Use of Real-Time Continuous Glucose Monitoring Technology in Children and Adolescents

CGM systems measure glucose in the interstitial fluid, not blood glucose,providing interstitial glucose readings every 1–5 minutes. Rapid blood glucose changes may result in a 6- to 18-minute delay in interstitial glucose readings.⁸ Approximately 6–8 minutes of the delay is the result of a physiological lag,⁹  and up to 12 minutes of delay may result from filters on the glucose sensor.⁸ 

CGM systems are currently approved for adjunctive use and should be used in addition to blood glucose testing. Blood glucose testing is used to calibrate CGM systems and should be the basis for treatment decisions, such as insulin dose adjustments and determinations of ability to drive. Blood glucose tests should also be performed to verify CGM data and alarms, as well as anytime there are symptoms of a low or high glucose level.

If an errant blood glucose test result is entered as a calibration, or if calibration is performed when the glucose level is changing rapidly, the accuracy of the CGM readings may be impaired. Calibration should be done when the glucose is not changing rapidly. Finding a good time to calibrate may be more challenging in children, whose eating and activity patterns are inconsistent and varied.

CGM systems are composed of several components. Disposable sensors are inserted in the subcutaneous tissue using an insertion device. The sensors generate a current that correlates with the level of glucose in the interstitial fluid. The transmitter then relays the information to the receiver, where it is displayed. Several CGM systems are presented in Table 1, some of which are approved for use in children.

The hope for a cure is a dream shared by many people with diabetes,especially children and their families. This strong desire is a positive force that provides hope and encourages research, but it can also contribute to high expectations for any novel drug or device. User expectations for CGM are no different, and having realistic expectations for the technology is an important first step in adapting its use. CGM systems are not a cure, and they are not an artificial pancreas.

These systems provide more and different data about glucose control than were previously available, but it is information that patients and their families need to be ready and willing to use. Although this technology has many advantages, it may not be suitable for all families and children with diabetes. The increased data may be overwhelming for some people, and the ability to interpret and respond to the data appropriately is an important consideration in patient selection. CGM is a behavior modification tool. Having real-time data will not prevent all hypo- or hyperglycemia, but the glucose alarms and the direction- and rate-of-change information may help prompt preemptive intervention. Children and their families should appreciate that CGM is not perfect, but the increased data can be an effective and powerful tool in diabetes management.

CGM technology is only effective when it is used. CGM use in children and adolescents provides unique challenges. In young children, especially those using insulin pump therapy, the addition of a transmitter may take up valuable body real estate. The smaller the child, the bigger the challenge. Transmitter size and the location of sensor insertion are important considerations in children. For children who are active or frequently perspire, the adhesion of the sensors and transmitters can be problematic and may require experimentation with various adhesives and skin preparations. Finding creative approaches to keeping receivers from getting lost, dropped, or damaged can be challenging.

It is important to consider the possibility that the desire to use CGM may lie with the parents and not the children. It is the children ultimately who will be wearing the system, and their opinion is key in the decision to use the technology. It is often beneficial to provide a trial period of sensor wear, and ideally a trial of inserting different sensors and using different CGM systems, especially if the child does not want to use the system because of anxiety about inserting or wearing the device. In older children, concerns about the violation of privacy should be addressed because this may create more conflict and result in decreased use and benefit from the technology. In this case, it may be worthwhile to work out an agreement between parents and children about how and when data will be reviewed. It is also possible that children may be willing to use the device intermittently.

Surveys of children with type 1 diabetes and their parents who were enrolled in a pilot trial using the FreeStyle Navigator system revealed a high level of satisfaction. More than 70% of both parents and children reported the use of CGM made adjusting insulin easier, made them more confident in their management decisions, and helped detect glucose patterns they were not previously aware of.¹⁰  Of particular interest in this small pilot trial was the finding that using CGM did not increase family conflict, and device use remained fairly constant throughout the study.¹⁰  During the pilot trial, which lasted 13 weeks, the CGM device was used 80% of the time in the first 4 weeks and 70% of the time in weeks 9–13.⁵ 

All of the current CGM systems have threshold alarms for low and high glucose. For children and their families, this feature is one of the major draws of the technology. These alarms are used in addition to traditional methods of detecting hypo- and hyperglycemia. CGM systems are not, however, an infallible method of detecting low or high glucose and should not be depended on as the only means of detecting these events. Lag time and errors in calibration, among other things, may prevent the system from alarming. In some cases, the system may alarm but the user may not respond.

In a study of responses to alarms generated by the GlucoWatch G2 Biographer, children and their parents were videotaped overnight using an infrared camera. Children awoke to 29% of alarms.¹¹  The alarm response rate increased with age: 4- to 6-year-olds responded to 17% of alarms, 7- to 11-year-olds responded to 20% of alarms, and there was a 53%response rate in adolescents.¹¹ Parents who were asleep in the same room responded to 37% of alarms.¹¹  It has been our experience that many patients and families do not hear the audible or feel the vibratory alarms overnight, and many feel they would benefit from louder alarms or alarms that were able to remotely alert others or turn on lights.

Children and adolescents in particular are at risk for missing meal boluses. This is especially true when they are at school. In a study of 48 children between 7 and 20 years of age (mean age of 15.3 ± 3.0 years)using insulin pump therapy but not meeting their glycemic goals, there was a correlation between the frequency of missed meal boluses and the elevation in A1C levels.¹²  This study found a 0.5% increase in A1C when two meal boluses per week were missed and an increase of 1% if four meal boluses per week were missed.¹² 

Real-time alarms for actual or projected high glucose can help alert children and adolescents to missed meal boluses and facilitate correction actions earlier than conventional blood glucose testing alone. Although preventing a missed bolus is the ideal scenario, responding rapidly to the postprandial hyperglycemia after a missed meal bolus is a step in the right direction.

Low and high glucose alarms and projected low and high alarms, can be a valuable adjunct to symptoms and blood glucose meter data in detecting hypo-and hyperglycemia. Alarm settings should balance sensitivity and specificity to maximize event detection and minimize nuisance alarms. Setting the low alarm at 80 mg/dl may improve the sensitivity, but if the patient and family are constantly warned of lows when the blood glucose is in the 100-mg/dl range, they may tire quickly, especially when the alarms are occurring overnight.

A study of the FreeStyle Navigator system found when the low threshold is< 70 mg/dl, the alarm sensitivity is 79.8% (defined as a glucose < 70 mg/dl within ± 30 minutes of the threshold or projected alarm). The rate of false-positive alarms (defined as a glucose > 85 mg/dl) was 7.2%.¹³  When the low alarm was raised to 80 mg/dl, the detection rate for blood glucose < 70 mg/dl increased to 92.5%, but the false alarm rate increased to 27.8%.¹³ 

CGM leaves little to the imagination. The continuous stream of data identifies glycemic excursions that were previously undetected, even with frequent blood glucose testing. Given the limitations of our current therapies and the numerous variables that affect glucose control, perfect glucose control is not a realistic goal for most patients. Never before have patients and families been so aware of the successes and missteps in care.

It is important to look at the data objectively, to identify patterns, and to work to make changes. As children gain independence with their diabetes in the age of CGM, it will be increasingly important to view the data objectively, focusing on the positives and working as a team to make changes. Children and adolescents need to understand that no one is perfect; we all make mistakes. Transitioning care from parents to children or adolescents can be very challenging, and focusing on what is working and providing positive feedback is as important as identifying areas for improvement. In fact, CGM may empower children and adolescents with the support of their families to take a more active role in their diabetes management. The example in Figure 1 shows the improvement in glucose control for a 15-year-old between when his mother managed his diabetes and 13 weeks later, when he had gained independence in his care.

For some patients, the desire to maintain autonomy and to demonstrate their diabetes management abilities may lead them to be overly aggressive with their insulin doses in response to glucose trends and alarms. In Figure 2, a 13-year-old who had received education on the dangers of over-correcting before starting CGM was alerted to a high glucose his first night of sensor use. He admitted that the high glucose was frustrating, and his instinct was to get it down as fast as possible without regard to insulin on board and without use of the bolus calculator. The figure shows that the patient gave two boluses close together,raised his basal rate, and used a square wave in attempts to achieve lower glucose values. As a result, his blood glucose dropped quickly, and he was woken again by alarms for a low glucose that needed treatment. This example highlights the need to address expectations and teach patients and parents the onset, peak, and duration of insulin to prevent over-correction.

When using CGM, patients and parents need to make decisions about how they will respond to the glucose readings on many levels. Data provided from CGM,including the rate and direction of glucose change and short- and long-term retrospective trend data, in combination with blood glucose tests, can guide therapy decisions. This more complete picture of glucose control warrants a comprehensive plan to maximize the use of all the data.

In the DirecNet pilot trial, subjects with a mean age of 11.2 ± 4.1 years of age using CGM tested algorithms designed to provide guidelines for the use of both real-time and retrospective data.¹⁴  The DirecNet Applied Treatment Algorithm (DATA) was written and tested using the FreeStyle Navigator in patients on insulin pump therapy. A condensed version of the algorithms that has been modified for use in a variety of systems appears in Table 2. A complete copy of the algorithms used in the study is available in the appendix of the original article.¹⁴ The algorithms provided guidance for patients on how to respond to glucose alarms, adjust insulin doses, and take into account the glucose trend, as well as how to review and respond to retrospective trends. These algorithms were used as part of a research study, and it is worth emphasizing that all treatment decisions should be confirmed with a blood glucose meter test.

The use of the DATA algorithms increased subjects' autonomy. This is evidenced by the frequency of therapy adjustments made by patients and their families. After 1 week of CGM use, 10% of subjects reported making changes to insulin-to-carbohydrate ratios, and 20% made changes in basal rates.¹⁴  By the end of the study, this increased to 25% reporting making a change in insulin-to-carbohydrate ratios and 32% adjusting basal rates.¹⁴  After only 3 weeks of use, 82% of subjects and 96% of parents felt DATA was helpful in guiding insulin management decisions.¹⁴  The use of algorithms decreased over time to 59% of subjects and 73% of parents by 13 weeks.¹⁴  This reduction may reflect increased confidence in making their own decisions and a better understanding of their individual needs as a result of CGM use.

The success of the DATA algorithms warrants further study and consideration when teaching patients how to respond to CGM data. Central to these algorithms is a thorough understanding of insulin action profiles. Teaching the onset,peak, and duration of all types of insulin patients are using is crucial. For patients using smart insulin pumps with bolus calculators, it is important to review their insulin-on-board settings and encourage the use of this feature to prevent insulin “stacking.” For patients not using smart pumps,discuss how to calculate insulin on board.

The ability to use direction- and rate-of-change data to prevent glucose from rising above or dropping below the threshold is an advantage not previously afforded with traditional blood glucose meters. The use of real-time data is exciting and empowering, but it is not the only benefit of CGM use. Retrospective glucose trends should be used to make medication or lifestyle modifications. When reviewing these trends, it is important to consider recurrent patterns of real-time proactive measures that were taken,so that changes can be made to basal rates, insulin sensitivity calculations,and insulin-to-carbohydrate ratios to decrease the need for recurring corrective measures.

A fear of hypoglycemia that keeps many parents from sleeping through the night, hormonal changes with puberty, varied activity and meal schedules, and the gradual transition of care from parents to children are all unique challenges in children and adolescents with diabetes. They are also common reasons that parents and children cite for using CGM. The additional data afforded by CGM use, especially data on overnight and postmeal glucose trends,can help increase confidence in therapy modifications. Although there are many reasons to use CGM, it is important to recognize that not everyone may be ready or willing to use the technology. For children and families who chose to add CGM to their diabetes regimen, having reasonable expectations for the therapy is important. It is also vital to have a thorough understanding of the onset, peak, and duration of all medications that affect glucose control. Education should include information about the technology itself and about how to safely use both the real-time data and the glucose trend and retrospective information.