No continuous glucose monitor (CGM) is FDA-cleared for use in hospitals. As of October 2026, Dexcom can supply its CGMs to US hospitals under FDA enforcement discretion, while FDA reviews Dexcom’s marketing submission for a dedicated Hospital CGM System. Clinical guidelines have moved faster: the Endocrine Society suggests real-time CGM, confirmed with bedside finger-stick testing, for some hospitalized adults at high risk of hypoglycemia.
That gap between clinical interest and regulatory clearance is where the hospital CGM market stands. This guide covers the current FDA status, how hospitals monitor glucose today, what inpatient studies show, and the problems device and software makers still have to solve.
Is CGM FDA-approved for hospital use?
No. Dexcom states that its CGM systems “are indicated for home use and have not been FDA-cleared or approved for use in hospital settings.” Hospital use rests on a series of FDA decisions that began during the COVID-19 pandemic.
Date | What happened |
|---|---|
April 2020 | FDA told Dexcom it would not object to hospitals using its CGMs during the COVID-19 public health emergency. Abbott announced similar FDA permission for FreeStyle Libre 14 Day the same month |
March 2022 | FDA granted Breakthrough Device Designation to Dexcom CGM for use in the hospital setting |
November 7, 2023 | The public health emergency transition period ended. FDA said it would not object to Dexcom continuing to supply hospitals until it decides on Dexcom’s Hospital CGM System submission |
October 2026 | No CGM has a hospital indication. Dexcom’s submission is still pending, according to its hospital page |
Two points often get confused. First, FDA’s COVID-era enforcement policy for non-invasive remote monitoring devices did not cover CGMs; hospital CGM use was handled through separate FDA communications with the manufacturers. Second, FDA’s position covers Dexcom’s distribution to hospitals, but it “does not expand Dexcom’s current authorized indications for use,” in the company’s words. Abbott’s current US hospital program focuses on starting patients on FreeStyle Libre at discharge.
How hospitals monitor glucose today
Most hospitals still rely on point-of-care (POC) blood glucose testing: a finger-stick reading on a bedside meter. Patients who are eating are typically tested before meals and at bedtime, and those on intravenous insulin may be tested as often as hourly. Connected meters can send results directly to the EHR, reducing transcription errors.
The need is large. In an analysis of more than 49 million POC readings from 575 US hospitals (Swanson et al., Endocrine Practice, 2011), glucose was above 180 mg/dL on 32.2% of ICU patient-days and 32.0% of non-ICU patient-days. Hypoglycemia below 70 mg/dL occurred on 6.3% and 5.7% of patient-days. The data are from 2009, but they show that high and low glucose are as common on general wards as in the ICU, where testing is far more frequent.
Finger-stick testing has a basic limitation: it only captures the moments it is performed. Lows between checks, especially overnight, can go unnoticed.
What inpatient CGM studies show
The evidence for CGM in non-critically ill hospital patients has grown since 2020. Four studies stand out:
Study | Design | Key findings |
|---|---|---|
FreeStyle Libre Pro vs POC in hospitalized adults with type 2 diabetes on basal-bolus insulin | POC detected glucose below 70 mg/dL in 14% of patients; CGM detected it in 56%. Below 54 mg/dL: 4.1% vs 36%. CGM MARD was 14.8% | |
Dexcom G6 accuracy in 218 insulin-treated, non-critically ill inpatients | MARD of 12.8% across 4,067 matched pairs; 98.7% of readings in Clarke error grid zones A and B. Accuracy was lower during hypoglycemia and severe anemia | |
Randomized trial of CGM-guided vs POC-guided insulin dosing in 185 general medicine and surgery patients | Time in range was similar (54.5% vs 48.6%, not significant). Among patients with hypoglycemia, the CGM group had fewer repeat episodes (1.80 vs 2.94 per patient) and less time below 70 mg/dL (1.89% vs 5.47%) | |
A hospital-wide CGM policy with EHR-integrated validation for insulin dosing | Showed that integrating CGM data into the EHR for inpatient dosing is feasible |
The pattern is consistent. CGM finds far more hypoglycemia than scheduled finger-sticks, it is accurate enough for trend monitoring in most patients, and it can reduce repeat lows. In the Spanakis trial, it did not significantly improve overall time in range, and its accuracy drops in exactly the situations where it matters most.
Guidelines reflect that balance. The Endocrine Society’s 2022 guideline suggests “real-time continuous glucose monitoring (CGM) with confirmatory bedside point-of-care blood glucose (POC-BG) monitoring for adjustments in insulin dosing” for adults with insulin-treated diabetes who are hospitalized for noncritical illness and at high risk of hypoglycemia, “in hospital settings where resources and training are available.” It is a weak recommendation based on low-certainty evidence.
Limitations device makers must solve
Each limitation below is also a product requirement for whoever builds the first hospital-indicated CGM, or the software around it.
Limitation | What is known | What it means for product design |
|---|---|---|
Accuracy at the extremes | Inpatient accuracy drops during hypoglycemia and severe anemia (Davis 2021) | Validate in the hospital population, including low glucose and anemic patients; design confirmatory POC workflows |
Interfering substances | Dexcom lists hydroxyurea and acetaminophen above 1 gram every 6 hours as causing falsely high readings. Abbott lists vitamin C above 1,000 mg per day for FreeStyle Libre 2 Plus and 3 Plus. A 2017 pilot study in 19 healthy adults found possible interference from lisinopril, albuterol, acetaminophen, and atenolol with older sensors | Test against drugs common in hospitals; link medication orders to CGM alerts in software |
Procedures and imaging | Labeling requires removing sensors for some procedures. Abbott, for example, requires removal before diathermy but allows 1.5T and 3T MRI for its Plus sensors under set conditions | Build procedure-specific guidance into nursing workflows and order sets |
Nursing workflow | Staff must place sensors, interpret trends, respond to alerts, and know when to confirm with a finger-stick | Simple insertion, clear alert design, and training built into the product |
Data and monitoring | Hospital EHR integration is feasible (Lee 2024), Abbott announced Libre data integration with Epic in April 2025, and Dexcom’s Epic integration is labeled for outpatient use | Central monitoring dashboards, EHR integration, and audit trails designed for inpatient units |
Cost | CGM adds a per-patient device cost, while the savings come from staff time and avoided harm | Make the value case with nursing time and hypoglycemia data |
Many patients also arrive wearing their own CGM, insulin pump, or automated insulin delivery (AID) system. Hospitals need policies on whether to continue them, and device makers need to support that transition safely.
Where the opportunities are
The first hospital indication will set the bar, but the bigger opportunity may be in the software and services around the sensor:
- Hospital-grade validation. Clinical evidence in inpatient populations, including the critically ill, is what a hospital indication will require.
- Central monitoring. Nurses need one view of every CGM patient on a unit, with alerts routed to the right person.
- EHR integration. CGM data, POC confirmations, and insulin doses should live in the same record. See why EHR interoperability matters and how diabetes device interoperability standards apply.
- Dosing decision support. Software that recommends insulin doses from CGM data is a regulated device function, so plan the regulatory path early. See our guide to FDA rules for diabetes apps.
- Transitions of care. Starting CGM in the hospital and continuing it at home connects inpatient care to remote patient monitoring for diabetes.
For the sensor technology itself, see our guide to continuous glucose monitoring technology.
Frequently asked questions
Is CGM FDA-approved for hospital use?
No. As of October 2026, no CGM is FDA-cleared for use in hospitals. FDA has said it will not object to Dexcom supplying its CGMs to hospitals until it decides on Dexcom’s Hospital CGM System submission, but that does not expand the product’s indications for use.
Can patients wear their CGM in the hospital?
It depends on the hospital’s policy. Some hospitals provide or allow CGMs, often with fingerstick checks to confirm readings before insulin doses. Some procedures and imaging require removal of the sensor, according to the manufacturer’s labeling.
How accurate are CGMs in hospitalized patients?
In a pooled analysis of 218 non-critically ill inpatients, the Dexcom G6 had a mean absolute relative difference (MARD) of 12.8%. A study of FreeStyle Libre Pro in hospitalized adults with type 2 diabetes found a MARD of 14.8%. Accuracy is lower during hypoglycemia and severe anemia.
Do hospitals still need finger-stick testing if patients use CGM?
Yes, in most cases. The Endocrine Society’s 2022 guideline pairs real-time CGM with confirmatory bedside point-of-care glucose testing for insulin dosing decisions.
What affects CGM accuracy in the hospital?
Low glucose levels, severe anemia, and certain medications can affect readings. Dexcom lists hydroxyurea and high-dose acetaminophen as interferents, and Abbott lists vitamin C above 1,000 mg per day as an interferent for its FreeStyle Libre 2 Plus and 3 Plus sensors.
Build hospital-ready CGM software with Sequenex
Hospital CGM will be won by the companies that solve accuracy, workflow, and data together, not by the sensor alone. Getting there takes software built for clinical environments and regulatory review from the start.
Sequenex builds software for diabetes and connected devices, from sensor apps to clinical data platforms. See our biosensor and CGM software development services, or contact us to discuss your project.
