The road from intermittently scanned glucose monitoring to hybrid closed-loop systems: Part A. Keys to success: subject profiles, choice of systems, education

The Reporting Quality of Machine Learning Studies on Pediatric Diabetes Mellitus: Systematic Review

BACKGROUND

Diabetes mellitus (DM) is a major health concern among children with the widespread adoption of advanced technologies. However, concerns are growing about the transparency, replicability, biasedness, and overall validity of artificial intelligence studies in medicine.

OBJECTIVE

We aimed to systematically review the reporting quality of machine learning (ML) studies of pediatric DM using the Minimum Information About Clinical Artificial Intelligence Modelling (MI-CLAIM) checklist, a general reporting guideline for medical artificial intelligence studies.

METHODS

We searched the PubMed and Web of Science databases from 2016 to 2020. Studies were included if the use of ML was reported in children with DM aged 2 to 18 years, including studies on complications, screening studies, and in silico samples. In studies following the ML workflow of training, validation, and testing of results, reporting quality was assessed via MI-CLAIM by consensus judgments of independent reviewer pairs. Positive answers to the 17 binary items regarding sufficient reporting were qualitatively summarized and counted as a proxy measure of reporting quality. The synthesis of results included testing the association of reporting quality with publication and data type, participants (human or in silico), research goals, level of code sharing, and the scientific field of publication (medical or engineering), as well as with expert judgments of clinical impact and reproducibility.

RESULTS

After screening 1043 records, 28 studies were included. The sample size of the training cohort ranged from 5 to 561. Six studies featured only in silico patients. The reporting quality was low, with great variation among the 21 studies assessed using MI-CLAIM. The number of items with sufficient reporting ranged from 4 to 12 (mean 7.43, SD 2.62). The items on research questions and data characterization were reported adequately most often, whereas items on patient characteristics and model examination were reported adequately least often. The representativeness of the training and test cohorts to real-world settings and the adequacy of model performance evaluation were the most difficult to judge. Reporting quality improved over time (r=0.50; P=.02); it was higher than average in prognostic biomarker and risk factor studies (P=.04) and lower in noninvasive hypoglycemia detection studies (P=.006), higher in studies published in medical versus engineering journals (P=.004), and higher in studies sharing any code of the ML pipeline versus not sharing (P=.003). The association between expert judgments and MI-CLAIM ratings was not significant.

CONCLUSIONS

The reporting quality of ML studies in the pediatric population with DM was generally low. Important details for clinicians, such as patient characteristics; comparison with the state-of-the-art solution; and model examination for valid, unbiased, and robust results, were often the weak points of reporting. To assess their clinical utility, the reporting standards of ML studies must evolve, and algorithms for this challenging population must become more transparent and replicable.

Impact on glucometric variables and quality of life of the advanced hybrid closed-loop system in pediatric and adolescent type 1 diabetes

BACKGROUND

In recent years, technological advances in the field of diabetes have revolutionized the management, prognosis, and quality of life of diabetes patients and their environment. The aim of our study was to evaluate the impact of implementing the MiniMed 780G closed-loop system in a pediatric and adolescent population previously treated with a continuous subcutaneous insulin infusion pump and intermittent glucose monitoring.

METHODS

Data were collected from 28 patients with type 1 diabetes aged 6 to 17 years, with a follow-up of 6 months. We included both glucometric and quality of life variables, as well as quality of life in primary caregivers. Metabolic control variables were assessed at baseline (before system change) and at different cutoff points after initiation of the closed-loop system (48 hours, 7 days, 14 days, 21 days, 1 month, 3 months, 6 months).

RESULTS

Time in range 70-180 mg/dL increased from 59.44% at baseline to 74.29% in the first 48 hours after automation of the new system, and this improvement was maintained at the other cutoff points, as was time in hyperglycemia 180-250 mg/dL (24.44% at baseline to 18.96% at 48 hours) and >250 mg/dL (11.71% at baseline to 3.82% at 48 hours).

CONCLUSIONS

Our study showed an improvement in time in range and in all time spent in hyperglycemia from the first 48 hours after the automation of the system, which was maintained at 6 months.

Insulin pumps in children - a systematic review

BACKGROUND

Insulin pump therapy is a real breakthrough in managing diabetes Mellitus, particularly in children. It can deliver a tiny amount of insulin and decreases the need for frequent needle injections. It also helps to maintain adequate and optimal glycemic control to reduce the risk of metabolic derangements in different tissues. Children are suitable candidates for pump therapy as they need a more freestyle and proper metabolic control to ensure adequate growth and development. Therefore, children and their caregivers should have proper education and training and understand the proper use of insulin pumps to achieve successful pump therapy. The pump therapy continuously improves to enhance its performance and increase its simulation of the human pancreas. Nonetheless, there is yet a long way to reach the desired goal.

AIM

To review discusses the history of pump development, its indications, types, proper use, special conditions that may enface the children and their families while using the pump, its general care, and its advantages and disadvantages.

METHODS

We conducted comprehensive literature searches of electronic databases until June 30, 2022, related to pump therapy in children and published in the English language.

RESULTS

We included 118 articles concerned with insulin pumps, 61 were reviews, systemic reviews, and meta-analyses, 47 were primary research studies with strong design, and ten were guidelines.

CONCLUSION

The insulin pump provides fewer needles and can provide very tiny insulin doses, a convenient and more flexible way to modify the needed insulin physiologically, like the human pancreas, and can offer adequate and optimal glycemic control to reduce the risk of metabolic derangements in different tissues.

Real-world prospective observational single-centre study: Hybrid closed loop improves HbA1c, time-in-range and quality of life for children, young people and their carers

Hybrid closed-loop (HCL) systems are characterised by integrating continuous glucose monitoring (CGM) with insulin pumps which automate insulin delivery via specific algorithms and user-initiated insulin delivery. The aim of the study was to evaluate the effectiveness of HCLs on Hba1c, time-in-range (TIR), time in hypoglycaemia, fear of hypoglycaemia, sleep and quality of life measure in children and young people (CYP) with T1D and their carers. Data on HbA1c, TIR and hypoglycaemia frequency were reviewed at baseline prior to starting HCL and 3 months after commencement. As part of clinical care, all patients and carers were provided with key education on the use of the HCL system by trained diabetes healthcare professionals. CYP aged 12 years and above independently completed the validated Hypoglycaemia Fear Survey (HFS). Parents of patients <12 were asked to complete a modified version of the HFS-Parent (HFS-P) survey. There were 39 CYP (22 men) with T1D included with a mean age of 11.8 ± 4.4 at commencement of HCL. Median duration of diabetes was 3.8 years (interquartile range 1.3-6.0). There were 55% of patients who were prepubertal at the time of HCL commencement. 91% were on the Control-IQ system and 9% on the CamAPS FX system. HCL use demonstrated significant improvements at 3 months in the following: HbA1c in mmol/mol (63.0 vs. 56.6, p = 0.03), TIR (50.5 vs. 67.0, p = 0.001) and time in hypoglycaemia (4.3% vs. 2.8%, p = 0.004). HFS scores showed improved behaviour (34.0 vs. 27.5.9, p = 0.02) and worry (40.2 vs. 31.6, p = 0.03), and HFS-P scores also showed improved behaviour (p < 0.001) and worry (p = 0.01). Our study shows that HCL at 3 months improves glucose control, diabetes management and quality of life measures such as fear and worry of hypoglycaemia for CYP and carers.

Rapid-acting insulin analogues: Theory and best clinical practice in type 1 and type 2 diabetes

Since the discovery of insulin 100 years ago, insulin preparations have improved significantly. Starting from purified animal insulins, evolving to human insulins produced by genetically modified organisms, and ultimately to insulin analogues, all in an attempt to mimic physiological insulin action profiles seen in individuals without diabetes. Achieving strict glucose control without hypoglycaemia and preventing chronic complications of diabetes while preserving quality of life remains a challenging goal, but the advent of newer ultra-rapid-acting insulin analogues may enable intensive insulin therapy without being too disruptive to daily life. Ultra-rapid-acting insulin analogues can be administered shortly before meals and give better coverage of mealtime-induced glucose excursions than conventional insulin preparations. They also increase convenience with timing of bolus dosing. In this review, we focus on the progress that has been made in rapid-acting insulins. We summarize pharmacokinetic and pharmacodynamic data, clinical trial data supporting the use of these new formulations as part of a basal-bolus regimen and continuous subcutaneous insulin infusion, and provide a clinical perspective to help guide healthcare professionals when and for whom to use ultra-fast-acting insulins.

Glucose control using fast-acting insulin aspart in a real-world setting: A 1-year, two-centre study in people with type 1 diabetes using continuous glucose monitoring

AIM

To evaluate the efficacy and safety of switching from traditional mealtime insulins to fast-acting insulin aspart (Fiasp) in a "real-world" clinical practice setting in adult people with type 1 diabetes (PWD1) who were using intermittently scanned or real-time continuous glucose monitoring (isCGM or rtCGM, respectively).

MATERIALS AND METHODS

Data from 438 adult PWD1 (60% men, age 44.6 ± 16.2 years, diabetes duration 21.5 ± 14.0 years, isCGM/rtCGM: 391/47, multiple daily injections/continuous subcutaneous insulin infusion: 409/29), who initiated Fiasp from January 2018 to May 2020, were analysed. The primary objective was the evolution of time in range (TIR; 70-180 mg/dL) at 6 and 12 months. Secondary objectives included change in HbA1c, body mass index (BMI), insulin doses, time below range (<70 and <54 mg/dL), and time above range (>180 and >250 mg/dL).

RESULTS

TIR improved from 50.3% ± 15.6% to 54.3% ± 15.1% at 6 months (n = 425) and to 55.5% ± 15.2% at 12 months (n = 385) (P < .001), corresponding to 57 min/d at 6 months and 75 min/d at 12 months. Time spent below 54 mg/dL evolved from 3.1% ± 3.3% to 3.1% ± 3.7% and 2.5% ± 3.0% at 6 and 12 months, respectively (P = .011). Also, time spent above 180 mg/dL decreased from 42.3% ± 16.7% at start by 4.2% at 6 months and by 4.6% at 12 months (P < .001). The proportion of people reaching TIR more than 70% increased from 11.0% to 14.8% (P = .002), and those spending less than 4% at time less than 70 mg/dL increased from 36.1% to 42.1% (P = .002). After 12 months, HbA1c, insulin doses, and BMI did not change significantly.

CONCLUSIONS

In a Belgian real-world setting of adult PWD1, switching to Fiasp was associated with a 5% increased TIR after 12 months, corresponding to 75 min/d, in combination with less time spent below and above range.

Diabetes Knowledge and Metabolic Control in Type 1 Diabetes Starting With Continuous Glucose Monitoring: FUTURE-PEAK

PURPOSE

To investigate whether diabetes knowledge and health literacy impact glycemic control after 1 year of intermittently scanned continuous glucose monitoring (isCGM) in people with type 1 diabetes  ≥ 16 years.

METHODS

In this prospective real-world cohort study, we assessed diabetes knowledge using a new 10-item questionnaire [Patient Education and Knowledge (PEAK)] and health literacy using the validated 6-item Newest-Vital Sign-D (NVS-D) questionnaire. Primary endpoint was association between PEAK score and change in hemoglobin A1c (HbA1c). Secondary endpoints were link between NVS-D score and change in HbA1c and that between time spent in/above/below range and PEAK/NVS-D scores.

RESULTS

851 subjects were consecutively recruited between July 2016 and July 2018. Median PEAK score was 8 (range: 0-10), and median NVS-D score was 6 (range 0-6). HbA1c evolved from 7.9% (7.8%-8.0%), 63 (62-64) mmol/mol, at start to 7.7% (7.6%-7.7%), 61 (60-61) mmol/mol (P < 0.001), at 6 months and to 7.8% (7.7%-7.9%), 62 (61-63) mmol/mol, at 12 months (P < 0.001). HbA1c only improved in subgroups with higher scores [PEAK subgroups with score 7-8 (P = 0.005) and 9-10 (P < 0.001) and NVS-D score 4-6 (P < 0.001)]. At 12 months, time spent below 70 mg/dL was reduced by 15% (P < 0.001), and time spent below 54 mg/dL was reduced by 14% (P < 0.001), irrespective of PEAK/NVS-D score. Multiple linear regression analysis demonstrated an association of PEAK score, scan frequency, and baseline HbA1c with evolutions in time in range and time in hyperglycemia.

CONCLUSIONS

isCGM reduced time in hypoglycemia, and HbA1c evolved favorably. Our findings suggest that diabetes and health literacy affect glucometrics, emphasizing the importance of education.

Extraction With Sweat-Sebum Emulsion as a New Test Method for Leachables in Patch-Based Medical Devices, Illustrated by Assessment of Isobornylacrylate (IBOA) in Diabetes Products

BACKGROUND

The increasing offering of patch-based medical devices is accompanied by growing numbers of reported adverse skin reactions. Procedures for testing leachables according to ISO 10993 may not be optimal for lipophilic substances that can be mobilized on skin by sweat and sebum. We propose an improved extraction method for targeted analysis of leachables using low volumes of a sweat-sebum emulsion. The approach is illustrated by the analysis of isobornylacrylate (IBOA), a compound found in some devices and suspected for allergenic potential.

METHOD

Three patch-based products were tested: an implantable device for continuous glucose monitoring (CGM), an intermittently scanned CGM (isCGM) device, and a micro-insulin pump. Quantification of IBOA was performed by gas chromatography and allergenic potential of IBOA levels was assessed by the KeratinoSens cell assay. Different combinations were used for extraction solvent (isopropanol, 5% ethanol-water solution, and sweat-sebum emulsion), extraction volumes (complete immersion vs partial immersion in 2 mm of solvent), and extraction time (3, 5, and 14 days).

RESULTS

Isobornylacrylate was only found in the isCGM device. About 20 mg/L IBOA were eluted after 3 days in isopropanol but only about 1 mg/L in ethanol-water. Sweat-sebum emulsion dissolves IBOA better and gives a more stable solution than ethanol-water. Decomposition of IBOA solutions requires adjusted extraction timing or correction of results. In the sweat-sebum extract, IBOA levels were about 20 mg/L after 3 days and about 30 mg/L after 5 days, clearly above the threshold found in the KerationSens assay for keratinocyte activation (10 mg/L).

CONCLUSION

Extraction by low volumes of sweat-sebum emulsion can be a superior alternative for the targeted simulating-use assessment of leachables in patch-based medical devices.

A review of flash glucose monitoring in type 2 diabetes

BACKGROUND

Continuous glucose monitoring systems are increasingly being adopted as an alternative to self-monitoring of blood glucose (SMBG) by persons with diabetes mellitus receiving insulin therapy.

MAIN BODY

The FreeStyle Libre flash glucose monitoring system (Abbott Diabetes Care, Witney, United Kingdom) consists of a factory-calibrated sensor worn on the back of the arm which measures glucose levels in the interstitial fluid every minute and stores the reading automatically every 15 min. Swiping the reader device over the sensor retrieves stored data and displays current interstitial glucose levels, a glucose trend arrow, and a graph of glucose readings over the preceding 8 h. In patients with type 2 diabetes (T2D) receiving insulin therapy, pivotal efficacy data were provided by the 6-month REPLACE randomized controlled trial (RCT) and 6-month extension study. Compared to SMBG, the flash system significantly reduced the time spent in hypoglycemia and frequency of hypoglycemic events, although no significant change was observed in glycosylated hemoglobin (HbA1c) levels. Subsequent RCTs and real-world chart review studies have since shown that flash glucose monitoring significantly reduces HbA1c from baseline. Real-world studies in both type 1 diabetes or T2D populations also showed that flash glucose monitoring improved glycemic control. Higher (versus lower) scanning frequency was associated with significantly greater reductions in HbA1c and significant improvements in other measures such as time spent in hypoglycemia, time spent in hyperglycemia, and time in range. Additional benefits associated with flash glucose monitoring versus SMBG include reductions in acute diabetes events, all-cause hospitalizations and hospitalized ketoacidosis episodes; improved well-being and decreased disease burden; and greater treatment satisfaction.

CONCLUSION

T2D patients who use flash glucose monitoring might expect to achieve significant improvement in HbA1c and glycemic parameters and several associated benefits.

Flash Glucose Monitoring and Patient Satisfaction: A Meta-Review of Systematic Reviews

Flash glucose monitoring (FGM) systems have been suggested to have clinical beneficial effects in patients with diabetes mellitus, although their improvements in terms of quality of life (QoL) and patients’ satisfaction are not always addressed or are considered a secondary outcome. Thus, the aim of this meta-review is to establish the benefits of FGM in terms of patients’ satisfaction and QoL in both type 1 and type 2 diabetes patients using evidence from past systematic reviews and meta-analyses. Major databases were searched for systematic reviews (with or without meta-analyses) that assessed the satisfaction or QoL of type 1 or 2 diabetes patients using FGM compared with other glucose monitoring systems. The quality of the included systematic reviews was addressed with the Assessment of Multiple Systematic Reviews 2 (AMSTAR-2) tool. Six systematic reviews (including two meta-analyses) were included in the meta-review. Evidence suggests that FGM systems seem to improve patients’ satisfaction and QoL compared with self-monitoring of blood glucose, although the high variability in the measurement tools, the clinical significance and the quality of the systematic reviews included do not allow us to state FGM benefits with any certainty. Further research, including high-quality randomised clinical trials, differentiating the needs of both type 1 and type 2 diabetes patients and focusing on psychosocial benefits for these patients is needed to optimise clinical decisions between patients and professionals by developing the right health technology assessment for FGM systems.

The changing face of paediatric diabetes

The purpose of this review is to provide an update on the changing face of paediatric type 1 diabetes and type 2 diabetes. Paediatric diabetes is on the rise, with extensive research dedicated to understanding its pathophysiology, comorbidities and complications. As obesity continues to increase among all youth, differentiating between type 1 diabetes and type 2 diabetes has become increasingly difficult but remains important for optimising treatment, anticipating complications and predicting disease risk. Novel treatments are emerging, with the ultimate goal being to achieve glycaemic control, limit weight gain, improve quality of life and reduce comorbidities. In this review, we focus on updates regarding the epidemiology, clinical presentation, comorbidities and complications of paediatric type 1 diabetes and type 2 diabetes and conclude with current and emerging treatments.

Continuous subcutaneous insulin infusion and flash glucose monitoring in diabetic hemiballism-hemichorea

A 71-year-old man without previous history of diabetes was hospitalized after suffering polyuria for 1 month and involuntary movement of the left arm for 1 week. His random serum glucose was 42.05 mmol/l and his hemoglobin A1C was 14% (129 mmol/mol). His serum osmolarity was normal and his urine ketone was negative. Cerebral CT revealed hyperdensity in the right basal ganglia. The patient was diagnosed with diabetic hemiballism-hemichorea (HH). Intravenous insulin was given and later shifted to continuous subcutaneous insulin infusion. During the hospital stay, insulin titration was guided mainly by flash glucose monitoring (FGM). Finger-prick glucose was occasionally checked to verify the accuracy of the FGM. Rapid correction of severe hyperglycemia was achieved without hypoglycemia. HH resolved within 1 week after euglycemia was achieved. This case emphasized the importance of being alert for HH as the initial presentation of diabetes and neuroimaging negative diabetic HH. In addition, interstitial glucose-monitoring technologies including continuous glucose monitoring and FGM can facilitate inpatient intensive insulin therapy in diabetic HH by avoiding hypoglycemia.