One in Four Hospital Beds, Zero FDA Clearance: The Inpatient CGM Question No One Wants to Answer Out Loud

A regulatory and clinical reality check for hospital leaders, quality teams, and the device industry that serves them.

About one in four people in a U.S. hospital bed has diabetes or hyperglycemia.¹ Almost none of the continuous glucose monitors used to manage them in 2026 is actually FDA-cleared for that setting.²

The COVID-era enforcement discretion that briefly made hospital CGM use legitimate expired on November 7, 2023.³ Two and a half years later, the devices are still on patients — sometimes brought from home, sometimes applied at the bedside, sometimes integrated (informally) into nursing workflows. The clinical literature now describes how to do it well. The 2026 ADA Standards of Care describe when to continue it. The FDA, meanwhile, has not cleared a single CGM for inpatient use, although Dexcom holds a Breakthrough Device Designation for that pathway granted in March 2022.⁴

That gap — between what is happening at the bedside and what the regulatory file actually authorizes — is the question this post is about. It is not a gotcha. The clinicians using CGMs in hospitals are doing thoughtful work, the manufacturers are investing real resources in the inpatient evidence base, and the FDA is navigating a genuinely novel device-context problem. But the gap is wide enough to deserve a serious conversation, particularly now that the same manufacturers have spent the last 18 months working through accuracy-related warning letters and Class I recalls in their cleared indications.

What the inpatient evidence actually shows

CGM performance in the hospital is not a single number. It is a distribution shaped by patient population, comparator method, sensor generation, and clinical context.

A 2024 scoping review of CGM accuracy in the ICU found a mean absolute relative difference (MARD) ranging from 6.6% to 30.5% across all subcutaneous CGM studies.⁵ For newer factory-calibrated systems, the range narrowed to 9.7% to 20.6%.⁵ Intra-arterial and intravenous CGM systems performed tighter still (5%–14.2% and 6.4%–13%, respectively),⁵ but those are not the devices most U.S. hospitals are actually deploying.

More recent prospective work has been more reassuring within defined populations:

• A 2024 study of 40 ICU patients on mechanical ventilation, IV insulin, and

  vasopressors using the Dexcom G6 reported an overall MARD of 12.7%, with 99.8% of readings in Clarke Error Grid zones A or B; accuracy improved over time and reached MARD <14% after the first 36 hours.⁶

• A 2025 observational cohort of 103 critically ill patients (97% mechanically ventilated, 92% on vasopressors, 25% on renal replacement therapy) reported a median MARD of 11.23% with no clinical hypoglycemia in patients with sensors in place.⁷

• A 2025 Italian study of FreeStyle Libre in critically ill patients found reproducible analytical and clinical accuracy over four days, although the use of vasopressors was associated with a higher MARD than baseline.⁸

• Earlier MICU work using subcutaneous CGM reported MARD as low as 6.6% versus capillary glucose and 8.8% versus plasma glucose, with no significant difference across vasopressor subgroups.⁹

  
  

The honest synthesis: in stable medical-surgical patients and in many ICU patients, modern CGMs perform within or near the ±20%/±20 mg/dL "20/20" criterion that the 2026 ADA Standards of Care identify as the validation threshold for hospital use.² Performance degrades — sometimes substantially — in specific scenarios: high-dose vasopressors, the operating room (electrocautery interference and signal loss observed in roughly 60% of patients in one CABG series),¹⁰ anasarca, and the first day or two after sensor insertion. None of that is disqualifying. All of it is workflow-relevant.

The regulatory backdrop, in four moves

Move one — April 2020. The FDA stated it would not object to inpatient CGM use during the COVID-19 public health emergency, framing it as a way to reduce healthcare-worker exposure and conserve PPE.¹¹

Move two — March 2022. Dexcom received Breakthrough Device Designation for hospital CGM use, with the company citing data from more than 800 patients across pandemic-era clinical experience.⁴

Move three — November 7, 2023. The COVID enforcement discretion formally expired. Hospital CGM use in the U.S. became, technically, off-label — although guideline bodies have been reasonably clear that continuing CGM in patients who already use one is appropriate when supported by institutional protocols.³

Move four — October 25, 2024. The American Diabetes Association published a consensus statement, Consensus Considerations and Good Practice Points for Use of Continuous Glucose Monitoring Systems in Hospital Settings, in Diabetes Care.¹² The document, led by Dr. Julie L.V. Shaw of The Ottawa Hospital, explicitly frames hospital CGM use as a clinician–laboratory collaboration. It addresses calibration, comparator-method selection, and decision-support workflows — the unglamorous infrastructure that determines whether a CGM number on a screen actually drives the right insulin dose.

The 2026 ADA Standards of Care have now folded this thinking forward, recommending continued use of personal CGM during hospitalization "when clinically appropriate" with point-of-care confirmation under an institutional protocol, and adopting the 20/20 validation criterion as the working accuracy benchmark.²

So the rules of the road are reasonably clear at the guideline level. They are not yet clear at the clearance level. That distinction matters when something goes wrong.

What changed in 2025 that compliance teams should sit with

The last 18 months have produced an unusual cluster of accuracy-related FDA actions in the cleared, ambulatory CGM indication. None of these actions is about hospital use. All of them are worth thinking about because of hospital use.

Dexcom received an FDA Warning Letter (CMS #700835) on March 4, 2025, following inspections of its San Diego and Mesa facilities. The agency cited 21 CFR 820.75(b) concerns tied to functional acceptance testing on the Dexcom G6 and G7 sensors — specifically glucose sensitivity slope, accuracy in the presence of acetaminophen (G6), and low-oxygen response. A central concern was a design change to a component used in the resistance layer of the sensors that, per the warning letter, was not adequately evaluated and validated; an associated clinical study did not meet its primary equivalence endpoints.¹³ The warning letter did not restrict Dexcom's ability to manufacture, market, or seek 510(k) clearance, and Dexcom has continued to advance its product roadmap, including FDA clearance of a 15-day G7 sensor in April 2025.¹⁴

Abbott received a warning letter dated January 23, 2026, posted to the FDA website on February 3, 2026, following an October 2025 inspection of its Alameda, California diabetes care site. The letter alleged, among other things, that Abbott did not transfer specific glucose sensor accuracy requirements to third-party manufacturers and released finished FreeStyle Libre CGMs without first testing them for accuracy after full assembly, sterilization, and programming.¹⁵ This action followed a Class I recall (classified February 4, 2026) of certain FreeStyle Libre 3 and 3 Plus sensors that may produce falsely low glucose readings; as of January 7, 2026, Abbott had reported 860 serious injuries and seven deaths globally associated with the issue, with the seven fatalities occurring outside the United States.¹⁶

A cross-cutting FDA Safety Communication on February 5, 2025 flagged that smartphone-delivered CGM and insulin-pump alerts are not always reaching users as intended, with serious adverse outcomes — including severe hypoglycemia, severe hyperglycemia, DKA, and death — reported in connection with missed alerts.¹⁷

I want to be careful here, because none of this means CGMs do not work. They do, every day, for millions of people. The manufacturers cited above have invested deeply in remediation, and the FDA's actions in this period have been notably focused on process controls and reporting — design change validation, accuracy verification at finished-device release, MDR timeliness — rather than on systemic device failure. That is an important distinction. But the ambulatory indication is the one where these companies have decades of post-market experience, the most mature complaint pipelines, and patients who can sanity-check a number with a fingerstick or a glass of juice.

The hospital indication has none of those buffers. The patient is often sedated, on a vasopressor, and being titrated by an IV insulin protocol. The "user" sanity-checking the CGM number is a nurse who may be carrying eight other patients.

The questions worth a real conversation

If you sit at the intersection of regulatory affairs, hospital quality, and diabetes technology — which a fair number of readers do — here is what I think is genuinely worth debating:

1. What does "off-label, but standard of care" mean for an enterprise risk register? Joint Commission surveyors, hospital legal teams, and device manufacturers each have different stakes in that phrase, and the answers are not aligned.

2. Should institutional CGM protocols specify performance disqualifiers explicitly? The literature is reasonably clear that high-dose vasopressors, the OR environment, severe anasarca, and the first 24–36 hours after insertion are higher-risk windows. Most institutional protocols are silent on these.⁶ ⁷ ¹⁰

3. How should hospitals respond when a CGM that a patient brought from home is the subject of an active recall or warning letter? The Abbott Libre 3 falsely-low-reading recall is the live test case. There is no consensus answer.¹⁶

4. Where does the comparator-method conversation actually live? The 2024 ADA consensus is clear that whole-blood capillary, venous, and arterial samples are not interchangeable comparators, and that timing from sample collection to measurement matters.¹² In practice, hospital CGM accuracy is often informally assessed against whichever POC meter is on the floor. That is not the same conversation.

5. What is the regulatory off-ramp? Breakthrough Device Designation is a pathway, not a clearance. The longer the gap stretches between a useful clinical practice and a formally authorized indication, the more the burden of harm-when-it-occurs sits with hospitals and clinicians rather than with the device file.

None of these questions has a clean answer. All of them, I think, are better answered together than alone.

Closing

The story of CGM in the hospital is, on balance, a story of clinical innovation outrunning regulatory architecture in a way that has — so far — produced more good than harm. The patients I have heard about who avoided severe hypoglycemia because a CGM alarm woke a night nurse are real. So are the patients who got hurt because a sensor read 80 points low at 3 a.m. Both stories are true at the same time.

The regulatory question is not whether to use these devices in hospitals. The 2024 ADA consensus and the 2026 ADA Standards of Care have effectively answered that.² ¹² The question is what the quality system around inpatient CGM use should look like — at the hospital, at the manufacturer, and at the agency — when the underlying device is technically off-label and the cleared indication has spent 2024 and 2025 working through its own accuracy-validation issues.

References

1. Shaw JLV, Bannuru RR, Beach L, et al. Consensus Considerations and Good Practice Points for Use of Continuous Glucose Monitoring Systems in Hospital Settings. Diabetes Care. 2024;47(12):2062–2075. https://doi.org/10.2337/dci24-0073

2. American Diabetes Association Professional Practice Committee. 16. Diabetes Care in the Hospital: Standards of Care in Diabetes—2026. Diabetes Care. 2026;49(Suppl. 1):S339–S358. https://diabetesjournals.org/care/article/49/Supplement_1/S339/163925/

3. Tucker ME. American Diabetes Association Advises on Hospital CGM Use. Medscape Medical News. October 28, 2024. https://www.medscape.com/viewarticle/american-diabetes-association-advises-hospital-cgm-use-2024a1000jrp

4. DexCom, Inc. FDA Grants Breakthrough Device Designation for Dexcom Hospital CGM System. Press release, March 1, 2022. https://investors.dexcom.com/news/news-details/2022/FDA-Grants-Breakthrough-Device-Designation-for-Dexcom-Hospital-CGM-System/

5. Tsur N, et al. Accuracy of continuous glucose monitoring systems in intensive care unit patients: a scoping review. J Diabetes Sci Technol. 2024. PMID: 39417874. https://pubmed.ncbi.nlm.nih.gov/39417874/

6. Friman R, et al. Performance of Subcutaneous Continuous Glucose Monitoring in Adult Critically Ill Patients Receiving Vasopressor Therapy. Diabetes Technol Ther. 2024. https://www.liebertpub.com/doi/full/10.1089/dia.2024.0035

7. Brown JE, Bellomo R, et al. Accuracy of continuous glucose monitoring in critically ill patients. Aust Crit Care. 2025. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12554099/

8. Continuous interstitial glucose monitoring in diabetic and non-diabetic critically ill patients is simple and accurate: comparison with venous, arterial and capillary glucose measurements. Acta Diabetologica. 2025. https://link.springer.com/article/10.1007/s00592-025-02531-1

9. The Accuracy of Continuous Glucose Monitoring in the Medical Intensive Care Unit. https://pmc.ncbi.nlm.nih.gov/articles/PMC9631519/

10. Perez-Guzman MC, Shang T, Zhang JY, Jornsay D, Klonoff DC. Continuous Glucose Monitoring in the Hospital. Endocrinol Metab (Seoul). 2021. https://www.e-enm.org/journal/view.php?number=2146

11. Implementing Inpatient Continuous Glucose Monitor Use During COVID-19. US Pharmacist. November 13, 2020. https://www.uspharmacist.com/article/implementing-inpatient-continuous-glucose-monitor-use-during-covid19

12. American Diabetes Association. New ADA Consensus Report: Consensus Considerations and Good Practice Points for Use of CGMs in Hospital Settings. October 25, 2024. https://diabetes.org/newsroom/press-releases/new-ada-consensus-report-consensus-considerations-and-good-practice-points

13. U.S. Food and Drug Administration. Warning Letter: Dexcom, Inc. (MARCS-CMS 700835), March 4, 2025. https://www.fda.gov/inspections-compliance-enforcement-and-criminal-investigations/warning-letters/dexcom-inc-700835-03042025

14. Reuter E. Dexcom nets FDA clearance for 15-day CGM. MedTech Dive. April 10, 2025. https://www.medtechdive.com/news/dexcom-fda-clearance-15-day-cgm/745051/

15. Reuter E. Abbott receives FDA warning letter over Freestyle Libre CGMs. MedTech Dive. February 3, 2026. https://www.medtechdive.com/news/abbott-receives-fda-warning-letter-over-freestyle-libre-cgms/811262/ (Warning letter dated January 23, 2026; FDA Warning Letters index: https://www.fda.gov/inspections-compliance-enforcement-and-criminal-investigations/compliance-actions-and-activities/warning-letters)

16. U.S. Food and Drug Administration. Glucose Monitor Sensor Recall: Abbott Diabetes Care Removes Certain FreeStyle Libre 3 and FreeStyle Libre 3 Plus Sensors. February 4, 2026. https://www.fda.gov/medical-devices/medical-device-recalls-and-early-alerts/glucose-monitor-sensor-recall-abbott-diabetes-care-removes-certain-freestyle-libre-3-and-freestyle

17. U.S. Food and Drug Administration. Safety Communication: Risk of Missing Smartphone Alerts on Continuous Glucose Monitors, Insulin Pumps, and Automated Insulin Dosing Systems. February 5, 2025.

    
    

The views above are mine and are intended to provoke discussion among regulatory affairs, quality, and clinical colleagues. They are not legal or clinical advice for any specific institution or patient. Mentions of specific manufacturers reflect publicly reported regulatory actions and peer-reviewed clinical evidence, and are not intended as criticism of the substantial work those organizations are doing to advance diabetes care.