FDA's New CGM Data Guidance: What Clinical Trial Sponsors Need to Know

In May 2026, FDA's Center for Drug Evaluation and Research (CDER) and Center for Biologics Evaluation and Research (CBER) jointly issued "Submitting Continuous Glucose Monitoring Data in Clinical Trials," a technical specifications document that, for the first time, lays out a standardized framework for how continuous glucose monitor (CGM) data should be structured, tabulated, and submitted in support of marketing applications for drugs and biologics. Though labeled nonbinding guidance, the document is a signal that FDA is treating CGM-derived endpoints as a maturing evidentiary category, and sponsors who plan to rely on CGM data for diabetes or metabolic disease programs should read it closely.

Why This Guidance Exists

CGM devices are digital health technologies (DHTs) that passively collect glycemic data at fixed sampling intervals, typically every one, five, or fifteen minutes. This produces far denser datasets than traditional point-in-time laboratory glucose draws, but that density creates both opportunity and risk. Sponsors can derive rich endpoints like time in range (TIR) and glucose variability2,3, but reviewers need a consistent way to trace those endpoints back to raw sensor output, and to distinguish genuine safety signals from sensor artifacts of missing or noisy data. The guidance builds on FDA's December 2023 guidance on Digital Health Technologies for Remote Data Acquisition, extending general DHT data-handling principles to the specific mechanics of glucose sensors.

What the Guidance Covers

The document maps a four-stage CGM data flow – epoch-level data, intermediate summary data, final analysis data, and "other CGM data" describing device characteristics – onto the CDISC SDTM and ADaM standards sponsors already use for other domains:

• SDTM Laboratory Test Results (LB) dataset for raw epoch-level readings, with defined values for LBSTAT and LBREASND to flag and explain missing readings.

• A standalone ADaM ADCGM dataset carrying epoch-level data forward for analysis, including new variables such as SPDEVID (sensor identifier), SENSFL (first record of a sensor session), and DTYPE = 'PHANTOM' for intermittent missing data not captured by the device itself.

• An ADaM final analysis dataset (ADCGMEN) reporting derived endpoints together with data-quality variables like Valid Epochs and Valid Percentage Expected.

• An optional ADaM paired glucose dataset (ADGLUCPR) that reconciles CGM readings against non-CGM confirmatory glucose values.

• Recommended tables for data completeness and missing-data summaries by treatment arm, plus a requirement that sponsors submit the actual source code used to derive CGM datasets.

What's Genuinely New for Sponsors

Historically, glucose data, whether from fingerstick meters or occasional CGM pilots, has been folded into the standard SDTM LB and ADaM ADLB structures alongside routine chemistry panels, with missingness handled inconsistently and often left to sponsor discretion. This guidance departs from that practice in several concrete ways:

1. A dedicated ADCGM dataset, separate from ADLB, is now expected, reflecting the sheer volume of CGM data relative to conventional labs.

2. Missing data gets a formal taxonomy for the first time: warmup periods, sensor non-wear, early termination, transmission failures, and upload failures each have recommended representations, and 'phantom' records must be added in ADaM even when the device never transmitted anything for that period.

3. Device-level traceability is elevated to a submission requirement: the FDA-required Unique Device Identifier (UDI) should be captured per sensor session and linked into the analysis datasets via SPDEVID, something rarely formalized in prior glucose-monitoring submissions.

4. Sponsors must submit source code for derivation and imputation logic, not just the final datasets: a higher transparency bar than typical ADaM submissions, where programs are described in the ADRG but not always required in full.

Implications for Clinical Trial Sponsors

Sponsors running trials with CGM-derived endpoints should revisit statistical analysis plans (SAPs), data management plans, and the Analysis Data Reviewer's Guide (ADRG) templates well before database lock. Windowing algorithms for mapping epoch-level timestamps to analysis visits must account for full timestamps (not just calendar dates) to avoid silently dropping post-midnight readings. Missing-data thresholds for 'complete' analysis windows (the guidance cites a 70% valid-epoch example) need prospective justification and sensitivity analyses. Early engagement with FDA, ideally at the End-of-Phase 2 meeting, is explicitly encouraged, particularly where sponsors intend to impute missing epochs or use a CGM model that changes mid-study.

Bottom Line

As a first-version technical specifications document, this guidance is nonbinding and explicitly non-exhaustive. But it represents a meaningful step toward standardizing how a widely used DHT modality is submitted for regulatory review, and it raises the bar on missing-data transparency and device-level traceability relative to how glucose data has traditionally been handled in SDTM and ADaM. Sponsors and vendors that begin aligning their data pipelines now will be better positioned as FDA's expectations for DHT-derived endpoints continue to mature.

About the Author

Lauren Crooks, MSc has comprehensive cross-functional experience within the life sciences and information technology sectors. At Signant, she combines this experience to provide scientific consultation and support to clients on the implementation of eCOA to optimize patient care and outcomes in clinical trials. She has supported eCOA projects across multiple therapeutic areas, including dermatology and neurology, with a current focus on respiratory diseases.  

References

1. Food and Drug Administration. (2026, May). Submitting Continuous Glucose Monitoring Data in Clinical Trials. Guidance for Industry. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/submitting-continuous-glucose-monitoring-data-clinical-trials

2. Battelino, T., Alexander, C. M, Amiel, S. A., et al. (2022). Continuous glucose monitoring and metrics for clinical trials: an international consensus statement. The Lancet Diabetes & Endocrinology, 11, 42-47.

3. Breyton, A. E., Lambert-Porcheron, S., Laville, M., et al. (2021). CGMs and Glycemic Variability, Relevance in Clinical Research to Evaluate Interventions in T2D, a Literature Review. Frontiers Endocrinology, 12, 666008.