Type 1 diabetes mellitus (T1DM) is characterized by absolute insulin deficiency caused by the autoimmune destruction of pancreatic islet cells that results in chronic hyperglycemia [1]. Consequently, individuals with T1DM require lifelong insulin therapy to achieve glycemic control and prevent long-term complications [2]. The global prevalence of T1DM in children and adolescents is rising, posing a growing public health burden [3,4]. Although its incidence remains lower in youth in Korea versus Western countries, it has shown a steady increase over recent decades [5,6]. From 2007 to 2017, the incidence and prevalence of childhood-onset T1DM in Korea increased annually by approximately 3%–4%, with the most notable rise being observed among boys [7]. More recent nationwide data indicate that from 2008 to 2021, the incidence among Koreans under 30 years increased from 3.02 to 3.75 per 100,000 population, with the greatest increases observed in children aged 0–14 years; over the same period, the prevalence more than doubled from 21.82 to 46.41 per 100,000 population [8].

Glucose monitoring has traditionally relied on fingerstick blood glucose meters (BGMs) and periodic glycated hemoglobin (HbA1c) assessments to guide treatment decisions and evaluate a patient's glycemic control [9]. Over the past 2 decades, the landscape of glucose monitoring has shifted significantly with the introduction and growing use of continuous glucose monitoring (CGM) systems. CGM systems measure interstitial glucose levels using subcutaneous enzyme- or fluorescence-based sensors that provide real-time data every 1–5 minutes and generate dynamic comprehensive glucose profiles. These technologies are broadly categorized into 3 types: retrospective (professional), real-time CGM (rtCGM), and intermittently scanned CGM (isCGM) [10].

CGM has transformed the management of T1DM in pediatric populations by improving glycemic outcomes, reducing the risk of hypoglycemia, and enhancing quality of life [11,12]. Current guidelines recommend the early initiation of CGM, ideally at or soon after diagnosis [10]. T1DM is now also being recognized as a progressive disease that advances through defined preclinical stages prior to clinical symptom onset [13]. Emerging evidence suggests that CGM may offer a non-invasive approach to identifying early glycemic abnormalities that indicate disease progression, particularly in presymptomatic stages [14]. This review aims to provide practical guidance for the clinical implementation of CGM, with a particular focus on the context of Korean healthcare, through a comprehensive overview of device characteristics, supporting clinical evidence, and future directions.

A typical CGM system consists of a subcutaneous sensor, wireless transmitter, and receiver or smart device that displays glucose readings and trends at frequent intervals. Technological advancements—such as improved sensor accuracy, nonadjunctive use, factory calibration (eliminating the need for a BGM), compact design, remote monitoring, and integration with insulin delivery systems—have supported the widespread adoption of these systems, particularly in pediatric populations [15]. The latest generations of rtCGM and isCGM systems—including the Dexcom G7 (Dexcom Inc., USA); Medtronic Guardian 4 and Medtronic Simplera (Medtronic Inc., USA); and Abbott FreeStyle Libre 3 and FreeStyle Libre 2 (Abbott Diabetes Care Inc., USA)— are factory calibrated and do not require user calibration using a capillary BGM. However, manual calibration remains an option on some rtCGM devices if persistent discrepancies between CGM and capillary blood glucose readings are observed. Modern CGM sensors are minimally invasive, designed for transdermal self-application, and worn for 6–15 days.

The CGM systems that are currently approved in Korea include the Guardian 4 (Medtronic Inc.), Dexcom G7 (Dexcom Inc.), FreeStyle Libre 2 (Abbott Diabetes Care Inc), and Caresens Air (i-SENS Inc., Seoul, South Korea) (Table 1). All 4 are classified as rtCGM systems that provide continuous glucose data without requiring active scanning, although they differ in terms of technical and regulatory characteristics such as device dimensions, approved age, sensor wear sites, transmitter integration, wear duration, warm-up time, and mean absolute relative difference (MARD). Device sizes vary slightly, and the minimum approved user age ranges from 2 to 19 years depending on the product. Sensor application sites differ by device and age group, with approved locations including the upper arm, abdomen, and upper buttocks. The Dexcom G7 (Dexcom Inc.), FreeStyle Libre 2 (Abbott Diabetes Care Inc), and Caresens Air (i-SENS Inc.) systems incorporate integrated transmitters, whereas the Guardian 4 (Medtronic Inc.) system requires a separate transmitter. Sensor wear duration ranges from 7 to 15 days, while warm-up times vary from 30 minutes to 2 hours. Reported MARD values in pediatric populations range from 8.1% to 12.3% depending on device, attachment site, and age group.

New-generation CGM systems already available in other countries are expected to be introduced in Korea in the near future. These include the Simplera (Medtronic Inc.), which features an all-in-one integrated sensor and is approved for younger children (≥2 years), and the FreeStyle Libre 3 (Abbott Diabetes Care Inc.), which offers a smaller and thinner sensor with improved accuracy. A 15.5-day wear version of the Dexcom G7 (Dexcom Inc.) was recently approved for adult use in the United States and is expected to become more widely available [16]. The FreeStyle Libre 2 Plus and Libre 3 Plus (Abbott Diabetes Care Inc.), which offer extended wear durations of up to 15 days and revised minimum age indications as young as 2 years, have also been developed.

The Advanced Technologies & Treatments for Diabetes consensus, first established in 2017 and revised in 2019, recommends the use of standardized CGM metrics to optimize diabetes management. These include the number of days of CGM use (≥14 days), time percentage of active CGM use (≥70%), mean glucose, glucose management indicator (an estimate of HbA1c based on CGM data), glycemic variability measured by the coefficient of variation (CV ≤36%), time above range (TAR) >250 mg/dL (<5%), TAR 181–250 mg/dL (<5%), time in range (TIR) 70–180 mg/dL (>70%), time below range (TBR) 54–69 mg/dL (<4%), and TBR <54 mg/dL (<1%) (Table 2) [17,18]. Standardized reports such as the ambulatory glucose profile (AGP), currently in version 5.0, provide visual summaries of these 10 key metrics to support informed decision-making by both patients with diabetes and their healthcare providers (Fig. 1) [19]. Each CGM manufacturer offers proprietary software used to visualize and interpret the CGM data—such as Dexcom CLARITY (Dexcom Inc.), which can be integrated with the Pasta app (Kakao Healthcare Corp., Korea); CareLink (Medtronic Inc.); LibreView^® (Abbott Diabetes Care Inc); and the Caresens Air App (i SENS Inc.)—that facilitate AGP report generation and core metric reviews (Table 1).

Time in tight range (TITR; also referred to as time in normoglycemia [TING])—defined as the percentage of time that blood glucose remains within 70–140 mg/dL—has recently emerged as a promising metric for evaluating optimal glycemic control [20]. Although a TITR of >55% may be required to achieve an HbA1c ≤6.5%, a more widely accepted threshold is >50% [21]. Although further research is warranted in pediatric populations, an increased TITR is thought to reflect reduced glycemic variability and a lower risk of complications. Importantly, the effective interpretation and use of CGM metrics must be supported by comprehensive education of both children and their caregivers to address the broader health needs beyond glycemic targets.

Multiple randomized controlled trials (RCTs) have demonstrated the clinical utility of CGM in pediatric patients with T1DM by reporting improvements in glycemic control and reductions in hypoglycemia (Table 3). In the Juvenile Diabetes Research Foundation (JDRF; renamed Breakthrough T1DM in 2024) CGM study, children aged 8–14 years using an rtCGM experienced significant reductions in HbA1c compared to those receiving standard care, with greater benefits observed in those with higher adherence [22]. Similarly, the ONSET (Paediatric Onset) study reported no overall difference in HbA1c levels between children who initiated sensor-augmented insulin pump therapy at diagnosis and those who used pump therapy with BGM. However, regular (≥1/wk) sensor users—particularly adolescents—showed significantly lower HbA1c levels, reduced glycemic variability, and better C-peptide preservation. No severe hypoglycemia was reported in the sensor group, which also supports CGM safety and metabolic benefits [23].

In one multicenter RCT, rtCGM use reduced the time spent in hypoglycemia (<63 mg/dL) by 48% among pediatric participants and produced a modest but significant reduction in HbA1c—particularly among those who were already within near-target glycemic ranges [24]. An additional analysis of the JDRF data showed that rtCGM use significantly reduced glycemic variability markers, reinforcing CGM’s role in mitigating acute glucose excursions [25]. The CITY (CGM Intervention in Teens and Young Adults with Type 1 Diabetes) study reported a small but statistically significant reduction in HbA1c levels (-0.37%) over 26 weeks in adolescents using CGM versus BGM along with increased TIR and reduced hypoglycemia; these benefits were observed as soon as 13 weeks [26].

In younger children (aged 2 to <8 years), the Strategies to Enhance New CGM Use in Early Childhood (SENCE) study showed that rtCGM use did not improve TIR, but it significantly reduced hypoglycemia and severe hypoglycemic events compared to BGM [27]. In a study of do-it-yourself rtCGM use, rtCGM use lead to significant improvements in TIR and reductions in hyperglycemia compared to isCGM without increasing hypoglycemia. Although it did not reduce parental fear of hypoglycemia, caregiver satisfaction and quality of life improved, supporting the short-term feasibility and psychosocial benefits of rtCGM [28].

An 8-week RCT that compared rtCGM to isCGM in children and adolescents revealed that rtCGM significantly reduced severe hypoglycemia. Among those with higher risks of hypoglycemia (baseline TBR, ≥5%), rtCGM use further reduced both TBR and glycemic variability, indicating greater benefits in high-risk pediatric populations [29]. A follow-up study of the SENCE confirmed that rtCGM use in children aged 2 to <8 years resulted in sustained reductions in hypoglycemia and consistently high device adherence regardless of behavioral interventions. However, TIR and HbA1c remained unchanged, highlighting persistent challenges in managing hyperglycemia in this age group [30].

In a multicenter RCT conducted in children aged 4–13 years with elevated HbA1c levels, isCGM (via the FreeStyle Libre 2.0 system) increased glucose monitoring frequency and reduced time in hypoglycemia compared to BGM, although no significant improvements were seen in terms of HbA1c, TIR, or psychosocial measures [31]. A recent meta-analysis of 5 RCTs involving 92 children and adolescents found that rtCGM versus isCGM use led to a greater increase in TIR, primarily by reducing hypo- and hyperglycemia. However, the TIR benefit was more pronounced in adults, and no significant difference in HbA1c was observed in the pediatric subgroup [32].

The superior glycemic outcomes with rtCGM versus isCGM are supported by recent nationwide cohort data showing greater and sustained HbA1c reductions and larger decreases in CV over 24 months in both adults and children with T1DM [33]. These benefits are likely attributable to rtCGM’s real-time data transmission and predictive alerts, which facilitate prompt corrective action, whereas isCGM requires intermittent scanning that may delay the detection and management of rapid or asymptomatic glucose excursions. Moreover, evidence from the coronavirus disease 2019 pandemic in Korean pediatric patients with T1DM indicates that CGM enabled effective glucose monitoring and supported timely insulin adjustments during periods of acute illness and restricted healthcare access, helping mitigate short-term glycemic deterioration and underscoring its value in maintaining metabolic stability during the public health crisis [34]. Newer-generation CGM devices that feature improved accuracy, ease of use, and integration with insulin pumps or decision-support algorithms are expected to further enhance their utility in young children and high-risk populations with T1DM.

Beyond its established role in glycemic management, CGM is increasingly being explored for the early detection and risk stratification of T1DM in its presymptomatic stages. In one prospective study of 91 children (median age, 11.5 years) with persistent islet autoantibodies identified through general population screening, achieving glucose levels of >140 mg/dL >10% of the time was associated with an 80% risk of progression to clinical T1DM within 12 months [35]. Similarly, in a study of 46 asymptomatic children and young adults (aged 4–25 years) with human leukocyte antigen–conferred genetic susceptibility to developing T1DM who were enrolled in the Finnish Type 1 Diabetes Prediction and Prevention (DIPP) study, a 10-day CGM assessment effectively identified individuals with asymptomatic stage 3 T1DM and stage 2 dysglycemia [36]. In the TrialNet study, which involved 105 first- or second-degree relatives of individuals with T1DM (aged 5–42 years) who tested positive for ≥2 islet autoantibodies, achieving glucose levels of ≥140 mg/dL ≥5% of the time was linked to a 40% risk of progressing to clinical T1DM within 2 years [37]. A longitudinal study of 34 nonoverweight first-degree relatives (aged 5–39 years) further demonstrated that CGM-derived metrics—particularly time spent above 120 and 140 mg/dL—had predictive value regarding progression to stage 3 T1DM. In this context, CGM combined with HbA1c may represent a practical alternative to oral glucose tolerance tests (OGTTs) for long-term monitoring [14].

Beyond early T1DM detection, CGM is being explored in various other clinical settings as well. In critically ill patients, CGM has demonstrated potential for real-time glucose monitoring in intensive care units, supporting insulin titration and reducing hypoglycemia risk [38]. In a Korean retrospective study of 18 critically ill pediatric patients with diabetic ketoacidosis (mean age, 11.0±4.2 years), CGM showed acceptable clinical accuracy (overall MARD, 13.0%), with performance improving as diabetic ketoacidosis resolved and accuracy negatively correlating with lower serum bicarbonate levels [39]. However, certain challenges remain, including variable accuracy, limited evidence of mortality benefits, and practical barriers such as sensor lag and cost [40-43]. Furthermore, CGM is also being evaluated as a screening tool for prediabetes because of its capacity to detect postprandial glucose excursions and glycemic variability that are not captured by fasting plasma glucose or HbA1c levels. Although CGM offers greater sensitivity and facilitates lifestyle modifications through real-time feedback, it currently serves as a supplemental rather than standalone diagnostic tool [44]. Emerging evidence supports the use of CGM in the diagnosis and management of gestational diabetes mellitus (GDM) as well. CGM can uncover glycemic excursions that may be missed by standard OGTTs, promote healthier behaviors, and enable earlier interventions [45-47]. It also aids the prediction of adverse maternal and neonatal outcomes, particularly when integrated with education and clinical support [48,49]. Collectively, these expanded applications underscore the evolving role of CGM as a minimally invasive and dynamic tool that can be used for diabetes management as well as diagnostics, staging, and risk stratifications of a broad range of other disorders related to glucose metabolism.

Looking ahead, the integration of CGM with automated insulin delivery system is expected to further transform diabetes care. These systems use rtCGM data to automatically adjust insulin delivery, reducing both hypo- and hyperglycemia while minimizing the burden of frequent manual insulin dose adjustments [50,51]. Advances in algorithm development, interoperability standards, and device miniaturization are anticipated to enhance the accessibility and effectiveness of closed-loop technology in pediatric populations in inpatient settings.

In Korea, the National Health Insurance Service (NHIS) began offering partial reimbursement for rtCGM sensors in January 2019. This coverage was expanded in January 2020 to include rtCGM transmitters and isCGM sensors for individuals with T1DM [52]. Transmitters are reimbursed at 70%, up to 210,000 Korean won (KRW) per 3-month period, while sensors are supported at a rate of 70% for a daily cost of up to 10,000 KRW; the remaining 30% is paid by the patient. These policy changes marked a pivotal step toward improving access to diabetes technology, particularly among pediatric patients and those at higher risk of glycemic instability. In 2021, the Committee of the Health Insurance and Government Relation of the Korean Diabetes Association released a position statement advocating for broader reimbursement criteria. The statement emphasized the importance of covering both CGM hardware and sensors and the associated consumables and services such as patient and caregiver education, data interpretation, and clinical management support [53]. As of February 2024, NHIS coverage was further expanded to cover up to 90% of the annual cost of CGM devices and sensors for individuals aged <19 years. Coverage was also extended to pregnant women with GDM, who now receive the same support as adult patients with T1DM (i.e., 70% coverage for a daily cost of up to 10,000 KRW).

According to NHIS claims data, 1,434 of 19,730 individuals with T1DM (7.3%) were prescribed CGM sensors between January and December 2019; of them, 751 (52.4%) were <20 years of age [52]. A more recent analysis of NHIS data showed that between 2019 and 2022, 1,911 individuals aged <19 years with T1DM used CGM; of them, 59.3% used rtCGM [33]. Although CGM adoption in Korea remains relatively low compared with that in other high-income countries, this increasing trend—particularly among pediatric users—reflects growing clinical acceptance and highlights the need for continued policy support and infrastructure development to enhance equitable access. However, despite expanded NHIS reimbursement since 2019, substantial barriers persist. These barriers include the financial burden from remaining out-of-pocket costs, lack of structured long-term education for patients and clinicians, adherence challenges (particularly in younger users), device-related discomfort and skin problems, complex data interpretation, psychological burden, and incomplete integration of CGM metrics into routine clinical care [54]. Overcoming these multi-faceted barriers will be essential to the promotion of broader and more effective CGM adoption in Korean pediatric T1DM. Moreover, given the clinical benefits of CGM across all age groups, a gradual expansion of NHIS coverage to include adults with T1DM should be considered. Such an approach would ensure continuous technology use during the transition from pediatric to adult care and help optimize long-term glycemic outcomes.

CGM has transformed the management of T1DM in pediatric populations, offering significant benefits for glycemic control, hypoglycemia prevention, and overall quality of life. Robust evidence from RCTs supports the use of rtCGM, particularly when initiated early and used consistently, with additional advantages being demonstrated in high-risk and younger children. Although isCGM also confers certain benefits, rtCGM generally provides superior outcomes in TIR and hypoglycemia reduction. In Korea, recent expansions in national health insurance coverage have improved access to CGM technologies, but uptake remains suboptimal compared to global benchmarks. Continued efforts are warranted to broaden reimbursement for these systems, support education and data interpretation services, and promote equitable access. As CGM devices become more accurate, user-friendly, and interoperable with insulin delivery systems, they hold increasing promise for improving long-term outcomes in Korean children and adolescents with T1DM.