Adverse Events, Skin Reactions, and Safety Warnings

Adverse Events, Skin Reactions, and Safety Warnings in Continuous [Glucose Monitoring](/)

Introduction

Most of a CGM's safety problems come from where it lives. A thin sensor filament sits in the fluid under your skin for days at a time, held in place by medical adhesive, and it measures glucose in that fluid rather than in blood. We sorted the problems that show up in FDA labeling and the medical literature into four groups: skin reactions, limits set by the body's own physiology, medicines that throw off the sensor's chemistry, and hardware failures.

Dermatological Issues

Skin trouble is the problem CGM users feel most directly. A 2020 systematic review of 54 studies counted about one skin complication per eight weeks of sensor wear, mostly redness, itching and hardened skin at the site [3]. Only 1.5% of those events were judged severe. Even so, skin irritation was the most common reason people stopped using a CGM [3].

Allergic contact dermatitis is a different, immune-driven reaction, and it has a named culprit. In 2017, dermatologists in Belgium and Sweden reported 15 people who reacted to FreeStyle Libre sensors; 12 of the 13 patch-tested for isobornyl acrylate (IBOA), an acrylate chemical, were allergic to it, and lab analysis found IBOA in the sensors themselves [2]. Skin problems also show up in FDA's adverse event reports. Skin reactions and bleeding made up 16.1% of 2022 injury reports for the Dexcom G6 [16], a sensor Dexcom stopped manufacturing after July 1, 2026 [11]. We cover the adhesive chemistry in more depth in Adhesive Chemistry and Dermatological Issues in CGMs. Which barrier films or skin treatments make sense for you is a decision for your care team.

Physiological Limitations

A CGM reads glucose in interstitial fluid, the fluid that surrounds cells under the skin, and glucose takes a few minutes to get there from the blood. When Mayo Clinic researchers traced labeled glucose in eight healthy adults after an overnight fast, it reached the interstitial fluid 5 to 6 minutes after entering the blood [4]. In six adults with type 1 diabetes, the median delay was 6.8 minutes, and every participant's was under 10 minutes [5]. That gap matters most when glucose is moving fast, because the sensor is reporting where glucose stood a few minutes earlier.

Pressure is the other physiological trap. In a 2013 study that video-recorded sleeping volunteers, aberrant sensor readings lined up closely with people lying on their sensors, and most of them were sudden drops, likely from compressed tissue and reduced local blood flow [6]. Engineers call these pressure-induced sensor attenuations, or PISA. The reading falls while blood glucose doesn't. In a system that pauses insulin when it predicts a low, a false low like that can shut off the pump unnecessarily and lead to inappropriate insulin dosing decisions [7]. Dexcom's and Abbott's labeling both tell users to rely on a blood glucose meter when sensor readings don't match their symptoms [9][13].

Chemical Interferences

Certain medicines and supplements can push CGM readings off, and which ones matter depends on the sensor. We checked each one against current FDA labeling.

Hydroxyurea affects Dexcom sensors. FDA's summaries for the Dexcom G7 and G7 15 Day say readings will be falsely high in people taking it, and that those users shouldn't use the sensor for treatment decisions [8][9]. A Cincinnati Children's Hospital study saw the effect on real patients: after hydroxyurea doses, readings from the older Dexcom G4 Professional system and from the G6 ran far above meter values for roughly half an hour to 6 hours [10]. Clinicians at Northwestern reported the same interference in a 2021 letter in Diabetes Care [1].

Acetaminophen is also a Dexcom labeling issue. The G7 and G7 15 Day labeling says a standard dose is fine, but readings will be falsely high after more than a standard dose [8][9].

Vitamin C (ascorbic acid) is an Abbott labeling issue. FDA's 2023 review of a redesigned FreeStyle Libre 2 and Libre 3 sensor tail says ascorbic acid "significantly" interfered with the earlier sensor; the redesign roughly doubled the daily vitamin C intake the labeling allows, and the labeling still warns that larger amounts can falsely raise readings and cause a user to miss a severe low [12]. FDA's 2024 review of the current Libre 2 Plus and Libre 3 Plus sensors relies on that same interference testing [13]. The older Libre 2 and Libre 3 sensors were sold only through September 30, 2025 [14].

The implantable Eversense 365 uses a different chemistry and has its own list: tetracycline-class antibiotics can falsely lower its readings, and mannitol or sorbitol given intravenously or in dialysis or irrigation solutions can falsely raise them [18].

A falsely high reading is dangerous because it can hide a real low. A 2023 case report described severe hypoglycemia caused by hydroxyurea interference with a sensor that was feeding an automated insulin delivery system [15]. Whether to change a medicine or monitor another way is a decision for your care team. Our CGM Interference Profiles Comparison lines up the interferents side by side.

Hardware Risks

The sensor filament can break off or detach. Dexcom's G7 labeling tells users not to ignore broken or detached sensor wires and to contact Dexcom technical support [8][9]. In FDA's 2022 adverse event reports, a detached or missing sensor wire accounted for 3.1% of Dexcom G6 injury reports [16]. Retained pieces can need surgery. A 2025 case report describes a 2-year-old using a Dexcom G6 whose swollen thigh nodule turned out to be retained wire fragments; surgeons removed two pieces and left two deeper ones in place because removing them risked injury or bleeding [17]. Applicator failures, where the inserter doesn't place the sensor correctly, are another hardware complaint, though we didn't find a primary source that measures how often they happen.

MRI rules differ by device. Dexcom says not to wear any G7 component during an MRI or diathermy (high-frequency electrical heat treatment); a CT scan is acceptable if the sensor is kept out of the scanned area and covered with a lead apron [8][9]. For the Libre 2 Plus and Libre 3 Plus, FDA's 2024 review removed the MRI and CT contraindications, while diathermy still requires removing the system [13]. The Eversense 365 smart transmitter is labeled MR Unsafe and must come off before an MRI [18]. Planning around a scan is something to work out with your care team and the imaging staff.

Conclusion

Most of these risks are manageable once you know which ones apply to your sensor, and the same substance can matter for one brand and not another. As of October 2026, FDA labeling ties hydroxyurea and larger acetaminophen doses to Dexcom G7 sensors, high-dose vitamin C to FreeStyle Libre sensors, and tetracyclines to Eversense 365 [8][12][18]. The gap we kept running into was measurement: skin reactions and drug interference are well studied, but we found almost no published data on how often applicators fail.