Continuous Glucose Monitor FDA Recalls & MAUDE: Dexcom & Abbott Teardown

Continuous Glucose Monitors (CGMs) have revolutionized diabetes management, transitioning from niche clinical tools to mainstream consumer health technologies. However, as their clinical utility grows—particularly in closed-loop Automated Insulin Delivery (AID) systems—so does the regulatory and safety scrutiny. For regulatory affairs specialists, quality managers, and healthcare technology management (HTM) teams, understanding the failure modes, recalls, and adverse events of these systems is crucial for risk management and post-market surveillance.

This article provides a detailed teardown of CGM recalls and FDA MAUDE (Manufacturer and User Facility Device Experience) database reports. We examine recent high-profile recall events from industry giants Dexcom and Abbott, identify their underlying failure modes, analyze historical recall trends, and extract raw metrics from the FDA's adverse event registries.

---

Scenario Question & Direct Answer

Question: Which continuous glucose monitor recalls and MAUDE failures actually matter—who recalled, what failed, how many units and injuries/deaths, and is it Class I or II?

Direct Answer: CGMs are under intense post-market scrutiny: in 2025–2026, the FDA classified Abbott FreeStyle Libre 3 and Libre 3 Plus (sensor inaccuracy; 860 injuries, 7 deaths) and Dexcom G6/G7/ONE/ONE+ receivers (speaker/alarm failure; 602,445 G7 + 36,824 G6 units) as Class I recalls, with a Dexcom G7 watchOS-app Class II recall in April 2026. Across CGM product codes QBJ and QLG, MAUDE logged 382,279 device-rows in 2024 (comprising 364,251 malfunctions, 17,987 injuries, and 30 deaths); the openFDA recall database holds 33 CGM recalls, dominated by Dexcom (20 recalls) and Abbott (8 recalls).

---

Major CGM Recalls (2024–2026) and What Failed

In recent years, both Abbott and Dexcom have initiated voluntary recalls that the FDA classified as Class I—the most serious recall category, reserved for situations where there is a reasonable probability that the use of or exposure to a violative product will cause serious adverse health consequences or death.

Below is a comparative matrix of the major CGM recalls from 2024 through mid-2026:

Recalling Manufacturer	Affected Device Model(s)	FDA Recall Class	Estimated Units Affected	Primary Failure Mode / Cause	Reported Injuries & Deaths
Abbott Diabetes Care	FreeStyle Libre 3 & Libre 3 Plus Sensors (Specific Lots)	Class I (Classified Feb 4, 2026)	Specific Lot Numbers	Incorrect low glucose readings due to sensor calibration issues, potentially leading to missed hyperglycemic alerts or inappropriate insulin administration.	860 serious injuries, 7 deaths (reported as of Jan 7, 2026)
Dexcom, Inc.	Dexcom G7 Receiver	Class I (Classified 2025)	602,445 units	A physical/hardware speaker malfunction where the audio alarm speaker becomes silent, preventing the user from hearing critical high/low glucose alarms.	0 deaths, specific injuries related to missed alerts
Dexcom, Inc.	Dexcom G6 Receiver	Class I (Classified 2025)	36,824 units	Physical speaker failure caused by a foam spacer degradation that dampens or silences audio alarms.	0 deaths, specific injuries related to missed alerts
Dexcom, Inc.	Dexcom G7 watchOS Mobile Application	Class II (Initiated Apr 14, 2026)	Software users	A software defect in the Apple Watch companion application (watchOS) where the app could crash or fail to display glucose updates in real time, interrupting critical alert sequences.	0 deaths, Class II status (lower immediate clinical risk)

These recall events illustrate that CGM vulnerabilities span the entire product design. They affect the physical sensor chemistry (Abbott's calibration issues), the hardware display units (Dexcom's receiver speakers), and the software interface (Dexcom's watchOS app).

---

Deep-Dive: Abbott FreeStyle Libre 3 Calibration Failures

Abbott’s recall of specific lots of FreeStyle Libre 3 and Libre 3 Plus sensors, classified as Class I by the FDA in early 2026, represents a significant case study in electrochemical sensor vulnerabilities.

The Electrochemical Underpinnings

The FreeStyle Libre 3 uses a wire-based subcutaneous sensor that operates on amperometric glucose detection. The sensor is coated with glucose oxidase (GOx) or glucose dehydrogenase (GDH) enzymes. As glucose in the interstitial fluid reacts with the enzyme, it transfers electrons via a mediator to the working electrode, generating an electrical current (in nanoamperes) directly proportional to the glucose concentration.

Interstitial Glucose + O2  --[GOx Catalyst]-->  Gluconic Acid + H2O2
H2O2  --[Electrode Potential]-->  O2 + 2H+ + 2e- (Measured Current)

In the recalled lots, manufacturing process inconsistencies or localized packaging seal degradation led to:

1. Enzyme Layer Desiccation: The active enzyme layer dried out prematurely, reducing its catalytic efficiency.
2. Mediator Leaching / Oxidation: The chemical mediator that transfers electrons from the enzyme to the electrode wire degraded.
3. Falsely Low Current: Because the current output is directly proportional to the rate of chemical reaction, a compromised enzyme or mediator layer results in a significantly reduced current output, even in the presence of high physical glucose concentrations.

The Clinical Risk Cascade

When the current output drops due to sensor degradation rather than physiological changes, the CGM interpreter registers a false low glucose level (frequently reading at or below the default hypoglycemia threshold of 70 mg/dL).

This triggers a cascade of clinical hazards:

• Hypoglycemia Alert Fatigue: The patient is bombarded with false low alarms, leading them to disable alerts or ignore subsequent genuine alarms.
• Inappropriate Carb Loading: Patients consume unnecessary fast-acting carbohydrates to treat a non-existent low, causing severe, unmanaged rebound hyperglycemia.
• AID Over-Correction: In closed-loop automated insulin delivery (AID) configurations, a false low reading prompts the pump control algorithm to suspend basal insulin. This prolonged suspension leads to rapid ketogenesis and diabetic ketoacidosis (DKA), which is a life-threatening clinical sign.
• Missed Real Hyperglycemia: Because the sensor cannot generate high current, it fails to alarm for rising blood sugar levels (hyperglycemia), preventing the patient from administering timely correction boluses.

The FDA noted that as of January 7, 2026, these calibration and chemical failures resulted in 860 reported serious injuries and 7 patient deaths, making it one of the most clinically severe CGM events in recent history.

---

Recommended Reading

FDA Heart Valve Recalls: Edwards, Medtronic & Abbott MAUDE Teardown

Regulatory Post-Market Surveillance2026-07-13 · 18 min read

Deep-Dive: Dexcom Receiver Speaker Failures

Dexcom's Class I recalls of G6 and G7 receivers in 2025 highlights a physical and mechanical engineering vulnerability in display unit hardware.

Acoustic Engineering Failures

A CGM receiver is designed to serve as a reliable, dedicated display unit with loud, piercing alarms (commonly in the 50–60 dBA range) to wake patients during nocturnal hypoglycemic episodes. To produce that alert, the receiver relies on a small piezoelectric or micro-dynamic audio speaker that must be properly seated and electrically coupled to the circuit board.

According to the FDA's recall record (FDA Determined Cause: Nonconforming Material/Component), the G6 and G7 receiver defect was traced to defective foam or assembly errors that could cause the speaker to lose contact with the circuit board. When that contact is lost, the speaker produces no audio output, even though the receiver's vibration and visual prompts may continue to work — meaning the most urgent alert channel (the audible alarm) goes silent.

Normal Receiver: Speaker properly seated on circuit board --> Audible high/low glucose alarm
Recalled Receiver: Foam/assembly defect breaks speaker contact --> Silent alarm (vibration/visual only)

The Clinical Danger of Silent Alarms

Because patients with type 1 diabetes, particularly pediatric and elderly populations, often suffer from hypoglycemia unawareness, they depend entirely on nocturnal alarms to wake them during rapid glucose drops. A muffled or silent speaker prevents the receiver from communicating these life-saving alerts.

The FDA classified this as Class I because a silent alarm fails to notify the patient of a life-threatening physiological state (nocturnal hypoglycemia), which can rapidly progress to severe seizures, cognitive damage, coma, or death.

---

MAUDE Adverse Event Analysis for CGMs

To understand the baseline safety profile of CGMs, we analyzed the FDA’s MAUDE database. MAUDE houses voluntary, mandatory, and user facility reports of medical device malfunctions, injuries, and deaths.

For continuous glucose monitors, the FDA categorizes devices under two primary product codes in the Clinical Chemistry panel:

1. QBJ: Integrated Continuous Glucose Monitoring System, Factory Calibrated. This is a Class II device code that covers systems designed to send glucose data directly to integrated systems like insulin pumps or smart pens.
2. QLG: Integrated Continuous Glucose Monitoring System, Factory Calibrated, Not For Use With Automated Insulin Delivery Systems. This code covers factory-calibrated CGMs meant solely for standalone glucose monitoring.

Our query of the 2024 MAUDE device-event export (over 2.6 million device-event rows for the calendar year) yielded a total of 382,279 device-rows matching codes QBJ and QLG. The events are broken down as follows:

• Total CGM Events (2024): 382,279 rows
  • QBJ (AID-compatible): 347,165 rows (90.8% of all CGM reports)
  • QLG (Standalone): 35,114 rows (9.2% of all CGM reports)

Here is the distribution of the reported event types for the combined CGM codes:

• Malfunctions: 364,251 reports (95.28%)
• Injuries: 17,987 reports (4.71%)
• Deaths: 30 reports (0.01%)
• Other/No Answer: 11 reports (<0.01%)

Critical Context on MAUDE Data

When presenting these figures, it is essential to emphasize that MAUDE is a passive surveillance database. While these numbers are large, they do not prove direct causality. The presence of a "death" or "injury" report in MAUDE means that a patient using a CGM experienced a serious event or died, but it does not necessarily prove that the CGM malfunctioned or caused the death. Many reports are filed out of caution by manufacturers when they learn of a patient death, regardless of whether the device was operating within specifications.

However, the high volume of malfunction reports (over 364,000 in a single year) highlights the sheer frequency of user-reported issues with sensor adhesion, transmitter battery life, signal dropouts, and calibration drift. Since QBJ devices are integrated into automated insulin pump recalls and MAUDE teardowns to drive automated insulin delivery, sensor malfunctions carry a higher risk of triggering incorrect automated dosing decisions.

---

Manufacturer Analysis: Dexcom vs Abbott

A review of the openFDA recalls database reveals a total of 33 CGM recalls registered under codes QBJ and QLG. The division among manufacturers is heavily concentrated:

• Dexcom, Inc. / Dexcom Inc: 20 recalls (60.6%)
• Abbott Diabetes Care, Inc.: 8 recalls (24.2%)
• Cardinal Health Inc.: 2 recalls (6.1%)
• Medtronic Inc.: 1 recall (3.0%)
• Bigfoot Biomedical: 1 recall (3.0%)
• Other (Get Tested International): 1 recall (3.0%)

Recall Trends over Time

The initiation dates of these CGM recalls show a distinct spike in recent years:

• Pre-2023: Scattered (e.g., 1 in 2018, 1 in 2019, 1 in 2020, 2 in 2021)
• 2023: 3 recalls
• 2024: 1 recall
• 2025: 20 recalls (representing a massive escalation, driven largely by Dexcom's receiver speaker recall across multiple models and regions)
• 2026 (Year-to-Date): 4 recalls (including the Abbott Libre 3 sensor recall and Dexcom G7 watchOS app recall)

Regulatory Scrutiny & Warning Letters

The spike in recalls has been accompanied by direct regulatory warnings. On March 4, 2025, the FDA issued a Warning Letter (MARCS-CMS 700835) to Dexcom, Inc. following an inspection of their San Diego manufacturing facility. The letter cited violations of Quality System Regulations (QSR) / Current Good Manufacturing Practices (cGMP), including:

• Failure to adequately establish and maintain procedures for implementing corrective and preventive action (CAPA).
• Implementing changes to Dexcom G7 device hardware and software designs without conducting appropriate verification/validation or filing a new premarket notification (510(k)).

This warning letter underscores the regulatory risks of iterating on digital health software in a regulated environment. Under the FDA’s software modification guidance, what a software team may view as a minor update or "hotfix" can be classified by regulators as an unauthorized design modification, necessitating a formal recall and corrective action. For context on how this fits into Dexcom's wider footprint, see our guide on the Dexcom FDA registration footprint.

---

Recommended Reading

FDA OTC Hearing Aid Regulation: Product Codes & MAUDE Adverse Events

Regulatory 510(k)2026-07-13 · 18 min read

Failure Modes Breakdown: Sensors, Alarms, and Apps

Understanding the failure modes is essential for quality assurance teams to build robust risk mitigation strategies. The historical data points to three distinct archetypes of failure:

1. Sensor Chemistry and Calibration Drift

CGM sensors rely on glucose oxidase or glucose dehydrogenase enzymes embedded on a subcutaneous wire to generate an electrical current proportional to interstitial glucose levels.

• The Failure Mode: Abbott’s 2026 Class I recall of the FreeStyle Libre 3 and Libre 3 Plus highlighted a critical failure mode: incorrect low glucose readings. During manufacturing or due to package seal integrity issues, the enzyme layer can degrade or dry out, causing the sensor to output a falsely low current.
• The Risk: If a sensor reads falsely low, it may trigger unnecessary high-carbohydrate ingestion or prompt an integrated automated insulin pump to suspend insulin delivery, leading to severe hyperglycemia. Alternatively, if a patient calibrates their system using a faulty sensor or if the factory calibration fails, the device may fail to alert them to genuine hypoglycemia, leading to cognitive impairment or coma. To compare this to earlier product launches, see the Abbott Libre Duo dual glucose-ketone CE mark launch.

2. Audio and Alarm Hardware Failures

For patients with hypoglycemia unawareness, auditory alarms are a life-saving safety net.

• The Failure Mode: Dexcom’s 2025 Class I receiver recalls (affecting over 600,000 G7 units and 36,000 G6 units) highlighted a mechanical vulnerability: audio speaker failure. Moisture, dust, or physical degradation of the foam spacer around the micro-speaker caused the speaker to malfunction, either completely silencing the alarm or drastically reducing the volume.
• The Risk: Users could sleep through critical low glucose alarms, leading to nocturnal hypoglycemia, seizures, or death. The FDA designated this a Class I recall because a silent alarm fails to notify the patient of a life-threatening physiological state.

3. Software App Crashes and OS Incompatibilities

Modern CGMs are frequently "Bring Your Own Device" (BYOD) systems, relying on consumer smartphones and smartwatches (iOS, Android, watchOS) to display readings.

• The Failure Mode: The April 2026 Class II recall of the Dexcom G7 watchOS application was triggered by a bug that caused the watch app to crash or stop syncing data.
• The Risk: Because smartwatches are secondary displays (the primary display is the smartphone or receiver), the immediate risk is lower, resulting in a Class II designation. However, when users rely on smartwatch complications for quick-glance safety checks, software freezes can delay the detection of rapid glucose swings.

---

Action Plan for HTM Teams and Clinicians

When a CGM recall is issued, healthcare technology management (HTM) departments and clinical teams must act swiftly to mitigate patient risk:

Step 1: Physical Inventory Audit

HTM teams must conduct a thorough physical audit of all hospital departments (emergency rooms, intensive care units, pediatric clinics) to locate all patient-loaner or clinical demonstration CGM receivers.

• Cross-reference the receiver serial numbers against the manufacturer’s recall registry.
• Quarantine any receiver that falls within the recalled lot ranges.
• Contact the manufacturer to arrange for speaker module replacement or receiver trade-in.

Step 2: Clinical Alarm Verification

For receivers currently in use that are not subject to immediate replacement, clinicians and biomed technicians should perform a manual alarm test.

• Access the receiver's setting menu and trigger a test alert.
• Measure the sound pressure level at a distance of 1 meter using a calibrated sound level meter.
• If the output is below 50 dBA or sounds distorted/muffled, immediately retire the receiver from clinical service.

Step 3: Patient Software Audits & Education

Clinicians prescribing CGMs must establish a standardized patient review workflow:

• Verify that patients are utilizing mobile applications that are officially validated for their specific smartphone OS version.
• Counsel patients that mobile smartwatch applications (such as watchOS or Wear OS complications) are secondary displays only. Dosing decisions must always be confirmed on the primary smartphone app or receiver.
• Advise patients to check their sensor packaging for any physical damage or liquid leakage, which could indicate a compromised seal leading to sensor desiccation and false low readings.

Step 4: System Integration FMEA Updates

For R&D teams and manufacturers designing automated closed-loop systems (such as artificial pancreas setups linking CGMs to insulin pumps), the recent recalls necessitate updating the system-level Failure Mode and Effects Analysis (FMEA).

• Incorporate False Low Safeguards: Program the control algorithm to require manual user confirmation or a capillary fingerstick validation before suspending basal insulin for longer than 60 minutes if the rate of glucose change does not match typical physiological profiles.
• Implement Cross-Sensor Inconsistency Check: If the system is dual-sensing or integrates multiple data streams, use algorithmic checks to flag inconsistencies between sensor current and auxiliary sensors.

---

Future Outlook of CGM Regulatory Standards (2027 and Beyond)

As continuous glucose monitoring technology moves toward wider adoption and tighter integration with smart delivery systems, regulatory agencies are actively revising their expectations:

1. Pre-Market Human Factors Testing Requirements: The FDA’s digital health division is planning to mandate more rigorous human factors validation for companion software. Manufacturers will need to demonstrate that users do not suffer from alert fatigue and can successfully identify and resolve sensor calibration drift without clinical intervention.
2. Cybersecurity Gates for AID Ecosystems: With the implementation of the PATCH Act and associated pre-market cybersecurity requirements, any CGM product code QBJ (integrated systems) must provide a full Software Bill of Materials (SBOM) and demonstrate robust encryption protocols for wireless Bluetooth communication with pumps.
3. Transition to Non-Invasive Technologies: While factory-calibrated enzymes represent the current standard, several startups are seeking clearances for optical and transdermal non-invasive sensors. The FDA is expected to establish new product codes for these systems, with distinct validation protocols for calibration accuracy.

For further reading on how the FDA manages post-market medical device surveillance across different sectors, consult our detailed analysis on MAUDE adverse-event datasets and trends.

---

Recommended Reading

Managing End-of-Service (EOS) Medical Equipment: Lifecycle Risks & HTM Fleet Controls

Quality Systems Post-Market Surveillance2026-07-12 · 16 min read

Frequently Asked Questions (FAQs)

Is the Dexcom G7 recall a Class I recall?

Yes, the 2025 recall of the physical Dexcom G7 Receiver (affecting 602,445 units) due to speaker/alarm failure was classified by the FDA as a Class I recall. However, the subsequent April 2026 recall of the Dexcom G7 watchOS companion app was classified as a Class II recall.

How many serious injuries and deaths are linked to the Abbott FreeStyle Libre 3 recall?

According to official FDA recall notices as of January 7, 2026, the Abbott FreeStyle Libre 3 and Libre 3 Plus sensor recall (Class I) was associated with 860 reported serious injuries and 7 reported deaths.

What product code covers continuous glucose monitors in the FDA database?

The two main FDA product codes for factory-calibrated continuous glucose monitors are QBJ (Integrated CGM, Factory Calibrated) and QLG (Integrated CGM, Factory Calibrated, Not For Use With Automated Insulin Delivery Systems).

What is the difference between a Class I and Class II FDA recall?

A Class I recall is the most urgent, indicating that the use of the device carries a reasonable probability of causing serious health consequences or death. A Class II recall indicates a lower risk, where the device may cause temporary or medically reversible health problems, or where the probability of serious adverse consequences is remote.

How does a user report a CGM failure to the FDA?

Patients and clinicians can report device issues directly to the FDA through the MedWatch program online or by submitting FDA Form 3500. These reports are subsequently indexed in the public MAUDE database.

What should a patient do if their CGM alarms for a low glucose reading but they feel normal?

The patient should immediately perform a capillary blood glucose fingerstick check using a standard handheld blood glucose meter. Dosing or treatment decisions should be based on the fingerstick reading if it contradicts the CGM.