Available online on 15.06.2026 at http://jddtonline.info

Journal of Drug Delivery and Therapeutics

Open Access to Pharmaceutical and Medical Research

Copyright  © 2026 The   Author(s): This is an open-access article distributed under the terms of the CC BY-NC 4.0 which permits unrestricted use, distribution, and reproduction in any medium for non-commercial use provided the original author and source are credited

Open Access Full Text Article                                                          Research Article

Initial Experience with a Novel 21-Day Continuous Glucose Monitoring System in Pediatric Type 1 Diabetic Patients: A Two-Center Initial Accuracy Evaluation

Erdal Eren ¹, Cemal Aslan ², Burhan Küçük ², Tayfun Aybek ²

1. Department of Pediatric Endocrinology, Bursa Uludag University Faculty of Medicine, Bursa, Türkiye
2. Department of Cardiovascular Surgery, Section Telemedicine, TOBB ETÜ University Faculty of Medicine, Ankara, Türkiye 

 

 

Continuous glucose monitoring (CGM) has become a cornerstone of modern type 1 diabetes (T1D) management, particularly in children and adolescents, in whom frequent fingerstick testing is burdensome and nocturnal hypoglycaemia is a persistent concern ^1,2. CGM systems estimate glucose in the interstitial fluid using a subcutaneous electrochemical sensor and have been shown to improve time-in-range and reduce hypoglycaemia exposure when integrated into routine care ³.

A practical limitation of the currently available factory-calibrated systems is sensor wear time. Widely used devices, including the world-renowned and reputable CGM systems, are approved for a maximum of 14 days of continuous wear ⁴. A longer functional sensor lifetime would reduce the number of insertions, lower per-patient cost, and improve adherence — advantages that are especially relevant in the pediatric setting and in resource-conscious health systems.

This study was conducted jointly by two academic centers in Türkiye. Pediatric patients were enrolled and followed at Bursa Uludağ University, while the remote monitoring, data aggregation and accuracy analysis were coordinated through the Department of Telemedicine of TOBB ETÜ University Faculty of Medicine. The Telemedicine Department, established under the Department of Cardiovascular Surgery, has since 2017 provided remote monitoring with wearable devices to approximately 14,000 patients, and therefore brings substantial operational experience in wearable-sensor data capture and analysis to the present evaluation.

The Perlanova (Instara-1) system evaluated here is, to our knowledge, the first CGM sensor designed for a 21-day continuous wear period — 50% longer than the established standard. Before such a device can be recommended for clinical use, it must be demonstrated that analytical accuracy is not only acceptable at baseline but is maintained throughout the extended wear period, as sensor drift and local tissue reactions can theoretically degrade performance over time. In the manufacturer’s YSI-referenced clinical study conducted in pediatric participants, the MARD was reported as 8.2% ⁵. 

We aimed to characterize the real-world accuracy of the 21-day Perlanova CGM system in pediatric T1D patients, with particular emphasis on how accuracy evolves across the full 21 days of wear.

2. Materials and Methods

This was a prospective, two-center accuracy evaluation. Patients were enrolled and clinically followed at the Department of Pediatric Endocrinology at Bursa Uludağ University Faculty of Medicine. Remote data capture, sensor data aggregation, and accuracy analysis were performed by the Department of Telemedicine, TOBB ETÜ University Faculty of Medicine, Ankara. The study was conducted in accordance with the Declaration of Helsinki.

Eligible participants were pediatric patients with type 1 diabetes. Inclusion criteria were age ≥ 4 years and ≤ 18 years, with written informed consent obtained from the patient and/or legal guardian (assent where applicable). Patients outside this age range and those without a diagnosis of type 1 diabetes were not analysed. In two patients, the sensor became detached due to the user handling; in both cases the sensor was immediately replaced and the patient continued in the study.

The investigated device was the Perlanova (Instara-1) CGM system (legal manufacturer: Teljane/Tangjian, China). The factory-calibrated sensor is designed for up to 21 days of continuous wear and reports interstitial glucose at 5-minute intervals; no user calibration was performed. The study was conducted using the mobile application version 1.0.1 for the sensor placed on the inner side of the upper extremity.

Reference glucose values were obtained by capillary (fingerstick) self-monitoring of blood glucose (SMBG) using Counter Plus (Ascensia Diabetes Care, NJ, USA), which is compatible with ISO 15197:2013. The comparator was a capillary point-of-care meter rather than a laboratory venous reference (e.g., YSI); this pragmatic, real-world design is discussed as a limitation in Section 4.4. Reference measurements were performed by patients or their parents at times of their own choosing rather than on a fixed protocol schedule.

Each reference SMBG value was paired with the temporally nearest CGM reading. Because reference and sensor timestamps rarely coincide exactly, three matching windows were pre-specified (±5, ±10, and ±15 minutes) and analyzed in parallel as a sensitivity analysis. MARD was calculated using capillary blood glucose measurements obtained with the Contour® Plus blood glucose monitoring system as the reference method. Sensor glucose values within ±5,10,15 minutes of each capillary measurement were matched, and the nearest sensor value was used for analysis. MARD was defined as the mean, across all paired points, of |CGM − reference| / reference × 100. Point-of-care agreement rates (proportion of pairs within ±15/15 and ±20/20 mg/dL or %) and Bland–Altman statistics were computed on the ±15-minute dataset, reported as the primary analysis. The day of sensor wear was derived for each sensor as the number of days elapsed since its first recorded reading, enabling characterization of the MARD trajectory across the 21-day period.

Continuous variables are summarised as mean ± SD or median (range/IQR). MARD is reported overall, by day of wear, and stratified by reference glucose range (<54, 54–69, 70–180, 181–250, >250 mg/dL) [3]. Pearson's correlation and linear regression describe the CGM–reference relationship; the Bland–Altman analysis describes systematic bias and the limits of agreement. Analyses were performed in Python (pandas, NumPy, Matplotlib). A formal power calculation was not performed for this initial evaluation.

The study protocol was approved by the Ethics Committee of TOBB ETÜ University Faculty of Medicine (approval no. 060, dated 19 February 2019), and Bursa Uludağ University, Faculty of Medicine (approval no. 2026/281/10-15). Written informed consent was obtained from all participants and/or their legal guardians.

3. Results

Thirty-two pediatric T1D patients aged 4–18 years contributed 1,523 paired CGM–reference measurements (median 52 pairs per patient, range 5–92). The median age was 13 years (IQR 11–15). Baseline characteristics are summarised in Table 1.

 

 

 

Table 1: Baseline characteristics of the study population.

Characteristic	Value
Patients (pediatric T1D), n	32
Paired data points, n	1,523
Median pairs per patient (range)	52 (5–92)
Age, median (IQR), years	13 (11–15)
Age 4–6 / 7–12 / 13–18 y, n	4 / 10 / 18
Diabetes type	Type 1 (all)
Maximum sensor wear, days	21
Sensors replaced for handling-related detachment, n	2

 

 

In the primary ±10-minute analysis, the overall MARD was 8.6% (median absolute relative difference 3.8%). CGM and reference glucose were strongly correlated (r = 0.963; regression slope 0.90, intercept 11.3 mg/dL; Figure 1). Point-of-care agreement was high: 87.5% of pairs fell within ±15/15 and 93.1% within ±20/20 (mg/dL or %). In the consensus error-grid analysis (Figure 2), 98.9% of paired points fell within the clinically acceptable zones A+B (Figure 2). Bland–Altman analysis demonstrated a small negative mean bias of −5.9 mg/dL (95% limits of agreement −50.2 to +38.4 mg/dL), indicating that the sensor reads marginally lower than the capillary reference on average (Figure 3).

 

 

 

Figure 1: CGM versus reference (capillary) glucose with identity line and linear fit (r = 0.963).

 

Figure 2: Consensus error-grid representation; the shaded band indicates ±20% around the reference. 98.9% of points fell within zones A+B.

 

 

 

Figure 3: Bland–Altman plot of CGM minus reference glucose. Mean bias −5.9 mg/dL; dashed lines show 95% limits of agreement.

 

The central question of this study — whether accuracy is maintained over the extended 21-day wear time — is addressed in Figure 4. Daily MARD remained largely within the 5–8% band throughout the wear period. MARD was slightly higher on day 1, consistent with the well-described early sensor warm-up and run-in behavior, then settled and remained stable through day 21, with no progressive deterioration within the labeled lifetime. Among the cohort sensors, 24 contributed data on day ≥15, and 19 reached day 21, supporting the feasibility of the extended-wear claim.

 

 

Figure 4: Mean daily MARD (± SEM) across the 21-day sensor wear period (±15-minute matching). Accuracy is preserved throughout the labelled lifetime

 

MARD varied by glycaemic range in the expected manner (Figure 5, Table 2). Accuracy was best in the euglycaemic and hyperglycaemic ranges most relevant to day-to-day management, and poorest in the low-glucose ranges, where both the small absolute denominators and the limited number of paired hypoglycaemic points amplify the relative error.

 

 

Table 2: Accuracy stratified by reference glucose range (±15-minute matching).

Reference range (mg/dL)	n pairs	Mean abs. rel. diff (%)	Median abs. rel. diff (%)
<54	9	16.5	13.7
54–69	76	12.4	7.0
70–180	854	7.2	3.5
181–250	328	5.8	2.8
>250	256	7.9	5.3
Overall	1,523	7.3	3.8

 

Figure 5: MARD stratified by reference glucose range. Relative error is highest in hypoglycemia, as expected for CGM.

 

 

Overall MARD depended on the reference-matching window: 10.8% at ±5 minutes, 8.6% at ±10 minutes, and 7.3% at ±15 minutes (Figure 6). We interpret this gradient as physiologically meaningful rather than as an artefact: CGM measures interstitial glucose, which lags capillary blood glucose. A wider matching window allows the paired reference and sensor values to reflect the same underlying glycaemic state once this physiological delay is accounted for, reducing apparent discordance (see Section 4.2).

Figure 6: Overall MARD as a function of the reference-matching window (±5, ±10, ±15 minutes).

4. Discussion

To our knowledge, this is the first clinical report of a CGM system with a 21-day labeled wear time, and the first such experience in a pediatric T1D population. In this two-center initial accuracy evaluation, the Perlanova Instara-1 CGM system demonstrated clinically acceptable accuracy in children and adolescents with type 1 diabetes across an extended 21-day wear period. The principal finding was that the overall MARD was 7.3% using the ±15-minute matching window, with 93.1% of paired values within the ±20/20 agreement criterion and 98.9% of values falling within consensus error-grid zones A and B. Importantly, accuracy was preserved throughout the labelled 21-day wear duration, with no evidence of progressive deterioration toward the end of sensor life. These findings suggest that the extended wear time of this novel CGM system was not achieved at the expense of late-wear analytical or clinical performance.

The present study used a capillary glucose measure rather than a venous YSI reference standard. Therefore, the findings should be interpreted as a pragmatic real-world accuracy evaluation, while future confirmatory studies using standardized venous/arterialized venous YSI sampling are needed to establish regulatory-grade analytical accuracy.

Established factory-calibrated systems are limited to 14 days of wear. A reliable 21-day sensor reduces the number of insertions by roughly one-third, with corresponding benefits for adherence, patient comfort and cost — considerations that are particularly important in children and in resource-conscious health systems. The stability of daily MARD through day 21 (Figure 4) is the key evidence that the longer wear time is not achieved at the expense of late-wear accuracy.

CGM does not measure blood glucose directly; it measures glucose in the subcutaneous interstitial fluid ⁶. Glucose reaches this compartment by diffusion across the capillary wall, introducing a time lag relative to blood glucose that is not a fixed constant: it depends on the direction and rate of glycaemic change (it is larger when glucose is changing rapidly), on the sensor insertion site and local perfusion, and on individual patient factors ^7,8. Consequently, a CGM value and a simultaneously obtained capillary value are expected to differ and pairing a reference value with a CGM reading taken a few minutes earlier or later may, in fact, reflect the same physiological state more faithfully. This is why a wider matching tolerance reduces apparent MARD, and it argues that a portion of the residual difference between CGM and capillary glucose is physiological rather than a true sensor error. The same physiology underlies the small negative Bland–Altman bias and the larger relative error in the hypoglycaemic range, and should be communicated to clinicians and families, especially regarding the interpretation of low sensor readings.

The accuracy observed here is broadly in the range reported for contemporary factory-calibrated 14-day systems. In a pivotal pediatric study of the factory-calibrated Dexcom G6, the overall MARD against a venous YSI reference was 7.7% ⁴; pediatric real-world and inpatient studies using capillary or point-of-care references have generally reported higher MARD values, in the range of roughly 10–16% ^1,9. Direct numerical comparison with the present data is nonetheless inappropriate, because most pivotal studies used a laboratory venous (YSI) reference rather than capillary SMBG, with standardised in-clinic sampling. Our figures should therefore be read as a real-world, comparator-limited initial experience rather than as a head-to-head accuracy claim.

Several limitations must be acknowledged. First, the reference was capillary fingerstick SMBG rather than a laboratory venous (YSI) standard, which limits comparability with regulatory-grade accuracy studies; these data support but do not replace a formal reference-standard validation. Second, reference sampling was not protocol-driven: patients and parents performed fingerstick measurements at times of their own choosing. In practice, these measurements were most often taken when the sensor indicated a very high or very low value — that is, the paired dataset is enriched for moments of rapid glycaemic change and extreme values, the conditions under which CGM accuracy is hardest to achieve. The favorable accuracy observed despite this enrichment suggests that the sensor performs well precisely in these dynamic, clinically important situations, although the same sampling pattern may bias the overall glycaemic distribution of pairs and should be interpreted with care. Third, paired-point counts per patient were uneven (range 5–92), and the hypoglycaemic range was sparsely sampled (9 pairs <54 mg/dL), limiting confidence in low-glucose accuracy estimates. Finally, this was an initial two-center evaluation without a pre-specified power calculation.

In this first clinical experience, the novel 21-day Perlanova CGM system showed clinically acceptable accuracy in pediatric T1D patients aged 4–18 years, with performance preserved across the full extended wear period and maintained even during the dynamic, extreme-glucose moments that dominated the reference sampling. These encouraging preliminary results justify a larger, adequately powered, multi-centre study using a venous/YSI reference standard and a pre-specified sampling protocol to establish definitive accuracy and formally substantiate the 21-day wear claim.

Declarations

Ethics approval: Ethics Committee of TOBB ETÜ University Faculty of Medicine, approval no. 060, dated 19 February 2019, and Bursa Uludag University Faculty of Medicine (approval no. 2026/281/10-15). Informed consent was obtained from all participants and/or guardians.

Funding: None.

Conflict of interest: Perlanova/Instara-1 (Teljane, Suzhou City, China). This relationship is disclosed. The academic analysis was performed independently at the participating university centers.

Data availability: De-identified data are available from the corresponding author on reasonable request. All patient identifiers must be removed from any supplementary data file before submission.

Author contributions / Acknowledgements. We would like to thank pediatric diabetes nurses [Zeynep Yildizer and Zehra Kuscu] for their valuable support in sensor insertion, patient and family education, and follow-up during the study.

References

1. de Bock M, Codner E, Craig ME, Huynh T, Maahs DM, Mahmud FH, et al. ISPAD clinical practice consensus guidelines 2022: Glycemic targets and glucose monitoring for children, adolescents, and young people with diabetes. Pediatr Diabetes. 2022;23:1270-1276. https://doi.org/10.1111/pedi.13455 PMid:36537523 PMCid:PMC10107615

2. American Diabetes Association Professional Practice Committee. 7. Diabetes technology: Standards of Care in Diabetes-2024. Diabetes Care. 2024;47(Suppl 1):S126-S144. https://doi.org/10.2337/dc24-S007 PMid:38078575 PMCid:PMC10725813

3. Battelino T, Danne T, Bergenstal RM, et al. Clinical targets for continuous glucose monitoring data interpretation: recommendations from the International Consensus on Time in Range. Diabetes Care. 2019;42(8):1593-1603. https://doi.org/10.2337/dci19-0028 PMid:31177185 PMCid:PMC6973648

4. Welsh JB, Zhang X, Puhr SA, et al. Performance of a factory-calibrated, real-time continuous glucose monitoring system in pediatric participants with type 1 diabetes. J Diabetes Sci Technol. 2019;13(2):254-258. https://doi.org/10.1177/1932296818798816 PMid:30198331 PMCid:PMC6399785

5. Clinical Trial Report. Continuous Glucose Monitoring System. Version Number: JT202403WMX-CSR/V1.0, Version Date June 17,2025.

6. Boscari F, Galasso S, Acciaroli G, et al. Performance of three different continuous glucose monitoring systems in children with type 1 diabetes during a diabetes summer camp. Pediatr Diabetes. 2021;22(1):e13190. https://doi.org/10.1111/pedi.13160 PMid:33219728 PMCid:PMC7984061

7. Cengiz E, Tamborlane WV. A tale of two compartments: interstitial versus blood glucose monitoring. Diabetes Technol Ther. 2009;11(Suppl 1):S11-S16. https://doi.org/10.1089/dia.2009.0002 PMid:19469670 PMCid:PMC2903977

8. Basu A, Dube S, Slama M, et al. Time lag of glucose from intravascular to interstitial compartment in humans. Diabetes. 2013;62(12):4083-4087. https://doi.org/10.2337/db13-1132 PMid:24009261 PMCid:PMC3837059

9. International Organization for Standardization. ISO 15197:2013 - In vitro diagnostic test systems: requirements for blood-glucose monitoring systems for self-testing in managing diabetes mellitus. Geneva: ISO; 2013.