Abstract
Introduction
Throughout the consecutive articles on continuous glucose monitoring (CGM) over the years, we have endeavored to demonstrate the maturity of this treatment modality for routine clinical use (1 –3). In this article, we can rest our case: the American Diabetes Association (ADA) practical guidelines update from July 2019 firmly places CGM among routine options with an overwhelming high‐quality evidence for its use. Importantly, an international consensus report on clinical targets for CGM data, formally endorsed by the ADA, American Association of Clinical Endocrinologists, American Association of Diabetes Educators, European Association for the Study of Diabetes, Foundation of European Nurses in Diabetes, International Society for Pediatric and Adolescent Diabetes, JDRF, and Pediatric Endocrine Society, as well as individuals with diabetes outside of the medical profession, proposed easy to understand time in range (TIR) targets, along with time below range (TBR) and time above range (TAR) targets for day‐to‐day diabetes management (4). After several months of preparations, including reanalysis of existing published data, the consensus participants agreed on the targets for type 1 and type 2 diabetes presented in Table 1. (At least 10 full days of CGM data are needed for reliable recommendations for individuals with diabetes.)
Targets for Assessment of Glycemic Control: Type 1/Type 2 and Older/High-Risk Individuals
Each incremental 5% increase in time in range is associated with clinically significant benefits for type 1/type 2 diabetes.
For age <25 yr., if the glycated hemoglobin goal is 7.5% then set time in range target to approximately 60%, and aim at incremental improvements towards the target.
“Older and/or high-risk individuals with diabetes” targets should be agreed upon in individual context.
Adapted from Battelino et al. Diabetes Care 2019;
The consensus participants contacted professional associations for pregnancy complicated with diabetes, and with their input agreed also on targets for pregnancy complicated with type 1 diabetes (T1D), which are presented in Table 2. The consensus participants also agreed on using the Ambulatory Glucose Profile as the standard way for presentation of the CGM data, which will greatly facilitate communication between the healthcare professionals and individuals with diabetes. This international consensus on TIR targets is being implemented by numerous national diabetes associations and was a welcome development for many groups of people with diabetes.
Targets for Assessment of Glycemic Control: Pregnancy
Each incremental 5% increase in time in range is associated with clinically significant benefits for pregnancy, type 1.
Glucose levels are physiologically lower during pregnancy.
Adapted from Battelino et al. Diabetes Care 2019;
The current article discusses some of the latest evidence for the use of CGM and intermittently scanned continuous glucose monitoring (isCGM), with an addition of a small selection of relevant self‐blood glucose monitoring data.
Key Articles Reviewed for the Article
Shah VN, DuBose SN, Li Z, Beck RW, Peters AL, Weinstock RS, Kruger D, Tansey M, Sparling D, Woerner S, Vendrame F, Bergenstal R, Tamborlane WV, Watson SE, Sherr J
Bosi E, Choudhary P, de Valk HW, Lablanche S, Castañeda J, de Portu S, Da Silva J, Ré R, Vorrink‐de Groot L, Shin J, Kaufman FR, Cohen O; SMILE Study Group
Zhu J, Volkening LK, Laffel LM
Burckhardt MA, Roberts A, Smith GJ, Abraham MB, Davis EA, Jones TW
Dovc K, Cargnelutti K, Sturm A, Selb J, Bratina N, Battelino T
Piona C, Dovc K, Mutlu GY, Grad K, Gregorc P, Battelino T, Bratina N
Tyndall V, Stimson RH, Zammitt NN, Ritchie SA, McKnight JA, Dover AR, Gibb FW
Yaron M, Roitman E, Aharon‐Hananel G, Landau Z, Ganz T, Yanuv I, Rozenberg A, Karp M, Ish‐Shalom M, Singer J, Wainstein J, Raz I
Halbron M, Bourron O, Andreelli F, Ciangura C, Jacqueminet S, Popelier M, Bosquet F, Rouanet S, Amouyal C, Hartemann A
Christiansen MP, Klaff LJ, Bailey TS, Brazg R, Carlson G, Tweden KS
Murphy HR, Feig DS, Sanchez JJ, de Portu S, Sale Aon behalf of CONCEPTT Collaborative Group
Klonoff DC, Parkes JL, Kovatchev BP, Kerr D, Bevier WC, Brazg RL, Christiansen M, Bailey TS, Nichols JH, Kohn MA
Weinstock RS, Braffett BH, McGuigan P, Larkin ME, Grover NB, Walders‐Abramson N, Laffel LM, Chan CL, Chang N, Schwartzman BE, Barajas RA, Celona‐Jacobs N, Haymond MW; for the TODAY Study Group
Ardilouze A, Bouchard P, Hivert MF, Simard C, Allard C, Garant MP, Ménard J, Ouellet A, Houde G, Pesant MH, Baillargeon JP, Ardilouze JL
Continuous glucose monitoring profiles in healthy nondiabetic participants: a multicenter prospective study
Shah VN1, DuBose SN2, Li Z2, Beck RW2, Peters AL3, Weinstock RS4, Kruger D5, Tansey M6, Sparling D7, Woerner S8, Vendrame F9, Bergenstal R10, Tamborlane WV11, Watson SE12, Sherr J11
1Barbara Davis Center for Diabetes, Aurora, CO; 2Jaeb Center for Health Research, Tampa, FL; 3Keck School of Medicine of the University of Southern California, Los Angeles, CA; 4SUNY Upstate Medical University, Syracuse, NY; 5Henry Ford Medical Center, Detroit, MI; 6University of Iowa, Iowa City, IA; 7University of Oklahoma Health Sciences Center, Oklahoma City, OK; 8Indiana University School of Medicine, Indianapolis, IN; 9University of Miami, Miami, FL; 10International Diabetes Center Park Nicollet, Minneapolis, MN; 11Yale School of Medicine, New Haven, CT; 12University of Louisville, Louisville, KY
Aims
The use of CGM is increasing for patients with diabetes who require insulin. Data on glycemic profiles of healthy, nondiabetic individuals exist for older sensors, but there are very few evaluations of glycemic metrics with new‐generation CGM devices. This study aims to establish reference sensor glucose ranges using a current generation CGM sensor in healthy, nondiabetic individuals across different age groups.
Methods
A multicenter, prospective study was conducted at twelve centers within the T1D Exchange Clinic Network. Study participants were nonpregnant, healthy, nondiabetic children (age ≥6 years) and adults with non‐obese BMI. Participants wore a blinded Dexcom G6 CGM, for up to 10 days, with once‐daily calibration. CGM metrics of mean glucose, hyperglycemia, hypoglycemia, and glycemic variability were calculated.
Results
In all, 153 participants aged 7–80 years were included in the analyses. For all age groups except those over 60 years, mean average glucose was 98–99 mg/dL (5.4–5.5 mmol/L). Mean average glucose for those older than 60 years was 104 mg/dL (5.8 mmol/L). The median time in range 70 to 140 mg/dL (3.9–7.8 mmol/L) was 96% (interquartile range, 93–98). Mean within‐individual coefficient of variation was 17±3%, and median percentage of time spent with glucose level >140 mg/dL was 2.1% (30 min/day). Median time spent with glucose levels <70 mg/dL (3.9 mmol/L) was 1.1% (15 min/day).
Conclusions
Normative sensor glucose data have been derived for a healthy, nondiabetic population including a wide array of age groups. This information will be useful as a benchmark for future research studies.
This is an extremely important study, complementing CGM data for preschool children (5) with a current latest‐generation sensor. Knowledge on continuous glucose concentrations in healthy populations is crucial for setting goals in different populations with diabetes, as well as for better understanding the physiology of glucose metabolism in different age groups. Mean CGM glucose concentration was remarkably consistent around 99 mg/dL (5.5 mmol/L), with a very low coefficient of variation (CV) around 17%. Around 28% of all participants had >1 hypoglycemic event <70 mg/dL (<3.9 mmol/L), with a mean duration of around 1 hour. This clearly indicated that glucose excursions <70 mg/dL (3.9 mmol/L) are physiologic and that overzealous avoidance of low‐end normoglycemia in individuals with diabetes may indeed be exaggerated. However, levels <54 mg/dL (<3.0 mmol/L) were rare and do not seem to be physiologic. Importantly, values >180 mg/dL (10 mmol/L) were as rare as values <54 mg/dL (3.0 mmol/L); it is curious why so many diabetologists, certified diabetes educators, and individuals with diabetes so lightheartedly accept values >180 mg/dL (10 mmol/L) in day‐to‐day diabetes management. Such an attitude may well need a more thorough consideration, particularly in the view of recent data strongly linking hyperglycemia to vascular dementia (6).
Efficacy and safety of suspend‐before‐low insulin pump technology in hypoglycaemia‐prone adults with type 1 diabetes (SMILE): an open‐label randomised controlled trial
Bosi E1, Choudhary P2, de Valk HW3, Lablanche S4, Castañeda J5, de Portu S6, Da Silva J6, Ré R6, Vorrink‐de Groot L6, Shin J7, Kaufman FR7, Cohen O6; SMILE Study Group
1Diabetes Research Institute, IRCCS San Raffaele Hospital and San Raffaele Vita Salute University, Milan, Italy; 2Department of Internal Medicine, King's College Hospital, Diabetes Research Group, Weston Education Center, London, UK; 3Department of Internal Medicine, University Medical Center, Utrecht, The Netherlands; 4Department of Diabetology, Grenoble University Hospital, Service d'Endocrinologie, Pôle Digidune, Grenoble, France; 5Bakken Research Centre, Maastricht, The Netherlands; 6Medtronic International Trading Sàrl, Tolochenaz, Switzerland; 7Medtronic, Northridge, CA
This manuscript is also discussed in the article on Insulin Pumps, page S‐17.
Aims
Hypoglycemia unawareness and severe hypoglycemia are complications of diabetes that can increase both the fear of hypoglycemia and the risk of subsequent hypoglycemic events in patients with diabetes. This study assessed the safety and effectiveness of insulin pump therapy with integrated CGM and a suspend‐before‐low feature (Medtronic MiniMed 640G with SmartGuard) in hypoglycemia‐prone adults with T1D.
Methods
The SMILE study was an open‐label randomized controlled trial carried out in patients from 16 centers in Canada, France, Italy, the Netherlands, and the UK. Participants were adults aged 24–75 years with T1D for 10 years or longer who had glycated hemoglobin (HbA1c) levels of 5.8%–10.0% (40–86 mmol/mol) and were at high risk of hypoglycemia [recent severe hypoglycemia or hypoglycemia unawareness (Clarke or Gold score ≥4)]. After a 2‐week baseline run‐in phase, patients were randomly assigned to either control group or intervention group. The control group used the MiniMed 640G pump (continuous subcutaneous insulin infusion [CSII]) with self‐monitoring of blood glucose; the intervention group used the MiniMed 640G system with the suspend‐before‐low feature enabled. Assigned treatments and data collection ran for 6 months. Due of the nature of the intervention, neither participants nor treating healthcare providers could be masked to group assignment. However, the study statistician who performed data analysis was masked to the group assignments until final database lock. Mean number of sensor hypoglycemic events (blood glucose 55 mg/dL, 3.1 mmol/L) or lower was the primary outcome and was analyzed on an intention‐to‐treat basis in all randomly assigned participants. This trial is registered with ClinicalTrials.gov number NCT02733991.
Results
Between Dec 7, 2016, and March 27, 2018, 153 participants with a mean age 48.2±12.4 years were randomly assigned to the control group (n=77 years; mean age 47.4±12.5 years) or the intervention group (n=76; mean age 49.0±12.2 years). The intervention group had significantly fewer hypoglycemic events per participant per week during the trial (1.1±1.2 vs 4.1±3.4 mean events, model‐based treatment effect −2.9 [95% CI −3.5 to −2.3]; P<0.0001) and experienced fewer severe hypoglycemic events (defines as situations requiring third‐party assistance with administration of carbohydrate or glucagon or requiring other resuscitative actions) overall (3 vs 18 compared with control; P=0.0036). Hypoglycemia and hypoglycemia were the most common adverse event, with hypoglycemia observed in 33% (10/77) of participants in the control group vs 5% (4/76) in the intervention group and hyperglycemia seen in 9% (7/77) of control and 9% (7/76) of intervention group patients. No serious adverse device effects or episodes of diabetic ketoacidosis (DKA) were reported.
Conclusions
The frequency of sensor hypoglycemic and severe hypoglycemic events in hypoglycemia‐prone adults was decreased in those using insulin pump therapy with integrated CGM and a suspend‐before‐low feature compared with use of CSII without real‐time CGM. Based on these results, this technology could be beneficial when used in this high‐risk population.
After the first randomized controlled trial with the suspend‐before‐low feature in a pediatric population demonstrated significant reduction in the number of hypoglycemic events (7), this trial added principal data for the hypoglycemia‐prone adult population. Importantly, severe hypoglycemia was also significantly reduced by 2.9‐fold. Interestingly, time in range (TIR) 70–180 mg/dL (3.9–10.0 mmol/L) was 57.8% in the control group and only 59.9% in the intervention group, a difference of probably marginal clinical relevance. The time >180 mg/dL (10.0 mmol/L) was 33.2% in the control group and even higher 37.3% in the intervention group (the difference was larger during the day), with time >240 mg/dL (13.3 mmol/L) at 13.1% in both groups, and time >300 mg/dL (16.7 mmol/L) at 3.8% and 3.5%. These times in hyperglycemia were considerably above the current recommendations (4), clearly demonstrating that additional features like hybrid closed‐loop are needed for an improvement in TIR without a high percentage of time above range (TAR).
Distinct patterns of daily glucose variability by pubertal status in youth with type 1 diabetes
Zhu J1,2, Volkening LK1, Laffel LM1,2
1Pediatric, Adolescent, and Young Adult Section, Section on Clinical, Behavioral and Outcomes Research, Joslin Diabetes Center, Boston, MA; 2Division of Endocrinology, Department of Pediatrics, Boston Children's Hospital, Boston, MA
Aims
This study investigated glycemia and glucose variability metrics in youth with T1D, and to assess patterns of 24‐h glucose variability according to pubertal status.
Methods
Participants included 107 youth (age 8–17 years, 50% female, 95% white) with T1D for ≥1 year. Glycemia, glucose variability, and glucose pattern were assessed using 4 weeks of CGM data. Glucose values per hour were expressed as percentages relative to the mean 24‐hour glucose for a 4‐week period. Glucose data were compared across pubertal status: prepubertal (Tanner stage T1), pubertal (T2–4), and postpubertal (T5), and HbA1c categories (<7.5% [<58 mmol/mol], ≥7.5% [≥58 mmol/mol]).
Results
Youth participants had a mean±SD age of 13.1±2.6 years, diabetes duration of 6.3±3.5 years, and HbA1c of 7.8±0.8% (62±9 mmol/mol); 88% of participating patients received treatment through a pump. A higher mean glucose level (86±12 mg/dL [4.8±0.7 mmol/L]; P=0.01) and coefficient of variation (CV) (43±5%; P=0.06) was observed in prepubertal youths than in pubertal (SD 79±13 mg/dL [4.4±0.7 mmol/L]; CV 41±5%) and postpubertal (SD 77±14 mg/dL [4.3±0.8 mmol/L]; CV 40±5%) youths. Over 24 h, prepubertal individuals had the largest glucose excursions and the highest CV across most hours versus pubertal and postpubertal individuals. Across all participants, CV was strongly correlated with the percentage of glucose TIR (<70 mg/dL, <3.9 mmol/L; r=0.79, P<0.0001).
Conclusions
Prepubertal participants with T1D had more glucose variability independent of HbA1c than did pubertal and postpubertal participants. HbA1c alone does not capture the full range of glycemic parameters, which highlights the added insight provided by the use of CGM in managing youth with T1D.
This very important study demonstrated that youth spent on average more than 50% of the total time above range (TAR >180 mg/dL, 10 mmol/L). This is double the recently recommended amount of time (4) and will likely need additional attention in the future. Furthermore, TAR >250 mg/dL (16.5 mmol/L) was on average at 25%, which is five times more than recently recommended and very likely clinically relevant. With the sensor accuracy demonstrated acceptable also in pediatric population (8), specific efforts should clearly be designed and implemented to increase TIR and reduce TAR, particularly in the view of demonstrated association of hyperglycemia with permanent changes in the central nervous system. The study also confirmed the well‐established correlation between TBR <70 mg/dL (3.9 mmo/L) and CV. Despite the fact that TBR <70 mg/dL (3.9 mmo/L) was well within the recommended target, CV was very high due to excessive hyperglycemic excursions. Interestingly, measured outcomes were worse in the prepubertal participants as compared with pubertal or postpubertal, independent of HbA1c. Prepubertal individuals with diabetes mostly depend on their caregivers at home, in kindergartens and schools. Implementation of the International Society for Pediatric and Adolescent Diabetes (ISPAD) guidelines for diabetes management in schools (9,10) seems paramount.
The use of continuous glucose monitoring with remote monitoring improves psychosocial measures in parents of children with type 1 diabetes: a randomized crossover trial
Burckhardt MA1–3, Roberts A1,2, Smith GJ1, Abraham MB1–3, Davis EA1–3, Jones TW1–3
1Children's Diabetes Centre, Telethon Kids Institute, The University of Western Australia, Perth, Australia; 2Division of Paediatrics, The University of Western Australia School of Medicine, Perth, Australia; 3Department of Endocrinology and Diabetes, Perth Children's Hospital, Perth, Australia
Aims
Hypoglycemia and fear of hypoglycemia can limit the ability of pediatric T1D patients to attain optimal glycemic control. This study explored the effect of CGM with remote monitoring on psychosocial outcomes in parents of children with T1D.
Methods
Children aged 2–12 years with T1D were studied, along with their parents, in a randomized crossover study. Participants in the control group used conventional blood glucose monitoring for 3 months, while the intervention group used the Dexcom G5 Mobile CGM system with remote monitoring for 3 months. The parental fear of hypoglycemia score, assessed using the Hypoglycemia Fear Survey, was the primary outcome.
Results
Childrens' use of CGM with remote monitoring was associated with lower Parental Hypoglycemia Fear Survey scores (P<0.001). Parental health‐related quality of life as well as family functioning, stress, anxiety, and sleep measures also improved significantly after intervention.
Conclusions
The use of CGM with remote monitoring by children with T1D was found to improve multiple measures of quality of life, reduce family stress, and to improve parental sleep.
Forty‐nine parents included in the final analysis expressed a fascinating improvement in the quality of life (QOL) over a wide range of investigated parameters. This is the first study in a pediatric population investigating QOL measures as a primary outcome and the first to demonstrate this significant improvement. Interestingly, mean metabolic control was suboptimal and did not improve with the CGM use; however, the TIR metrics were not available for the control arm, and therefore no comparison of TIR metrics was reported. Additionally, the authors state that no specific CGM‐related intervention was used, which may explain no improvement in mean metabolic control. Reducing the burden of families with children with T1D seems to be a very difficult task and results from this study strongly support the use of CGM, possibly with the option of remote control. As reimbursement and accessibility of CGM remain major obstacles even in affluent environments (11), advocacy should use more powerful means including social media to help individuals with diabetes and their families reduce the disease burden, hopefully resulting also in better metabolic outcomes.
Continuous glucose monitoring use and glucose variability in pre‐school children with type 1 diabetes
Dovc K1,2, Cargnelutti K3, Sturm A4, Selb J5, Bratina N1, Battelino T1,4
1Department of Pediatric Endocrinology, Diabetes and Metabolic Diseases, University Children's Hospital, University Medical Centre, Ljubljana, Slovenia; 2Wellcome Trust‐MRC Institute of Metabolic Science, University of Cambridge, Cambridge, UK; 3Faculty of Medicine and Surgery, University of Udine, Udine, Italy; 4Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia; 5University Clinic of Respiratory and Allergic Diseases Golnik, Kranj, Slovenia
Aims
The correlation between CGM use and glucose variability was evaluated in preschoolers with T1D.
Methods
A nationwide population‐based cohort study was conducted to measure the difference in glucose variability between those periods during which participants were using CGM versus periods during which they used CGM over the course of 5 years, using data from the Slovenian National Registry.
Results
Data were collected for a total of 40 children under 8 years of age who were followed for an observational period of approximately 116 patient‐years. Mean age at CGM initiation was 3.5 (±1.7) years. Standard deviation of mean glucose level was lower during intervals of CGM use [3.6 mmol/L (3.2–3.9) with CGM and 4.3 mmol/L (3.8–4.7) without CGM, P < 0.001]. The same was true for coefficient of variation [44.0% (40.4–47.0) with CGM and 46.1% (42.3–49.4) without CGM, P=0.021]. Frequent use of CGM (>5 days/week) was associated with a 0.4% (4.4 mmol/mol) decrease in HbA1c (7.6% compared with 7.2%, P=0.047).
Conclusions
The results of this population cohort study showed that the use of CGM was associated with reduced glucose variability during a 5‐year follow‐up period among young patients with T1D.
Several randomized controlled trials failed to demonstrate efficacy of CGM in young children for various reasons, with low sensor utilization and excessive hypoglycemia avoidance usually being the most prominent. This retrospective national cohort analysis over 5 years focused on comparison between periods when CGM was utilized and periods when only self‐monitoring of blood glucose (SMBG) data were used for diabetes management. Frequent use of CGM was associated with significantly better metabolic outcomes. When CGM parameters from this study are compared with the recent report from Zhu and colleagues (2019) on prepubertal children, the mean glucose (163.8 mg/dL [9.1 mmol/L] vs. 202 mg/dL [11.2 mmol/L]), SD (59.4 mg/dL [3.3 mmol/L] vs. 86 mg/dL [4.8 mmol/L]), and CV (36.0% vs. 43%) are considerably better in the former. It is not possible to directly compare these values as the methodology of the two reports is different, however, the data from this retrospective cohort study demonstrate that frequent use of CGM can significantly improve metabolic outcomes also in preschool children with T1D with parameters for GV close to the proposed target (12).
Non‐adjunctive flash glucose monitoring system use during summer‐camp in children with type 1 diabetes: the free‐summer study
Piona C1, Dovc K2, Mutlu GY3, Grad K4, Gregorc P4, Battelino T2,4, Bratina N 2
1Pediatric Diabetes and Metabolic Disorders Unit, Regional Center for Pediatric Diabetes, University City Hospital, Verona, Italy; 2Department of Paediatric Endocrinology, Diabetes and Metabolic Diseases, University Children's Hospital, University Medical Centre, Ljubljana, Slovenia; 3Department of Pediatrics, Koç University Hospital, İstanbul, Turkey; 4Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia
Aims
A factory‐calibrated sensor for intermittently scanned continuous glucose monitoring (isCGM) is accurate and safe in children with T1D. Data regarding the effectiveness of isCGM in place of SMBG in this population is scarce. This study assessed the nonadjunctive use of isCGM in children with T1D during 2 weeks in a challenging summer‐camp setting.
Methods
A two‐arm, parallel, randomized, outpatient clinical trial was carried out in 46 children (n=25 female) aged 11.1±2.6 years, with HbA1c 7.4%±0.7%. In the isCGM (intervention) group, all participants (n=26) were blinded for SMBG, and insulin dosing was isCGM‐based; in the control group, participants (n=20) were blinded for isCGM and performed SMBG‐based insulin dosing. Between‐group difference in the proportion of TIR 70–180 mg/dL (3.9–10 mmol/L) was the primary outcome.
Results
No significant difference was found between the two groups for TIR. A significant reduction in time spent above 10 mmol/L (P<0.05) was found in patients with suboptimal metabolic control (HbA1c > 7%), and an improvement in TIR (P=0.05) was observed in the isCGM group versus the SMBG group. There were no severe hypoglycemic or serious adverse events during the study. The mean absolute relative difference (MARD) overall between isCGM and SMBG was 18.3%, and the median absolute relative difference was 8%. Consensus error grid analysis showed 82.2% in zone A and 95.2% of results in zone A + B.
Conclusions
In this subpopulation of children with T1D and suboptimal glycemic control, the nonadjunctive use of isCGM was found to be as safe and effective as SMBG, and isCGM was found to reduce time spent in hyperglycemia.
This small randomized controlled trial was performed during a two‐week summer camp at the Mediterranean Sea, where children with diabetes swim, play games, and compete in sports several times during the day. Despite this challenging condition the TBR <54 mg/dL (<3.0 mmo/L) was 1.3%, close to the recent recommendations, without a hypoglycemia requiring glucagon or i.v. glucose. TIR, however, was only at 50%. isCGM significantly improved TIR and TAR in those with higher HbA1c. Routinely, isCGM was used for insulin dosing; the >2 mmol (36 mg/dL) discrepancy predefined in the study protocol was noticed on 42 occasions (1.5%), and on this occasion, insulin dosing was based on the SMBG rather than on the isSGM. A larger non‐randomized single arm study in children demonstrated a significant improvement of TIR by 1 hour/day, with reduced TAR >180 mg/dL (10 mmol/L) by 1.2 hours and no change in TBR (13); importantly the Diabetes Treatment Satisfaction Questionnaire “Total Treatment Satisfaction” score improved for parents (P<0.0001) and teenagers (P<0.0001).
Marked improvement in HbA1c following commencement of flash glucose monitoring in people with type 1 diabetes
Tyndall V1, Stimson RH2,3, Zammitt NN2, Ritchie SA1, McKnight JA1, Dover AR2, Gibb FW2,3
1Edinburgh Centre for Endocrinology and Diabetes, Western General Hospital, Edinburgh, UK; 2Edinburgh Centre for Endocrinology and Diabetes, Royal Infirmary of Edinburgh, Little France Crescent, Edinburgh, UK; 3Centre for Cardiovascular Science, University of Edinburgh, Edinburgh, UK
Aims
There is very little evidence supporting the effectiveness of flash monitoring in lowering HbA1c. We sought to assess the impact of introducing flash monitoring in individuals with T1D.
Methods
A prospective observational study was conducted in 900 individuals with T1D to assess change in HbA1c following flash monitoring compared with a group of 518 patients with T1D who did not use flash monitoring. The primary outcome was change in HbA1c—the difference between initial HbA1c (prior to commencement of any flash monitoring) and the next available value after the flash monitoring education session. Secondary outcomes included changes in hypoglycemia, quality of life, flash monitoring data, and hospital admissions.
Results
Participants with starting HbA1c level ≥58 mmol/mol (7.5%) achieved a median change in HbA1c of −7 mmol/mol (interquartile range [IQR] −13 to −1; −0.6%, IQR −1.2% to −0.1%; P<0.001). The percentage of subjects achieving HbA1c <58 mmol/mol increased from 34.2% to 50.9% (P<0.001). Median follow‐up was 245 days (IQR 182 to 330). Participants who did not use flash monitoring experienced no change in HbA1c across a similar length of time (P=0.508). Some patient characteristics, including higher HbA1c (P<0.001), younger age at diagnosis (P=0.003), and lower social deprivation (P=0.024), were independently associated with an HbA1c fall of ≥5 mmol/mol (0.5%). After flash monitoring, more symptomatic (OR 1.9, P < 0.001) and asymptomatic (OR 1.4, P<0.001) hypoglycemia was reported. Regimen‐related and emotional components of the diabetes distress scale improved after flash monitoring, although scores increased for the proportion of subjects with elevated anxiety (OR 1.2, P=0.028) and depression (OR 2.0, P<0.001). The number of blood glucose test strips used per day fell from 3.8 to 0.6 (P<0.001). DKA admissions fell significantly following flash monitoring (P=0.043).
Conclusions
Flash monitoring is associated with significant improvements in HbA1c levels and fewer hospital admissions for DKA. Higher rates of hypoglycemia may relate to greater recognition of previously unrecognized events. The effect of flash monitoring on quality of life parameters was mixed, but overall treatment satisfaction was overwhelmingly positive.
The unprecedented success of isCGM use in the United Kingdom is a landmark achievement essentially attributable to the devoted advocacy and professional determination of healthcare professionals tired of the unreasonable attitudes of the National Health Service England (NHS): overwhelming, high‐quality evidence from randomized controlled trials (RCTs) supporting the use of CGM and isCGM was not sufficient for NHS to grant reimbursement. Consorted action of several professional associations and people with diabetes under the lead of @ParthasKar on social media resulted in a so‐far limited reimbursement for isCGM with immediate huge clinical impact: all parameters of metabolic control along with quality of life parameters improved with the use of isCGM. Recently, a series of publication demonstrated different aspects of metabolic improvement with several thousand individuals with T1D benefiting from the isCGM use. This report form Edinburgh reports a swift clinically meaningful improvement of metabolic control within an average follow‐up of 8 months. A concomitant increase in hypoglycemia was likely due to previously unrecognized hypoglycemia reported in several UK trials. Number of daily scans and both final HbA1c (r=−0.255, P<0.001) and change in HbA1c (r=−0.279, P<0.001) were strongly correlated. Additionally, there was a significant reduction in admissions for diabetic ketoacidosis (DKA), falling from 10 to 2 episodes (P=0.043) in the 6 months following NHS‐funded isCGM use (compared with the corresponding period 2 years earlier). Although user satisfaction was very high, there were potentially issues related to hypoglycemia detection and psychological distress, at least in the early phase of treatment intensification. Higher depression scores may be related to the initial visualization of how poor the glycemic control really was. Indeed, an RCT comparing the use of isCGM with or without education demonstrated a significant effect of isCGM education compared with the control group receiving no isCGM education, reflecting also in significant improvements in time spent in the target glucose range, in diabetes‐related distress scores and in satisfaction with the glucose monitoring method when isCGM education was provided (14).
Effect of flash glucose monitoring technology on glycemic control and treatment satisfaction in patients with type 2 diabetes
Yaron M1,2, Roitman E1, Aharon‐Hananel G1, Landau Z1,2,3, Ganz T3, Yanuv I4, Rozenberg A4, Karp M1, Ish‐Shalom M1,2, Singer J1,2, Wainstein J1,2,3, Raz I1,4
1Diabetes Medical Center, Tel Aviv, Israel; 2Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel; 3Diabetes Unit, E. Wolfson Medical Center, Holon, Israel; 4Hebrew University, Jerusalem, Israel
Aims
Treatment satisfaction and the effectiveness of a flash glucose monitoring (isCGM) system was investigated in patients with type 2 diabetes (T2D) using insulin.
Methods
Individuals with T2D who had been on multiple daily insulin injections (MDI) for at least 1 year were included in the study. A total of 101 participants were randomly assigned to the isCGM intervention group (n=53) or the standard care (control) group (n=48) and followed for 10 weeks. Adjust to insulin doses were made in face‐to‐face and telephone visits for both groups. Satisfaction with treatment, quality of life, comfort using isCGM, HbA1c, and frequency of hypoglycemic events were all evaluated.
Results
Treatment was deemed significantly more flexible by members of the intervention group (P=0.019), and these individuals would recommend it to their counterparts (P=0.023). Satisfaction ratings for the isCGM system were high. HbA1c decreased more in the intervention group [−0.82% (9 mmol/mol)] then in the control group [−0.33% (3.6 mmol/mol)] (P=0.005). In nonprespecified post hoc analysis, 68.6% of individuals in the intervention group saw their HbA1c reduced by ≥0.5% (5.5 mmol/mol) versus 30.2% in the control group (P<0.001), and 39.2% in the intervention group experienced a reduction of ≥1.0% (10.9 mmol/mol) versus 18.6% in the control group (P=0.0023) without an increased frequency of hypoglycemia.
Conclusions
isCGM tends to improve treatment satisfaction and may lead to improved glycemic control in individuals with T2D on MDI without increasing the frequency of hypoglycemia.
Similar to previous reports on CGM in T2D (15), isCGM is—not surprisingly—also effective and improves treatment satisfaction. Marked improvement in HbA1c characteristically came without any increase in hypoglycemia parameters. Also, in this diverse population with T2D using very different therapeutic combinations, significantly more participants achieved target HbA1c with the use of isCGM, particularly if the starting HbA1c was high, confirming the now well‐proven fact that individuals with higher HbA1c may benefit most from diabetes technology in relation to managing HbA1c. With accumulating high‐quality data on the beneficial effects of CGM and isCGM in individuals with T2D using MDI, different insurance systems are likely to expand reimbursement of CGM and isCGM to individuals with T2D on MDI in order to improve outcomes and reduce cost.
Insulin pump combined with flash glucose monitoring: a therapeutic option to improve glycemic control in severely nonadherent patients with type 1 diabetes
Halbron M1,2, Bourron O1,3, Andreelli F1,3, Ciangura C1, Jacqueminet S1, Popelier M1, Bosquet F1, Rouanet S4, Amouyal C1, Hartemann A1,3
1Assistance Publique‐Hôpitaux de Paris, Diabetes Department, Hôpital Pitié‐Salpêtrière, Paris, France; 2Assistance Publique‐Hôpitaux de Paris, Department of Pharmacology, INSERM, Paris, France; 3Assistance Publique‐Hôpitaux de Paris, Diabetes Department, Hôpital Pitié‐Salpêtrière, Sorbonne Université, Paris, France; 4Statistician Unit, StatEthic SASU, Levallois‐Perret, France
Aims
Some individuals with T1D have severe difficulties achieving target metabolic control; they rarely perform self‐blood glucose measures and miss insulin injections. Their HbA1c is far above the target rate. Current guidelines do not recommend starting treatment with an insulin pump CSII for these persons. The aim of this study was to determine whether a CSII associated with a flash glucose monitoring (isCGM) device could reduce HbA1c without increasing the risk of acute events, DKA, and severe hypoglycemia (SH).
Methods
We conducted a 6‐month nonrandomized, pilot prospective study. Individuals with T1D on MDI who performed less than two self‐blood glucose tests/day and had an HbA1c >9% were equipped with CSII and an isCGM device. The primary composite endpoint was defined as a change in HbA1c ≥1% without any episode of DKA or SH during 6 months. Change in mean HbA1c, weight, treatment satisfaction, frequency of minor hypoglycemia, and ketoacidosis were secondary endpoints.
Results
Nineteen adults were included. Median (Q1–Q3) HbA1c at baseline was 10.8 (10.3–13.0), 14 participants did not perform any self‐monitoring, and 5 performed maximum two tests daily. Twelve participants (63%) (95% confidence interval 41%–81%) met the primary composite endpoint. Seventeen participants completed the study. HbA1c decreased by 2% (1.0–3.3) (P<0.001), and satisfaction with treatment significantly improved. Three participants experienced SH and one a DKA, versus, respectively, five and eight in the year preceding the study. Participants scanned the sensor 4 (3–6) times per day and injected 3 (2.7–4.1) boluses per day. Weight increased significantly. An association of an insulin pump with an isCGM device can be an effective and safe therapeutic option in individuals with severe difficulties in achieving glycemic targets, with high HbA1c.
This interesting pilot trial focused on the population with the most difficulties achieving glycemic targets and is therefore of particular importance. After 6 months of concomitant use of an insulin pump and isCGM, the HbA1c dropped by 1.5%, with a significant reduction in total daily insulin dose, marginal increase in weight by 1 kg, marginal increase in minor hypoglycemia, and a significant increase in treatment satisfaction without any increase in hypoglycemia fear. This together seems to be a remarkable attainment, defying prejudices against individuals with more challenges in accomplishing glycemic targets. The study clearly suggests that failure to do SMBG should not be a contraindication to CSII or CGM/isCGM technology. The willingness to modify day‐to‐day behavior is principal.
A prospective multicenter evaluation of the accuracy and safety of an implanted continuous glucose sensor: the PRECISION study
Christiansen MP1, Klaff LJ2, Bailey TS3, Brazg R2, Carlson G4, Tweden KS5
1Diablo Clinical Research, Walnut Creek, CA; 2Rainier Clinical Research Center, Inc., Renton, WA; 3AMCR Institute, Inc., Escondido, CA; 4Carlson Consulting, San Francisco, CA; 5Senseonics, Inc., Germantown, MD
Aims
A prior study (PRECISE II) demonstrated that an implantable continuous glucose monitoring (CGM) system (Eversense CGM System) provided accurate glucose readings through the 90‐day sensor life with a favorable safety profile in participants with T1D or T2D. This study was performed to further characterize the accuracy of the system.
Methods
PRECISION was a prospective multicenter study that evaluated the accuracy and safety of Eversense among adults with T1D or T2D through 90 days (NCT02647905). Accuracy measures included percentage system agreement and mean absolute relative difference (MARD) between Eversense and Yellow Springs Instrument reference measurements from 40 to 400 mg/dL. The primary safety endpoint was incidence of device‐related or sensor insertion/removal procedure‐related serious adverse events (SAEs) through 90 days. An updated glucose calculation algorithm was also applied to the sensor data from the PRECISE II study to evaluate consistency of accuracy results.
Results
Thirty‐five participants received the CGM system. Eighty‐five percent of CGM values were within 15/15% of reference and the MARD value against reference was 9.6% [95% confidence interval (CI) 8.9–10.4]. All sensors were functional through day 90. No device‐ or procedure‐related SAEs occurred. Application of the updated algorithm to PRECISE II sensor data resulted in 87% of readings within 15/15% of reference and an MARD value against reference of 8.5% (95% CI 8.0%–9.1%).
Conclusions
PRECISION corroborated prior accuracy and safety findings of the Eversense CGM System through the 90‐day sensor life. The updated algorithm improved accuracy of measurements in PRECISE II.
An implantable CGM is an important treatment option. Eversense CGM demonstrated robust accuracy data, corroborated in the recent study with MARD consistently below 10 in all glucose ranges. An international registry enrolled 3,023 Eversense users, with 5,417 inserted sensors (1,260 patient‐years [PYs] of follow‐up); 969 participants had used the system for at least 6 months and 173 patients had used the system for at least one year (16). The most frequently reported Eversense related AEs were sensor location site infection (0.96%; 2.46 events per 100 PYs), failure to remove the sensor at first attempt (0.76%; 1.90 events per 100 PYs), and adhesive area site irritation (0.66%; 1.59 events per 100 PYs). The full intended sensor life was achieved by 91% of 90‐day sensors and 75% of 180‐day sensors. International guidelines on the implementation and use of the implantable sensor were developed (17), suggesting best candidates and most successful technical approaches. Long‐term implantable CGM with no on‐body component may become the preferred CGM choice by individuals with diabetes if/when commercialized. With several long‐term technologies under clinical development, the CGM market may see considerable development in the foreseeable future.
Modelling potential cost savings from use of real‐time continuous glucose monitoring in pregnant women with type 1 diabetes
Murphy HR1,2, Feig DS4,5,6, Sanchez JJ7, de Portu S8, Sale A3 on behalf of CONCEPTT Collaborative Group
1Norwich Medical School, University of East Anglia, Norwich, UK; 2Women's Health Academic Centre, Division of Women's and Children's Health, King's College London, London, UK; 3Medtronic Ltd, Watford, UK; 4Mt Sinai Hospital, Toronto, Ontario, Canada; 5Lunenfeld‐Tanenbaum Research Institute, Toronto, Ontario, Canada; 6Department of Medicine, University of Toronto, Toronto, Ontario, Canada; 7Sunnybrook Research Institute, Toronto, Ontario, Canada; 8Medtronic International Trading Sàrl, Tolochenaz, Switzerland
Aims
This study investigated the potential for cost savings associated with using real‐time continuous glucose monitoring (RT‐CGM) throughout pregnancy in women with T1D.
Methods
A budget impact model was created from the perspective of NHS England to estimate the total costs attributable to managing pregnancy and delivery in women with T1D SMBG with and without RT‐CGM. It was assumed that the entire modelled cohort (n=1441) would use RT‐CGM from 10 to 38 weeks' gestation (7 months). Sources of data regarding pregnancy and neonatal complication rates and related costs included published literature, national tariffs, and device manufacturers.
Results
The cost of glucose monitoring was determined to be £588 with SMBG alone and £1820 with RT‐CGM. The total annual costs of managing pregnancy and delivery in women with T1D were £23.725.648 with SMBG alone and £14.165.187 with SMBG and RT‐CGM. These data indicate that the use of RT‐CGM could potentially result in a cost savings of approximately £9.560.461. Most of the cost savings could be attributed to the daily cost of neonatal intensive care unit (NICU) admissions (£3.743) and the shorter duration of NICU stays (mean 6.6 vs. 9.1 days respectively). Sensitivity analyses revealed that RT‐CGM continued to provide cost savings, albeit to lesser extents, across a range of NICU costs and lengths of hospital stays, and with varying numbers of daily SMBG measurements.
Conclusions
The primary source of the substantial cost‐savings associated with routine use of RT‐CGM by pregnant women with T1D are attributable to reductions in NICU admissions and shorter duration of NICU care.
This analysis provides a robust, simple, and straightforward proof of the fact that the use of CGM during pregnancy complicated by T1D is actually not expensive, as so commonly gossiped, but cost‐saving. An enormous amount of €10 million (U.S. $11 million) could be saved by using CGM during pregnancy in the UK alone. If the increased risks of serious pregnancy complications such as congenital abnormalities, stillbirth, and neonatal mortality, imposing particular clinical, societal, and financial burdens on healthcare systems were not sufficient to persuade public insurance systems to universally reimburse CGM during pregnancies complicated by diabetes, perhaps this money‐talk will.
Investigation of the accuracy of 18 marketed blood glucose monitors
Klonoff DC1, Parkes JL2, Kovatchev BP3, Kerr D4, Bevier WC4, Brazg RL5, Christiansen M6, Bailey TS7, Nichols JH8, Kohn MA9
1Mills‐Peninsula Medical Center, Diabetes Research Institute, San Mateo, CA; 2Joan Lee Parkes Consulting, Inc., Bristol, IN; 3Center for Diabetes Technology, University of Virginia, Charlottesville, VA; 4William Sansum Diabetes Center, Santa Barbara, CA; 5Rainier Clinical Research Center, Inc., Renton, WA; 6Diablo Clinical Research, Walnut Creek, CA; 7AMCR Institute, Inc., Escondido, CA; 8School of Medicine, Vanderbilt University, Nashville, TN; 9Department of Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, CA
Aims
Blood glucose monitoring systems (BGMs) for personal use must meet minimum requirements for accuracy set forth by regulatory agencies to ensure that the self‐monitoring of blood glucose to be safe and clinically meaningful. Inaccurate BGMs have the potential to put users at significant risk. Studies published in recent years have shown that the performance of several BGMs was suboptimal.
Methods
Eighteen BGMs cleared by U.S. Food and Drug Administration (FDA) and representing the majority of systems available in the United States between 2013 and 2015 were assessed for accuracy. Products were tested in three triple‐blinded studies. The model of the meter, person reading the BGMs, and the reference laboratory were all blinded. The study included 1,035 adult (≥ 18 years of age) patients with T1D, T2D, or no diabetes. Capillary blood glucose was tested on six different systems and a reference capillary sample was prepared for plasma testing at a reference laboratory for each of the study subjects. In each study, a BGM was tested on 115 participants. To be considered in compliance, a BGM's reading should be within 15% of a reference plasma value ≥100 mg/dL (5.55 mmol/L) and within 15 mg/dL (0.83 mmol/L) of a reference value below 100 mg/dL (5.55 mmol/L). Readings were categorized as compliant or noncompliant, and the percentage of compliant readings in each study was compared with a predetermined accuracy that was standard similar to, but more lenient than, current regulatory standards. Additional accuracy metrics were included in the overall compliance rate: the percentage of extreme outlier readings, defined as differing from the reference value by >20%; modified Bland–Altman analysis including average bias, coefficient of variation, 95% limits of agreement; and percentage of meter readings with no clinical risk as determined by the surveillance error grid.
Results
All tested BGMs produced accuracy metrics with nearly identical BGM rankings. Six of the 18 BGMs [Bayer Contour Next, Roche Accu‐Chek Aviva Plus and AccuChek SmartView, ARKRAY Walmart ReliOn Confirm (Micro), AgaMatrix CVS Advanced, and Abbott Diabetes Care Freestyle Lite] met the predetermined accuracy standard in all three studies. Five BGMs met the predetermined standard in two studies, and three BGMs met the standards in one study. Four BGMs failed to meet the accuracy standard in any of the three studies. Overall, ≥95% of data pairs that were compliant for every BGM in all three studies met the accuracy standard in all three studies. In the six top‐performing BGMs on other metrics, 97% of readings were within the “no risk” category; none of the other BGMs had 97% of readings in this category.
Conclusions
BGM performance levels reported in the study represent the performance of BGM products when used by trained medical professional, which does not necessarily reflect results of when BGMs are used by patients. Cleared BGMs do not always meet the level of analytical accuracy currently required for regulatory clearance. Of 18 meters tested, the 6 top‐performing BGMs met a predefined accuracy standard on three out of three studies, as well as performing the best according to four additional metrics. The results of this study could assist patients with diabetes, healthcare professionals, and insurers in choosing products and regulators when evaluating postmarket performance.
Postmarket surveillance of blood glucose meters (BGMs) demonstrated discrepancies between the accuracy of the off‐the‐shelf meters and strips and accuracy reported to the regulator in approval submission documentation. As there is currently no mandatory BGM postmarket surveillance, this accuracy study of FDA‐cleared BGMs using a consensus protocol created with input from the FDA is the largest ever reported in the literature, and provides very important information: only a third of the FDA cleared BGM met the accuracy standards. This postmarket deterioration in accuracy may be due to manufacturing errors, scale‐up problems, changes in components between strip lots, other production problems, or inappropriate shipping. Empirical and modeling studies reported adverse clinical and economic outcomes due to inaccuracies of BGMs (18); moreover, BGMs are used to calibrate automated insulin delivery systems like the commercially available hybrid closed loop. Finally, the accuracy of the BGMs may additionally deteriorate with the decreasing market volume, as BGMs are increasingly replaced by CGM. Special BGMs specifically designated and/or cleared for calibration/verification of CGM devices and automated insulin delivery systems may be needed in the future.
Self‐monitoring of blood glucose in youth‐onset type 2 diabetes: results from the TODAY study
Weinstock RS1, Braffett BH2, McGuigan P3, Larkin ME4, Grover NB2, Walders‐Abramson N5, Laffel LM6, Chan CL5, Chang N7, Schwartzman BE8, Barajas RA9, Celona‐Jacobs N5, Haymond MW10; for the TODAY Study Group
1State University of New York Upstate Medical University, Syracuse, NY; 2The Biostatistics Center, George Washington University, Rockville, MD; 3Case Western Reserve University, Rainbow Babies and Children's Hospital, Cleveland, OH; 4Massachusetts General Hospital Diabetes Center, Harvard Medical School, Boston, MA; 5University of Colorado School of Medicine, Aurora, CO; 6Joslin Diabetes Center, Harvard Medical School, Boston, MA; 7Children's Hospital of Los Angeles, Los Angeles, CA; 8Children's Hospital of Philadelphia, Philadelphia, PA; 9University of Texas Health Science Center San Antonio, San Antonio, TX; 10Baylor College of Medicine, Houston, TX
Aims
The utility of self‐monitoring of blood glucose (SMBG) in non‐insulin‐treated young individuals with T2D has not yet been evaluated. The aim of this exploratory analyses of Treatment Options for Type 2 Diabetes in Adolescents and Youth (TODAY) study results was to explore whether SMBG is associated with lower HbA1c in youth with T2D taking oral medications only or after starting insulin for persistently elevated HbA1c.
Methods
Study participants (n=699) of TODAY study on oral medications were asked to perform SMBG twice daily. Insulin glargine was started after reaching primary outcome (PO), which was HbA1c ≥8% (64 mmol/mol) over 6 months, or after an inability to wean from temporary insulin because of metabolic decompensation. HbA1c and proportion of days when the meter was used one or more times on ≥80% of days (SMBG%) before and after PO were analyzed.
Results
A total of 319 youth participants reached PO during the study, 298 of them started insulin, and 282 patients had SMBG data. At PO, 65.3% were female, mean age was 15.8±2.3 years, BMI 35.5±7.9 kg/m2, and HbA1c 9.6±2.0% (81±21.9 mmol/mol). Median SMBG% was 40.0% at PO (n=282), which increased to 49.0% 6 months after starting insulin therapy (n=181) and returned to 40.5% and 40.0% after 1 year (n=145) and 2 years (n=94) of insulin therapy, respectively. SMBG declined over time and was inversely related to HbA1c (P<0.0001). At 6 months and 1 year after insulin initiation, participants with SMBG ≥80% had a ≥1% reduction in HbA1c. HbA1c reduction for those checking at least twice daily was 1.5% and 2.5%, respectively, at 6 months and 1 year after starting insulin therapy. Participants who checked ≥80% were younger and had a lower a HbA1c, BMI, and blood pressure.
Conclusions
Low SMBG adoption in youth with T2D was common and associated with increased HbA1c. Further studies on optimal SMBG frequency in youth using or not using insulin are needed.
This TODAY survey basically demonstrates efficacy of SMBG if/when used by adolescents with T2D on oral medications. Despite the largely negative view on the SMBG use in adults with T2D treated without insulin, a recent meta‐analysis included 15 RCTs comprising 3,383 individuals with non‐insulin‐treated T2D and demonstrated that SMBG improved HbA1c levels in the short term (≤6‐month follow‐up) as well as in long term (≥12‐month follow‐up) use (19). As adolescent‐onset T2D often becomes a life‐long disease with devastating and costly chronic complications, the early adoption of SMBG or, for that, is CGM, seems prudent.
Self‐monitoring of blood glucose: a complementary method beyond the oral glucose tolerance test to identify hyperglycemia during pregnancy
Ardilouze A1, Bouchard P2, Hivert MF1,3,4, Simard C2, Allard C1, Garant MP1, Ménard J1, Ouellet A1,5, Houde G1,2, Pesant MH1,2, Baillargeon JP1,2, Ardilouze JL1,2
1Research Centre of the Centre Hospitalier Universitaire de Sherbrooke, Sherbrooke, Quebec, Canada; 2Endocrine Division, Department of Medicine, Université de Sherbrooke, Sherbrooke, Quebec, Canada; 3Department of Population Medicine, Harvard Pilgrim Health Care Institute, Landmark Center, MA; 4Diabetes Unit, Massachusetts General Hospital, MA; 5Department of Obstetrics and Gynecology, Université de Sherbrooke, Sherbrooke, Quebec, Canada
Aims
Gestational diabetes mellitus is a state of glucose intolerance of variable severity during pregnancy, commonly diagnosed using oral glucose tolerance test (OGTT). This study aimed to compare OGTT and SMBG in identifying gestational diabetes mellitus and other hyperglycemic statuses in pregnant women, glycemic status‐related pregnancy outcomes and perceived convenience of both diagnostic methods.
Methods
Pregnant women with a glucose level equal or above 7.2 mmol/L after screening 50 OGTT at 24 to 28 weeks of gestation and singleton pregnancy were asked to participate in this prospective study. Participants underwent blinded 75 g OGTT at day 1, followed by a week of SMBG with four daily fasting and 2‐hour postprandial measurements, without modifying diet or lifestyle. Gestational diabetes mellitus was diagnosed according to current International Association of the Diabetes and Pregnancy Study Groups recommendations, while pregnancy hyperglycemia was defined as ≥4/7 glucose values ≥5.3 after fasting or ≥6.7 mmol/L 2 h after any meal of the day with equal management for both states.
Results
Enrolled participants (n=109; age 29.5±5.0 years; prepregnancy BMI 25.3±5.4 kg/m2) were divided into four groups according to test results: positive both OGTT and SMBG (n=12, 11.7%); positive OGTT and negative SMBG (n=14, 13.6%); positive SMBG and negative OGTT (n=9, 8.7%); and both tests negative (n=68, 66.0%). Clinical characteristics and maternal outcomes were not different between groups. Neonatal hypoglycemia rates were greater in groups with hyperglycemia compared with the group with both tests negative, (9/12, 7/14, 5/9 vs. 6/68; P<0.001). Participants reported no convenience difference between methods but would prefer OGTT for a future pregnancy.
Conclusions
More than half of the women with a positive OGTT result were normoglycemic in daily life. Conversely, 11.7% of women with a negative OGTT had pregnancy hyperglycemia. Positive OGTT and/or SMBG were equally associated with greater neonatal complications. The results of this study suggest that regular SMBG in addition to OGTT could improve the care of pregnant women.
This study, in a larger view, demonstrates that the standard OGTT during pregnancy may not detect all individuals with hyperglycemia and gestational diabetes (GD). Importantly, a recent study demonstrated that mothers of large for gestational age (LGA) infants had significantly higher glucose overnight compared with mothers with normal for gestational age infants. Detecting and addressing nocturnal glucose control may help to further reduce rates of LGA in women with GD (20). The efficacy of CGM for GD is yet to be demonstrated in a well‐designed RCT, however it seems logical that with proper training and education CGM could help pregnant women with GD prevent delivering LGA babies.
Conclusion
This 2019 CGM article slowly starts to reorient the CGM discussion from the CGM efficacy and cost‐effectiveness toward CGM outcomes. With clinical targets for TIRs agreed upon and endorsed by several major international professional associations, implementation of this novel paradigm into day‐today diabetes management remains a gigantic task. Development of national TIRs targets implementation guidelines endorsed and promoted by national diabetes association is in progress. However, to really improve the depressing recently published data on metabolic control in T1D, where compared with HbA1c <7%, HbA1c >9% was associated with 2‐fold and 12‐fold higher incidences of SH and DKA, respectively, particularly in younger adults (21), people with diabetes will have to decide whether the new treatment paradigm of TIRs targets suits them better. The TIR targets of maintaining glucose roughly 17 hours per day within the range between 70 and 180 mg/dL (3.9–10 mmol/L), with less than 1 hour below 70 mg/dL (3.9 mmol/L), and less than 6 hours above 180 mg/dL (10 mmol/L) seem intuitive. Most importantly, these targets are prospective: an individual with diabetes can improve TIR every day with ongoing CGM readings and actions to maintain glucose within the target range. This is diametrically different from the inherently retrospective nature of the SMBG/HbA1c–based glucose control. The reply to the big question therefore stays with people with diabetes: will they adopt the novel prospective treatment strategy with TIR targets and engage more intimately in day‐to‐day diabetes management? If so, will CGM finally help them improve their outcomes!
Footnotes
Author Disclosure Statement
T.B. served on advisory boards of Novo Nordisk, Sanofi, Eli Lilly, Boehringer, Medtronic, and Bayer Health Care. T.B.'s employer University of Ljubljana and UMC Ljubljana received research grant support, with receipt of travel and accommodation expenses in some cases, from Abbott, Medtronic, Novo Nordisk, GluSense, Sanofi, Sandoz, and Diamyd. T.B. received honoraria for participating on the speaker's bureau of Eli Lilly, Bayer, Novo Nordisk, Medtronic, Sanofi, and Roche. T.B. owns stocks of DreaMed Diabetes.
K.D. declares that no competing financial interests exist.
B.B. received consultancy and speaker fees from Adocia, Astra Zeneca, Bayer, Diasome, Intarcia, Janssen, Mannkind, Medtronic, Novo Nordisk, and Sanofi. B.B.'s employer, Atlanta Diabetes Associates, has received research and grant support from Abbott, Becton Dickson, Boehringer Ingleheim, Diasome, DexCom, Janssen, Lilly, Mannkind, Medtronic, Novo Nordisk, Sanofi, and Senseonics.