clarke’s score ≥ 3, resulting in an absolute risk reduction of 20% (95% confidence interval: 7–32) of having IAH. ConclusionsSwitching from any type of insulin administration to AHCL system improves restoration of hypoglycemia awareness and metabolic control in patients with T1D, particularly in adults with impaired perception of hypoglycemia symptoms. Trial RegistrationClinicalTrial.gov ID NCT04900636.

A comparison of daily glucose fluctuation among GCK-MODY and type 2 diabetes using continuous glucose monitoring technology.

Glucokinase variant-induced maturity-onset diabetes of the young (GCK-MODY) exhibits the unique clinical features of mild fasting hyperglycaemia. However, formal studies of its glucose excursion pattern in daily life in comparison with those with or without other types of diabetes are lacking. We conducted a case-control study including 25 patients with GCK-MODY, 25 A1c-matched, drug naive patients with type 2 diabetes (T2DM) and 25 age-, BMI- and sex-matched subjects with normal glucose tolerance (NGT). All the subjects wore flash glucose monitoring (FGM) sensors for 2 weeks, and glucose readings were masked. Glucose excursion was significantly lower in the GCK-MODY than that in A1c-matched T2DM during the daytime, but was similar during the nighttime. The daytime coefficient of variation [CV] driven by postprandial glucose could separate GCK-MODY from well-controlled T2DM, but the nighttime CV could not. In discriminating between GCK-MODY and T2DM, the area under the curve (AUC) of the CV was 0.875. However, in GCK-MODY and NGT subjects, the CVs were similar at 24h, whereas the other four excursion parameters were significantly higher in GCK-MODY than those in NGT subjects. FGM confirmed the stability and mildness of hyperglycemia in GCK-MODY patients. Postprandial regulation is a key driver of the difference in excursion between GCK-MODY and T2DM.

The Clinical Impact of Flash Glucose Monitoring-a Digital Health App and Smartwatch Technology in Patients With Type 2 Diabetes: Scoping Review.

Type 2 diabetes has a growing prevalence and confers significant cost burden to the health care system, raising the urgent need for cost-effective and easily accessible solutions. The management of type 2 diabetes requires significant commitment from the patient, caregivers, and the treating team to optimize clinical outcomes and prevent complications. Technology and its implications for the management of type 2 diabetes is a nascent area of research. The impact of some of the more recent technological innovations in this space, such as continuous glucose monitoring, flash glucose monitoring, web-based applications, as well as smartphone- and smart watch-based interactive apps has received limited attention in the research literature. This scoping review aims to explore the literature available on type 2 diabetes, flash glucose monitoring, and digital health technology to improve diabetic clinical outcomes and inform future research in this area. A scoping review was undertaken by searching Ovid MEDLINE and CINAHL databases. A second search using all identified keywords and index terms was performed on Ovid MEDLINE (January 1966 to July 2021), EMBASE (January 1980 to July 2021), Cochrane Central Register of Controlled Trials (CENTRAL; the Cochrane Library, latest issue), CINAHL (from 1982), IEEE Xplore, ACM Digital Libraries, and Web of Science databases. There were very few studies that have explored the use of mobile health and flash glucose monitoring in type 2 diabetes. These studies have explored somewhat disparate and limited areas of research, and there is a distinct lack of methodological rigor in this area of research. The 3 studies that met the inclusion criteria have addressed aspects of the proposed research question. This scoping review has highlighted the lack of research in this area, raising the opportunity for further research in this area, focusing on the clinical impact and feasibility of the use of multiple technologies, including flash glucose monitoring in the management of patients with type 2 diabetes.

Management of Glycemia during Acute Aerobic and Resistance Training in Patients with Diabetes Type 1: A Croatian Pilot Study

(1) Background: The increased risk of developing hypoglycemia and worsening of glycemic stability during exercise is a major cause of concern for patients with type 1 diabetes mellitus (T1DM). (2) Aim: This pilot study aimed to assess glycemic stability and hypoglycemic episodes during and after aerobic versus resistance exercises using a flash glucose monitoring system in patients with T1DM. (3) Participants and Methods: We conducted a randomized crossover prospective study including 14 adult patients with T1DM. Patients were randomized according to the type of exercise (aerobic vs. resistance) with a recovery period of three days between a change of groups. Glucose stability and hypoglycemic episodes were evaluated during and 24 h after the exercise. Growth hormone (GH), cortisol, and lactate levels were determined at rest, 0, 30, and 60 min post-exercise period. (4) Results: The median age of patients was 53 years, with a median HbA1c of 7.1% and a duration of diabetes of 30 years. During both training sessions, there was a drop in glucose levels immediately after the exercise (0'), followed by an increase at 30' and 60', although the difference was not statistically significant. However, glucose levels significantly decreased from 60' to 24 h in the post-exercise period (p = 0.001) for both types of exercise. Glycemic stability was comparable prior to and after exercise for both training sessions. No differences in the number of hypoglycemic episodes, duration of hypoglycemia, and average glucose level in 24 h post-exercise period were observed between groups. Time to hypoglycemia onset was prolonged after the resistance as opposed to aerobic training (13 vs. 8 h, p = NS). There were no nocturnal hypoglycemic episodes (between 0 and 6 a.m.) after the resistance compared to aerobic exercise (4 vs. 0, p = NS). GH and cortisol responses were similar between the two sessions, while lactate levels were significantly more increased after resistance training. (5) Conclusion: Both exercise regimes induced similar blood glucose responses during and immediately following acute exercise.

Use of the FreeStyle Libre system and diabetes treatment progression in T2DM: Results from a retrospective cohort study using a Canadian private payer claims database.

Up to one-third of Canadians are estimated to be living with prediabetes or diabetes. A retrospective study using Canadian private drug claims data was conducted to investigate whether flash glucose monitoring using the FreeStyle Libre system (FSL) among people with type 2 diabetes mellitus (T2DM) in Canada can be associated with changes in treatment intensification when compared with blood glucose monitoring (BGM) alone. Using a Canadian national private drug claims database comprising approximately 50% coverage of insured individuals in Canada, cohorts of people with T2DM using FSL or BGM were identified algorithmically based on treatment history and followed over a 24-month study period, tracking their progression in diabetes treatment therapy. The Andersen-Gill model for recurrent time-to-event data was used to evaluate whether the rate of treatment progression differs between the FSL and BGM treatment cohorts. The survival function was used to calculate comparative treatment progression probabilities between the cohorts. In total, 373 871 people with T2DM met the inclusion criteria. Across treatment (FSL) and control (BGM) groups, people using FSL had a higher probability of treatment progression compared with BGM alone, with a relative risk ranging between 1.86 and 2.81 (p < .001). A higher probability of treatment progression was independent of the diabetes treatment at the enrolment date (index date) or the patient status, and independent of whether patients were treatment naïve or on established diabetes therapy. Assessment of the ending treatment relative to the starting therapy indicated that dynamic treatment changes were most evident for patients in the FSL cohort and that the FSL cohort had a much greater portion of patients who ended with insulin treatment (when they started with non-insulin treatment) compared with the BGM cohort. People with T2DM using FSL had a greater probability for treatment progression compared with BGM alone, irrespective of the starting therapy, which may suggest that FSL can be used to support escalation of diabetes therapy to improve therapeutic inertia in T2DM.

Monitoring of Diabetes Mellitus Using the Flash Glucose Monitoring System: The Owners’ Point of View

The flash glucose monitoring system (FGMS) has recently become one of the most common monitoring methods in dogs and cats with diabetes mellitus. The aim of this study was to evaluate the impact of FGMS on the quality of life of diabetic pet owners (DPOs). Fifty DPOs were asked to answer a 30-question survey. More than 80% of DPOs considered FGMS easier to use and less stressful and painful for the animal compared to blood glucose curves (BGCs). Overall, 92% of DPOs reported that their pet had better diabetes control since using FGMS. The most challenging aspects of using the FGMS were ensuring proper sensor fixation during the wearing period (47%), preventing premature detachment (40%), and purchasing the sensor (34%). Moreover, 36% of DPOs reported that the device cost was difficult to afford in the long term. Comparing dogs and cats, a significantly higher number of dogs' owners found the FGMS to be well-tolerated (79% vs. 40%), less invasive than BGCs (79% vs. 43%), and easier to maintain in situ (76% vs. 43%). In conclusion, FGMS is considered by DPOs to be easy to use and less stressful compared to BGCs, while enabling better glycemic control. Nevertheless, the costs related to its long-term use might be difficult to sustain.

IDF2022-1246 Flash glucose monitoring to correct nocturnal and fasting hyperglycemia with basal insulin in T2DM with CKD Stage 3B

IDF2022-1246 Flash glucose monitoring to correct nocturnal and fasting hyperglycemia with basal insulin in T2DM with CKD Stage 3B

Using flash glucose monitoring in pregnancies in routine care of patients with gestational diabetes mellitus: a pilot study

AimFlash glucose monitoring (FGM) has been approved for the care of pregnant women with preexisting diabetes since 2017. However, its use in gestational diabetes (GDM) has been critically discussed. Inaccuracy and missing recommendations for target values are the main arguments against the use of FGM in GDM. To date, there is a lack of data to justify routine use of FGM in GDM pregnancies. Consequently, this new technology has been withheld from GDM-patients. Aim of our pilot study was to analyze the impact of FGM use on pregnancy outcomes, patient’s satisfaction and to confirm the safe use in GDM pregnancies.MethodsCohort study of 37 FGM-managed GDM pregnancies compared with 74 matched women using self-monitoring of blood glucose (SMBG). Group comparison using nonparametric testing concerning patients characteristic and perinatal outcome focusing on adverse outcomes (preeclampsia, preterm delivery, large for gestational age, C-sections, neonatal intensive care unit admission, hyperbilirubinemia and hypoglycemia). Evaluation of patient’s treatment satisfaction using the “Diabetes Treatment Satisfaction Questionnaire change” (DTSQc) and patient interviews.ResultsNo significant differences in patient’s characteristics despite gestational age at diagnosis (FGM with 20 vs. SMBG with 25 weeks). No difference in gestational weight gain, HbA1c progression and perinatal outcome. Treatment satisfaction obtained by the DTSQc revealed a high level of satisfaction with FGM use.ConclusionFGM use was well accepted and did not affect perinatal outcome. Use of FGM during pregnancy is safe and non-inferior to the management with SBGM. FGM should be considered as an option in the management of GDM patients.

Orthorexia nervosa and type 1 diabetes: results of a cross-sectional study.

Orthorexia nervosa (ON) is a condition characterized by an excessive importance attributed to the intake of healthy foods. This study was aimed at investigating the prevalence of ON in subjects with type 1 diabetes (T1D) compared to control subjects. Patient with T1D using either flash glucose monitoring or continuous glucose monitoring were enrolled. For the selection of control group, each patient was asked to indicate one non-diabetic subject of their same sex and approximate age among colleagues at work and school. Patients and controls completed the following questionnaires: ORTO-15, Dusseldorf Orthorexie Scale (DOS), Eating Disorder Examination Questionnaire (EDE-Q) and Brief Symptom Inventory (BSI). The principal outcome was the prevalence of ON among T1D and control subjects. We enrolled 44 patients with T1D aged 39.7 ± 15.7years, with BMI 24.3 ± 4.3kg/m2, and mean HbA1c 53.5 [49-57] mmol/mol. Control subjects were similar to T1D with respect to sex, age and BMI. Thirty-two [72%] and 29 [65%] subjects among patients and controls, respectively, had ORTO15 < 40 (between-group p = 0.48). Two (4.5%) and zero subjects among patients and controls, respectively, had DOS ≥ 30 (p = 0.29). Median scores of DOS, but not of ORTO-15, were significantly higher in patients than in controls. None of the metabolic variables showed a correlation with psychometric tests in T1D. Although the prevalence of ON was not significantly higher in T1D than in controls, patients with T1D showed higher scores of some, but not all, tests assessing orthorexia, without any significant correlation with metabolic parameters.

Glycemia Risk Index Assessment in a Pediatric and Adult Patient Cohort With Type 1 Diabetes Mellitus.

To evaluate the glycemia risk index (GRI) as a new glucometry in pediatric and adult populations with type 1 diabetes (T1D) in clinical practice. A cross-sectional study of 202 patients with T1D receiving intensive treatment with insulin (25.2% continuous subcutaneous insulin infusion [CSII]) and intermittent scanning (flash) glucose monitoring (isCGM). Clinical and glucometric isCGM data were collected, as well as the component of hypoglycemia (CHypo) and component of hyperglycemia (CHyper) of the GRI. A total of 202 patients (53% males and 67.8% adults) with a mean age of 28.6 ± 15.7 years and 12.5 ± 10.9 years of T1D evolution were evaluated.Adult patients (>19 years) presented higher glycated hemoglobin (HbA1c) (7.4 ± 1.1 vs 6.7 ± 0.6%; P < .01) and lower time in range (TIR) (55.4 ± 17.5 vs 66.5 ± 13.1%; P < .01) values than the pediatric population, with lower coefficient of variation (CV) (38.6 ± 7.2 vs 42.4 ± 8.9%; P < .05). The GRI was significantly lower in pediatric patients (48.0 ± 22.2 vs 56.8 ± 23.4; P < .05) associated with higher CHypo (7.1 ± 5.1 vs 5.0 ± 4.5; P < .01) and lower CHyper (16.8 ± 9.8 vs 26.5 ± 15.1; P < .01) than in adults.When analyzing treatment with CSII compared with multiple doses of insulin (MDI), a nonsignificant trend to a lower GRI was observed in CSII (51.0 ± 15.3 vs 55.0 ± 25.4; P= .162), with higher levels of CHypo (6.5 ± 4.1 vs 5.4 ± 5.0; P < .01) and lower CHyper (19.6 ± 10.6 vs 24.6 ± 15.2; P < .05) compared with MDI. In pediatric patients and in those with CSII treatment, despite a better control by classical and GRI parameters, higher overall CHypo was observed than in adults and MDI, respectively. The present study supports the usefulness of the GRI as a new glucometric parameter to evaluate the global risk of hypoglycemia-hyperglycemia in both pediatric and adult patients with T1D.

A dual-attention based coupling network for diabetes classification with heterogeneous data

Diabetes Mellitus (DM) is a group of metabolic disorders characterized by hyperglycaemia in the absence of treatment. Classification of DM is essential as it corresponds to the respective diagnosis and treatment. In this paper, we propose a new coupling network with hierarchical dual-attention that utilizes heterogeneous data, including Flash Glucose Monitoring (FGM) data and biomarkers in electronic medical records. The long short-term memory-based FGM sub-network extracts the time-dependent features of dynamic FGM sequences, while the biomarkers sub-network learns the features of static biomarkers. The convolutional block attention module (CBAM) for dispersing the feature weights of the spatial and channel dimensions is implemented into the FGM sub-network to endure the variability of FGM and allows us to extract high-level discriminative features more accurately. To better adjust the importance weights of the characteristics of the two sub-networks, self-attention is introduced to integrate the characteristics of heterogeneous data. Based on the dataset provided by Peking University People’s Hospital, the proposed method is evaluated through factorial experiments of multi-source heterogeneous data, ablation studies of various attention strategies, time consumption evaluation and quantitative evaluation. The benchmark tests reveal the proposed network achieves a type 1 and 2 diabetes classification accuracy of 95.835% and the comprehensive performance metrics, including Matthews correlation coefficient, F1-score and G-mean, are 91.333%, 94.939% and 94.937% respectively. In the factorial experiments, the proposed method reaches the maximum area under the receiver operating characteristic curve of 0.9428, which indicates the effectiveness of the coupling between the nominated sub-networks. The coupling network with a dual-attention strategy performs better than the one without or only with a single-attention strategy in the ablation study as well. In addition, the model is also tested on another data set, and the accuracy of the test reaches 94.286%, reflecting that the model is robust when it is transferred to untrained diabetes data. The experimental results show that the proposed method is feasible in the classification of diabetes types. The code is available at https://github.com/bitDalei/Diabetes-Classification-with-Heterogeneous-Data.

Effect of switching from intermittently scanned to real-time continuous glucose monitoring in adults with type 1 diabetes: 24-month results from the randomised ALERTT1 trial

Comparing Continuous With Flash Glucose Monitoring In Adults With Type 1 Diabetes (ALERTT1) examined whether switching from first-generation intermittently scanned continuous glucose monitoring (isCGM) without alerts to real-time continuous glucose monitoring (rtCGM) with alert functionality offers additional benefits to adults with type 1 diabetes. The extension of the randomised ALERTT1 trial assessed the effect of switching from isCGM to rtCGM up to 24 months. In this 6-month, double-arm, parallel-group, non-masked, randomised, controlled trial, done across six hospitals in Belgium, 254 adults aged 18 years or older with type 1 diabetes previously using isCGM were randomly assigned (1:1) to rtCGM with alerts (intervention; n=127) or isCGM without alerts (control; n=127). Upon completion of the 6-month trial, the control group switched to rtCGM (is-rtCGM group), and the intervention group continued rtCGM (rt-rtCGM group). The extension focused on within-group changes in time in range (TIR; 3·9-10·0 mmol/L; primary outcome), HbA1c, time in clinically significant hypoglycaemia (<3·0 mmol/L), and Hypoglycaemia Fear Survey worry (HFS-worry) score (all prespecified key secondary outcomes). Mean within-group change versus the start of rtCGM is reported, with a positive value referring to a lower value at start of rtCGM. This trial is registered at ClinicalTrials.gov (NCT03772600). 119 participants were assigned to the is-rtCGM group of whom 112 (94%) completed the 24-month trial, and 123 participants were assigned to the rt-rtCGM group of whom 117 (95%) completed the 24-month trial. TIR increased from 51·8% (95% CI 49·1-54·5) at start of rtCGM (month 6) to 63·5% (60·7-66·3) at month 12 in the is-rtCGM group, and remained stable up to month 24 (change 11·7 percentage points [pp] [9·4-14·0; p<0·0001). In the rt-rtCGM group, TIR increased from 52·5% (95% CI 49·8-55·1) at start of rtCGM (month 0) to 63·0% (60·3-65·8) at month 12, also remaining stable up to month 24 (change 10·5 pp [8·2-12·8]; p<0·0001). HbA1c decreased from 7·4% (57 mmol/mol; month 6) to 6·9% (52 mmol/mol) at month 24 (change -0·54 pp [95% CI -0·64 to -0·44]; -5 mmol/mol [95% CI -6 to -4]; p<0·0001) in the is-rtCGM group, and from 7·4% (57 mmol/mol; month 0) to 7·0% (53 mmol/mol) at month 24 (change -0·43 pp [95% CI -0·53 to -0·33]; -4 mmol/mol [95% CI -5 to -3]; p<0·0001) in the rt-rtCGM group. The change in HFS-worry score was -2·67 (month 24 vs month 6; p=0·0008) in the is-rtCGM group and -5·17 points (month 24 vs month 0; p<0·0001) in the rt-rtCGM group. Time in clinically significant hypoglycaemia was unchanged in both groups after month 12. Severe hypoglycaemia decreased from 31·0 to 3·3 per 100 patient-years after switching to rtCGM. Glycaemic control and hypoglycaemia worry improved significantly up to 24 months after switching from isCGM without alerts to rtCGM with alerts, supporting the use of rtCGM in the care of adults with type 1 diabetes. Dexcom.

Accuracy of Flash Glucose Monitoring in Hemodialysis Patients With and Without Diabetes Mellitus.

Glucose and insulin metabolism are altered in hemodialysis patients, and diabetes management is difficult in these patients. We aimed to validate flash glucose monitoring (FGM) in hemodialysis patients with and without diabetes mellitus as an attractive option for glucose monitoring not requiring regular self-punctures. We measured interstitial glucose using a FreeStyle Libre device in eight hemodialysis patients with and seven without diabetes mellitus over 14 days and compared the results to simultaneously performed self-monitoring of capillary blood glucose (SMBG). In 720 paired measurements, mean flash glucose values were significantly lower than self-measured capillary values (6.17±2.52 vs. 7.15±2.41 mmol/L, p=1.3 E-86). Overall, the mean absolute relative difference was 17.4%, and the mean absolute difference was 1.20 mmol/L. The systematic error was significantly larger in patients without vs. with diabetes (- 1.17 vs. - 0.82 mmol/L) and on dialysis vs. interdialytic days (-1.09 vs. -0.90 mmol/L). Compared to venous blood glucose (72 paired measurements), the systematic error of FGM was even larger (5.89±2.44 mmol/L vs. 7.78±7.25 mmol/L, p=3.74E-22). Several strategies to reduce the systematic error were evaluated, including the addition of +1.0 mmol/L as a correction term to all FGM values, which significantly improved accuracy. FGM systematically underestimates blood glucose in hemodialysis patients but, taking this systematic error into account, the system may be useful for glucose monitoring in hemodialysis patients with or without diabetes.

Patient Perspectives on the Ambulatory Glucose Profile Report for Type 1 Diabetes Management in Adults: A National Online Survey

ObjectivesContinuous and flash glucose monitoring devices produce data reports (e.g. ambulatory glucose profile [AGP]) that can be used by people with diabetes and health-care providers (HCPs). Clinical benefits of these reports have been published, but the patient perspective is underreported. MethodsWe conducted an online survey for adults with type 1 diabetes (T1D) using continuous/flash glucose monitoring to understand their use and attitudes toward the AGP report. Barriers and facilitators related to digital health technology were explored. ResultsThe survey included 291 respondents: 63% were <40 years of age, and 65% had been living with T1D for >15 years. Nearly 80% reviewed their AGP report, with 50% often discussing it with their HCP. Support from family and HCPs was positively associated with the use of the AGP report, and there was a positive relationship between motivation and better understanding of the AGP report (odds ratio=2.61; 95% confidence interval, 1.45 to 4.71). Nearly all respondents (92%) indicated that the AGP report is important for their diabetes management, but most indicated dissatisfaction with the cost of the device. Open-ended responses suggested some concern with the complexity of information from the AGP report. ConclusionsThe online survey showed that there may be few barriers to the use of the AGP report by people with T1D, with the main barrier being its cost of the devices. Facilitators for the use of the AGP report included motivation and support from both family and the HCP. Facilitating discussion between HCPs and patients may be a strategy to improve the use and potential benefit of the AGP.