Continuous Glucose Monitoring in Diabetes Clinical Trials: Endpoints, Regulatory Guidance, and Operational Best Practices

Introduction

Advances in continuous glucose monitoring (CGM) technology, including improvements in accuracy, usability, and device integration, have led to increasing adoption in both clinical care and research settings. Although uptake in clinical trials has historically been limited, use has steadily increased, supported by growing consensus on standardized CGM metrics and reporting.^1-3

CGM systems provide near-continuous measurement of interstitial glucose, enabling detailed characterization of glycemic patterns. Commonly derived metrics include time in range (TIR), time above and below range, glycemic variability, and glucose management indicator.

These measures complement glycated hemoglobin (HbA1c), which remains the primary endpoint for long-term glycemic control but does not capture short-term fluctuations or variability.⁴

How do CGM and BGM differ?

Blood glucose monitoring (BGM) has been available for clinical use since the 1970s, offering intermittent glucose readings. Adherence to recommended testing frequency is, however, often suboptimal, with less than 50% of patients with type 1 diabetes (T1D) achieving four or more daily checks.⁵

In clinical studies, BGM is primarily used to document hypoglycemia and guide insulin titration. BGM detects hypoglycemic events only at specific times when a fingerstick test is performed. Relying on participant adherence can lead to incomplete data and may fail to capture asymptomatic or nocturnal hypoglycemia.

CGM provides ongoing surveillance of glucose levels, recording measurements every few minutes throughout the day and night. Consequently, CGM offers a much more comprehensive record of hypoglycemic events, including asymptomatic (“silent”) events that would otherwise go undetected by traditional BGM.

This continuous data stream supports a fuller understanding of glycemic control and reduces reliance on participant-initiated testing. Regulatory guidance reflects these differences. BGM remains the preferred method for confirming individual hypoglycemic events, particularly severe (level 3) events requiring adjudication.

CGM is increasingly used to assess glycemic variability and trends across populations. In this context, CGM and BGM should be viewed as complementary: CGM enhances endpoint sensitivity and contextual understanding, whereas BGM supports confirmatory accuracy (Table 1).

Regulatory Framework

Guidance from both FDA and the European Medicines Agency (EMA) positions HbA1c as the primary endpoint for evaluating glycemic control and recognizes hypoglycemia as a crucial safety and efficacy outcome.^6,7 Severe hypoglycemia requires rigorous documentation and adjudication, with BGM measurements included when available.

Regulators increasingly acknowledge the value of CGM-derived endpoints, particularly for capturing hypoglycemia not detected through intermittent testing and for assessing glycemic variability. However, FDA has noted limitations in CGM accuracy at low glucose levels that reinforce the continued need for BGM in confirmatory contexts.

Overall, regulatory expectations emphasize integrating CGM and BGM within a framework that ensures data reliability, traceability, and interpretability. Implementing CGM in clinical trials requires careful attention to device validation, data integrity, management of data visibility (e.g., blinded versus unblinded CGM system use), and regulatory submission standards.

These requirements are essential to ensure that CGM-derived endpoints are reliable, reproducible, and acceptable to regulatory authorities.

Implementation Considerations

Study-provisioned CGM devices

The dual use of CGM for both clinical management and research data collection requires thoughtful operational and scientific considerations. CGM data used for routine care are typically managed outside of the study’s controlled environment, limiting the ability to ensure consistency, completeness, and real-time oversight.

Moreover, although Dexcom devices have a near-global footprint, other devices are also widely used globally or are popular in specific regions. This variety complicates planning and data collection for studies that rely on data captured from participants’ personal devices.

Clinical trials require CGM data to be captured, standardized, and continuously monitored within a single system to support data integrity, protocol compliance, and regulatory acceptability.⁸ Sponsors should allow participants to continue using their personal CGM for clinical care and insulin dosing and add a study-provisioned CGM for discrete data collection periods.

This dual-device strategy, although it may require participants to wear two sensors temporarily, minimizes confounding and supports regulatory compliance, particularly with 21 CFR Part 11 for electronic records.⁹

Blinding

Access to real-time glucose data may influence participant behavior and introduce bias into study outcomes. This effect may be especially pronounced in participants unfamiliar with CGM, including many individuals with type 2 diabetes who are not treated with insulin.

For these participants, CGM may be a novel source of information, and real-time glucose visibility may heighten their awareness of glucose fluctuations. Blinded CGM approaches mitigate this risk by limiting behavioral modification while preserving consistent data capture.

Blinded study-provisioned CGMs for participants with T1D who also wear their personal CGM avoid potential confusion around clinical care decisions if readings differ. Given these considerations, sponsors should provision blinded CGM devices to all participants.

In such designs, participants continue using personal CGM devices for routine care while wearing an additional study-provisioned CGM during defined collection periods. This approach minimizes disruption to clinical management while enabling standardized, centralized data capture.

Data Quality

Device Consistency

Using the same CGM device model throughout a trial maintains data consistency and comparability. Switching participants between CGM systems can introduce variability related to sensor performance characteristics (e.g., approved sensor wear times), user interaction, and learning effects, which may affect adherence and data quality.

Global trials must also account for regional regulatory requirements, device availability, and logistical constraints. Differences in market approval status or device support can affect study feasibility and data standardization if these differences are not addressed during study planning.

Data Integrity

Regulatory frameworks such as FDA 21 CFR Part 11 and EMA guidance require validated systems that provide secure, auditable data handling across the full data lifecycle.^9,10 In practice, these requirements include structured processes for regular data review, real-time visibility into data completeness, and clear procedural documentation for CGM-specific activities, such as sensor insertion, calibration (where applicable), and data synchronization.

These processes minimize the risk of data loss or corruption. Achieving these standards requires more than conceptual alignment of CGM with electronic patient-reported outcome (ePRO) systems.

It requires well-defined, validated data architectures. These should encompass device connectivity, workflow synchronization, and a centralized environment.

Such integration is essential to ensure that data remain complete, consistent, and contemporaneous in accordance with ALCOA++ expectations.¹¹ Fragmented or externally managed data sources may compromise completeness and auditability, increasing regulatory risk.

Early alignment with these requirements during study design can reduce downstream regulatory risk and streamline submission review.

Compliance and Training

Compliance

Participant compliance with CGM wear is a key determinant of data quality. Because CGM data are collected at high frequency, gaps in wear time or delays in synchronization can substantially affect endpoint reliability.

Consensus recommendations suggest that CGM data should be collected over defined intervals (e.g., at least 14 days per period) with a minimum completeness threshold (e.g., ≥70%) to ensure interpretability.² Operationally, CGM sensors have a defined wear period and must be replaced at regular intervals.

This requirement can increase participant burden and create opportunities for missed replacements. As a result, data collection may be interrupted and completeness reduced, particularly in longitudinal studies.

Regulatory expectations for data integrity, as described in FDA’s 2018 current good manufacturing practice (CGMP) guidance, require that “all data be reliable and accurate ... CGMP regulations and guidance allow for flexible and risk-based strategies to prevent and detect data integrity issues.”¹² Mitigation strategies should include automated monitoring of wear-time adherence, alerts for delayed synchronization or nonwear, and participant reminders to maintain continuous data capture.

Finally, regulatory guidance highlights the importance of ongoing oversight, noting that systems should enable timely detection of data issues and support audit readiness. Centralized, near-real-time monitoring at the site and sponsor level facilitates early identification of missing or inconsistent data, enabling prompt intervention.

Training

Participant training is fundamental to compliance and data integrity. FDA guidance on digital health technologies emphasizes ensuring that participants are adequately trained to use digital health technologies in accordance with protocol requirements.¹³

Training should include device use, data entry workflows, troubleshooting procedures, and compliance expectations. Participants should also understand key protocol requirements, including the correct timing of symptom reporting relative to glucose measurements.

Standardized training delivered at a global scale, such as interactive on-device videos or study-specific guides for participants, supports consistency in data reporting, compliance, and accuracy. By reinforcing study expectations, improving ePRO reporting, and providing accessible resources throughout participation, these tools can enhance engagement, reduce variability, and improve the overall reliability of study findings.

Integration of ePRO and CGM

Integration of ePRO with CGM enhances the interpretability of glycemic data by providing contextual information on symptoms, behaviors, and external factors. Effective integration requires alignment of data capture timing and structured workflows that support reconciliation across data streams.

Automated alerts based on ePRO or CGM data can further support participant safety and data quality by enabling timely site intervention. These capabilities strengthen both clinical oversight and endpoint reliability by notifying sites when a participant’s reported data indicate potential issues, such as frequent hypoglycemic events, or when a participant has not submitted data for a defined period.

Integration of ePRO and CGM data also enables the development of cross-data quality metrics, such as comparing participant-reported hypoglycemic symptoms with corresponding CGM glucose values and flagging discordant findings. Furthermore, ePRO measures can be designed to capture contextual factors, such as meal intake and physical activity, which provide insight into potential drivers of observed glucose fluctuations.

Conclusion

CGM has substantially advanced the ability to capture detailed, continuous glycemic data in clinical trials. Regulatory agencies support its use for assessing glycemic variability and hypoglycemia and maintain HbA1c as the primary endpoint and BGM as the standard for event confirmation.

Successful implementation requires careful attention to study design, particularly with respect to blinding, data standardization, and centralized monitoring. Alignment with data integrity principles, robust system integration, and comprehensive participant training are crucial to ensuring that CGM data are reliable and suitable for regulatory evaluation.

As adoption continues to grow, CGM is expected to play an increasingly important role in clinical trial design, enabling more precise endpoint assessment and supporting advances in diabetes research.

About the Authors

• Dr. Réka Balogh is a Scientific Advisor within the eCOA Clinical Science and Consulting team at Clario, part of Thermo Fisher Scientific. Her expertise spans electronic clinical outcome assessments (eCOAs), clinical outcome strategy, and patient-centered measurement. With a background in neuropsychology and neurodegenerative disorders, she has extensive experience in patient assessment methodologies and the measurement of cognitive, behavioral, and patient-reported outcomes. Dr. Balogh earned her PhD in Clinical Medicine, specializing in Clinical and Experimental Neuroscience, from the Doctoral School of Clinical Medicine at the University of Szeged.
• Dr. Brian Rundle is a neuropsychologist and Scientific Advisor within the eCOA Science & Consulting team at Clario, part of Thermo Fisher Scientific, where he helps pharmaceutical and biotechnology sponsors optimize the collection of patient-centered and cognitive outcomes data in clinical trials. His work focuses on the design and implementation of research technology that balances regulatory compliance, scientific rigor, and intuitive user experiences to maximize data quality and participant engagement. He received his Ph.D. from Baylor University and completed a post-doc in neurology at the David Geffen School of Medicine at UCLA. With expertise spanning neuroscience, cognitive assessment, metabolic disorders, and connected devices, Brian works closely with sponsors to develop fit-for-purpose solutions for complex clinical research programs.
• Dr. Charles M. Alexander is an endocrinologist who has more than 40 years of experience in diabetes, both direct patient care and research. Since 2016, he has provided consulting services for biopharmaceutical and medical device companies. He has also been helping The diaTribe Foundation since 2016 and serving as its Scientific and Medical Advisor since 2021. He has been a member of Penn Medicine’s Institute for Diabetes, Obesity and Metabolism (IDOM) Leadership Council since 2015. Dr. Alexander retired from Merck in 2015 after 23 years in both Outcomes Research and Medical Affairs. Roles at Merck included Global Medical Director for Diabetes in Medical Affairs 2008-2015 and Senior Director Outcomes Research from 1996-2008 and Director in Medical Affairs 1992-1996. Prior to joining Merck, he practiced diabetes, endocrinology, and internal medicine in Los Angeles for ten years and was Clinical Professor of Medicine at the Keck School of Medicine of the University of Southern California (KSOM/USC). He is a Past President of the Academy of Pharmaceutical Physicians and Investigators (APPI). In addition, he also served as a volunteer consultant in diabetes and endocrinology at the HealthLink Medical Clinic, Southampton, Pennsylvania 2010-2015. He was a member of the board of directors of the Eastern Pennsylvania & Delaware Chapter of the American Diabetes Association 2012-2018. Dr. Alexander received his medical degree from KSOM/USC where he also did his postgraduate training in internal medicine and diabetes/endocrinology. He is the author of numerous scientific articles and a frequent reviewer for medical journals.
• Dr. Jessica Emerson is a behavioral scientist with over 15 years of experience in health science research focusing on the role of technology in measuring and intervening on health behaviors and symptoms. She has extensive experience in electronic clinical outcome assessment (eCOA) design, data management and analysis across phase I-IV clinical trials. On the eCOA Clinical Science and Consulting team at Clario, part of Thermo Fisher Scientific, she has supported sponsors and CROs in eCOA design best practices to optimize the reliability and validity of patient-reported electronic data for over 40 global clinical trials in the diabetes and weight loss space alone. Dr. Emerson earned her PhD from Brown University’s School of Public Health and received her postdoctoral training in behavioral medicine focused on obesity, diabetes and related health behaviors at the Weight Control & Diabetes Research center at the Warren Alpert Medical School of Brown University.

References

1. 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. doi:10.2337/dci19-0028
2. Battelino T, Alexander CM, Amiel SA, et al. Continuous glucose monitoring and metrics for clinical trials: an international consensus statement. Lancet Diabetes Endocrinol. 2023;11(1):42-57. doi:10.1016/S2213-8587(22)00319-9
3. Fox BQ, Benjamin PF, Aqeel A, et al. Continuous glucose monitoring use in clinical trials for on-market diabetes drugs. Clin Diabetes. 2021;39(2):160-166. doi:10.2337/cd20-0049
4. Diabetes Control and Complications Trial Research Group; Nathan DM, Genuth S, Lachin J, et al. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med. 1993;329(14):977-986. doi:10.1056/NEJM199309303291401
5. Moström P, Ahlén E, Imberg H, Hansson PO, Lind M. Adherence of self-monitoring of blood glucose in persons with type 1 diabetes in Sweden. BMJ Open Diabetes Res Care. 2017;5(1):e000342. doi:10.1136/bmjdrc-2016-000342
6. US Food and Drug Administration. Diabetes Mellitus: Efficacy Endpoints for Clinical Trials Investigating Antidiabetic Drugs and Biological Products; Draft Guidance for Industry. May 2023.
7. European Medicines Agency. Draft Guideline on Clinical Investigation of Medicinal Products in the Treatment or Prevention of Diabetes Mellitus. CPMP/EWP/1080/00 Rev. 2. 2018.
8. US Food and Drug Administration. Submitting Continuous Glucose Monitoring Data in Clinical Trials; Guidance for Industry. May 2026.
9. Electronic Records; Electronic Signatures. 21 CFR §11 (2003).
10. European Medicines Agency. Guideline on Computerised Systems and Electronic Data in Clinical Trials. EMA/INS/GCP/112288/2023. 2023.
11. European Commission. EudraLex Volume 4: EU Guidelines for Good Manufacturing Practice for Medicinal Products for Human and Veterinary Use; Chapter 4: Documentation. 2025.
12. US Food and Drug Administration. Data Integrity and Compliance With Drug CGMP: Questions and Answers; Guidance for Industry. December 2018.
13. US Food and Drug Administration. Digital Health Technologies for Remote Data Acquisition in Clinical Investigations; Guidance for Industry. December 2023.