Complications associated with a flash glucose monitoring system in diabetic dogs

Abstract

Interstitial glucose monitoring systems are commonly being used in diabetic dogs. The aim of this study was to document the incidence of complications associated with the use of a flash glucose monitoring system (FGMS) in dogs. Medical records of dogs that had placement of a 14-day FGMS during a 1-year period were reviewed. Data retrieved included the number of days the sensor remained attached and functional, sensor detachment, sensor failure prior to the end of the 14-day monitoring period, and dermatologic changes at the sensor site. Descriptive statistics were used to characterize the data. Thirty-four dogs had FGMSs placed. Most [32/34 (94%)] sensors were placed over the dorsolateral aspect of the thorax caudal to the scapula. Twenty-four sensors (71%) remained attached for the full 14 days. Incidence of complications associated with FGMS use was 13/34 (38%). The most frequent complication was mild dermatologic changes at the sensor site [6/34 (18%)]. Erythema and crusting at the attachment site were common and could be related to contact dermatitis, hypersensitivity, or skin preparation prior to placement. Flash glucose monitoring systems are safe in dogs, although there are some potential complications that should be discussed with dog owners.

Introduction

The mainstay of treatment in the management of canine diabetes mellitus is insulin therapy. The primary goals of therapy are palliation of clinical signs and avoidance of life-threatening complications such as diabetic ketoacidosis and hypoglycemia. In order to achieve these goals, individualized treatment plans and frequent reassessments are necessary. Most diabetic dogs require serial blood glucose monitoring (i.e., blood glucose curves) to adjust their insulin dose to establish glycemic control.

There are various glucose monitoring methods in the diabetic dog. Direct glucose monitoring methods provide the information needed for safe and effective dose adjustments (1). Direct monitoring methods include in-hospital or at-home blood glucose curves, spot blood glucose measurements, and continuous glucose monitoring. Indirect monitoring methods include urine glucose monitoring, measurement of fructosamine concentrations, glycosylated hemoglobin concentrations, and assessment of clinical signs (2). Each of these techniques has limitations.

Blood glucose curves vary from day to day and can be affected by variation in the patient’s normal routine (3). Not all owners are willing and able to obtain at-home blood glucose curves. In-hospital curves are less convenient and prove challenging to interpret, especially for feline patients whose curves are confounded by stress hyperglycemia. Spot blood glucose measurements used as a sole method may not provide enough information to allow safe dosing changes. Urine glucose measurements via dipsticks reflect glucose concentrations over the time urine was formed and are often positive. Negative readings may be due to hypoglycemia, normoglycemia, or mild hyperglycemia. Fructosamine measurement reflects glycemic control over the previous 2 wk, but this does not provide information about glucose nadirs, duration of action of insulin, or change in response to insulin administration (4). Canine glycosylated hemoglobin concentration (HbA1c) may correlate well with glycemic control but is rarely used in veterinary medicine. In clinically stable dogs with chronic diabetes, HbA1c has been reported to be more consistent with clinical evaluation of diabetic control compared to serum fructosamine concentrations (5,6). Moderate to severe anemia in humans has been reported to increase the level of HbA1c (7). This has not been well-documented in canines and should be taken into consideration when assessing HbA1c levels on diabetic dogs with moderate to severe anemia.

Continuous glucose monitoring systems (CGMSs) utilize a small filament inserted in the subcutaneous tissue and measure glucose concentrations in the interstitial fluid. An electrical signal is produced by the reaction between the glucose oxidase enzyme within the filament and the interstitial glucose (8). This electrical signal is detected by the sensor. Interstitial glucose concentrations are similar to that of blood glucose concentrations; however, there is a delay in interstitial glucose with rapid changes in blood glucose such as postprandially (9). Glucose values are stored in the recording device. The first-generation systems were only capable of retrospective analysis of the glucose concentrations following removal of the sensor and uploading of the data. Real-time CGMSs automatically display each glucose reading and can integrate with insulin pump systems in order to adjust insulin dosing; however, these devices are expensive and most need blood glucose measurement once to twice daily for calibration. Continuous glucose monitoring systems are regularly used in human diabetics and their potential use in veterinary medicine has been previously described (9–14).

A relatively new flash glucose monitoring system (FGMS) (FreeStyle Libre 14-day system, Abbott Diabetes Care, Alameda, Califoania 94502, USA) that eliminates some of the pitfalls of the older CGMS has been designed for human diabetic patients. The FGMS measures the interstitial glucose every minute via a disposable round sensor with a small catheter inserted under the skin that can be worn for up to 14 d. The FGMS is factory calibrated, therefore it does not require peripheral blood sampling for calibration. The system begins working 1 h after application. The sensor is “flashed” by the reader to display the interstitial glucose concentration. As long as the sensor is “flashed” every 8 h, continuous data are collected, stored, and can be downloaded to generate comprehensive glucose reports. The FGMS was studied in stable diabetic dogs with good correlation reported between interstitial glucose and peripheral blood glucose concentrations. Accuracy was 99% in the euglycemic and hyperglycemic ranges, and 93% in the hypoglycemic range (15). Another study showed good accuracy of the FGMS in canines with diabetic ketoacidosis (16).

The FGMS is considered to be generally safe in both human and veterinary patients, but very few studies have specifically examined the safety of these devices. Mild skin reactions were noted in 2 studies evaluating the clinical usefulness of the FGMS in diabetic dogs. The first study documented transient mild to moderate skin irritation in 80% of the dogs. In addition, one dog developed Horner’s syndrome shortly after the owner had applied the sensor on the ventral neck (17). The second study documented pruritus in 2/4 cases (18). Another study evaluating the effect of skin thickness in the accuracy of the FGMS documented early detachment in 2/14 dogs and sensor error in 7/14 dogs (19). The safety of the FGMS has been evaluated in cats in which the overall incidence of complications was 10/33 (30%). The most frequent complication observed was early sensor detachment, followed by mild dermatologic changes (erythema, crusts) (20). Studies on complication rates and safety of the devices in dogs do not exist.

The aim of this study was to identify the prevalence and nature of complications associated with the use of the FGMS in dogs. Our hypothesis was that the FGMS would be safe and well-tolerated, without any major adverse effects.

Materials and methods

Medical records of dogs that had the FGMS 14-day system placed between December 2020 and December 2021 at the Foster Small Animal Hospital at the Cummings School of Veterinary Medicine at Tufts University were examined. The following terms were used to search the electronic medical record: “Freestyle,” “Libre,” “Omni-freestyle,” and “blood glucose monitoring system.”

All sensors were placed by trained veterinary technicians or veterinarians. The desired location for placement was shaved and disinfected with alcohol wipes provided by the manufacturer. The site was allowed to air dry for ~1 min. An adhesive wipe (SkinTac Wipes; Torbot, Cranston, Rhode Island 02910, USA) was used on the cleaned site. The FGMS was placed using the applicator provided by the manufacturer and following the manufacturer’s instructions. For some dogs a small amount of cyanoacrylate tissue glue (3M Vetbond; St. Paul, Minnesota 55144, USA) was placed on the contact surface of the sensor prior to placement. The device was scanned immediately after placement to initiate the 1-hour warm-up period that is required before data collection.

Cases were included if the FGMS was placed between December 2020 and December 2021 for the purpose of interstitial glucose monitoring, if the date of sensor placement was recorded and if the sensor remained attached for at least 24 h. For dogs that had multiple sensors placed, only the first sensor placement was included. Cases were excluded if the date of placement was not recorded, if the placement data was incomplete, or if the sensor fell off within 24 h of placement.

Medical records were reviewed for patient and FGMS placement data, as well as complications documented in hospital, at home, or during follow-up visits. Patient data included signalment, body condition score (BCS: 1 to 9 scale), primary disease processes, and comorbidities. Flash glucose monitoring system placement data included site of sensor placement, technique for placement (i.e., if tissue glue was used), duration of time that the sensor remained attached and functional, and complications. The types of complications are defined in Table I. For patients with complications, the medical records were reviewed further to determine if additional treatment was necessary.

Table I.

Types of complications associated with flash glucose monitoring system (FGMS) use.

Complication	Definition
None n = 21 (62%)	Sensor remains attached and functional for 14 d, or is removed early for owner convenience. Patient is tolerant of the sensor and no adverse effects are observed.
Early detachment n = 4 (12%)	Sensor inadvertently detaches before the 14-day monitoring is complete.
Mild dermatologic changes n = 5 (15%)	Erythema, mild crusting, abrasions, mild pruritus/discomfort noted upon removal/detachment of the FGMS. Additional intervention not required.
Major dermatologic changes n = 1 (3%)	Changes include marked crusting, erosions, ulceration, scabbing, abscessation and severe pruritus noted upon removal of/detachment of the FGMS. Additional intervention required.
Dysfunctional sensor n = 4 (12%)	Sensor remains attached but error message is given suggesting sensor is no longer operational.

Descriptive statistical analysis was performed for applicable variables (age, weight, BCS). Data were analyzed for normality using Microsoft Excel formulae for kurtosis and skewness, and were expressed as median and range (non-parametric) or mean ± SD (parametric).

Results

Thirty-four dogs met the criteria for enrollment in the study from December 2021 and December 2022; 11 were spayed females and 23 were castrated males. Mixed breed dog represented the most common breed [n = 10 (29%)], followed by Labrador retriever [n = 4 (12%)], dachshund [n = 2 (6%)], Yorkshire terrier [n = 2 (6%)], West Highland white terrier [n = 2 (6%)], and one each of miniature Australian shepherd, Maltese, bichon frise, Boston terrier, Jack Russell terrier, silky terrier, petite basset griffon Vendeen, shih tzu, schnauzer, Samoyed, Chihuahua, Shetland sheepdog, basset hound, and miniature pinscher. The mean age of the included dogs was 10.2 y ± 2.8 y. Mean weight was 17.5 kg ± 12.4 kg. Mean BCS was 6/9 ± 1.6.

Twenty-three FGMS were placed for routine diabetes monitoring. Ten were placed in dogs hospitalized for management of diabetic ketoacidosis. One FGMS was placed for monitoring of a dog with an insulinoma. Comorbidities were present in 30 dogs and included hepatopathy [n = 16 (47%)], systemic hypertension [n = 6 (18%)], cardiac disease [n = 6 (18%)], hyperadrenocorticism [n = 5 (15%)], kidney disease [n = 3 (9%)], seizures [n = 3 (9%)], otitis externa [n = 3 (9%)], pancreatitis [n = 2 (6%)], dyslipidemias [n = 2 (6%)], hypoadrenocorticism [n = 2 (6%)], hypothyroidism [n = 1 (3%)], and immune-mediated neutropenia [n = 1 (3%)]. Several dogs had more than one comorbidity [n = 21 (62%)].

Locations for FGMS attachment included overlying the dorsolateral aspect of the thorax just caudal to the scapula (n = 32 (94%)] (Figure 1), between the shoulders on the dorsum [n = 1 (3%)], and on the lateral aspect of the hip [n = 1 (3%)]. Twenty-one of 34 (62%) FGMSs remained attached and functional for the full 14 d, with two of those remaining attached longer (15 and 22 d, respectively). The sensors that remained attached longer only provided 14 d[ of data. Two FGMSs (6%) were removed or replaced early (despite ongoing functionality) because it was convenient for the owners. Cyanoacrylate tissue glue was used in 4/34 FGMS placements due to a history of early sensor detachment. Cyanoacrylate tissue glue was avoided for 30/34 FGMS placements. Sensor removal was performed using an adhesive remover (TacAway Wipes; Torbot, Cranston, Rhode Island 02910, USA).

Figure 1.

Placement of the flash glucose monitoring system over the right dorsolateral aspect of the thorax of a domestic dog.

Eleven of 32 (34%) FGMSs applied over the dorsolateral aspect of the thorax, 1/1 (100%) FGMS applied on the dorsum, and 1/1 (100%) FGMS applied on the lateral aspect of the hip had associated complications. The most common complication recorded was mild skin erythema, superficial abrasions and/or crusts that were noted when sensors were removed [n = 5 (15%)] (Figure 2). Marked crusting was noted at the FGMS site following removal in one dog [n = 1 (3%)] (Figure 3). Two of 4 patients in which sensors were placed with cyanoacrylate tissue glue had dermatologic complications. Four of 34 (12%) of the FGMS detached before 14 d of data had been collected. Four of 34 FGMS (12%) remained attached appropriately but stopped functioning (i.e., dysfunctional sensor) earlier than expected due to sensor/reader error. One of these sensors became dysfunctional after being hit during an abdominal ultrasound. Most sensors [n = 21/34 (62%)] were placed and used without reported or observed complications. One patient had rapid hair growth beneath the sensor site, but the sensor remained attached and functional for the full 14 d. The incidence of complications associated with FGMS use in dogs are summarized in Table I.

Figure 2.

Mild erythema and abrasion noted after removal of flash glucose monitoring system.

Figure 3.

Marked crusting noted after early detachment of the flash glucose monitoring system.

Discussion

Our study provides an assessment of complications associated with the use of FGMS in dogs. This device is being used much more readily by both general practitioners and specialists. The device provides veterinarians with consecutive days of monitoring in the home environment as opposed to in-hospital glucose curves. The accuracy of the FGMS has been established in diabetic dogs (15,16). Complication rates have been briefly mentioned in larger studies, but the present study specifically evaluated dermatologic and technical complications associated with the FGMS.

The most common complication was mild dermatologic changes (erythema, mild crusting, abrasions). Erythema and mild crusting at the FGMS placement site were relatively common. This could be related to contact between the skin and the device or secondary to skin preparation prior to placement. Owners reported that the erythema improved after 1 to 2 d following sensor removal. Marked crusting was noted at the FGMS site in one dog following early detachment due to severe pruritus during the attachment time. The crusting required a medicated shampoo for resolution. Due to the dog’s response following placement of the first sensor, the owner declined additional sensors. Contact dermatitis is a significant ongoing complication associated with glucose sensors in humans. This is thought to be related to an allergy to plastic and adhesive materials in the sensors (21). The adhesive chemicals (isobornyl acrylate and N,N-dimethylacrylide) within the FGMS evaluated in this study have been reported as a contact allergen in humans (22,23). Studies in humans have shown as high as 46% of glucose sensor users have skin problems related to their sensors (21,24). Contact dermatitis should be considered as a possibility in our dogs, namely those with development of erythema, crusts, and pruritus. Further studies are necessary to examine the possibility of a similar contact dermatitis associated with glucose sensors in dogs.

More serious dermatologic changes such as marked crusting were observed, but there was no documentation of erosions or abscess formation. Although, one dog that had cyanoacrylate tissue glue applied to the skin-facing surface of the FGMS had a very difficult removal process with mild dermatologic complications at the placement site following removal on day 14. Cyanoacrylate adhesives are safe and effective when used as a sole agent for superficial skin closure in humans, dogs, and cats (25–27). The effect of cyanoacrylate use in conjunction with the chemical adhesives contained in the FGMS is unknown, although there is potential that it could result in a more severe skin reaction. Although larger studies are needed to determine if additional adhesives, sensor location, and timing of sensor removal have an effect on the incidence or type of complications, we are currently avoiding the use of cyanoacrylate tissue glue (if possible). In addition, sensor placement over the dorsolateral thorax is preferred with alternating of sides if repeat sensor placement is indicated.

The complications reported with the next highest prevalence after mild dermatologic changes were early detachment of the sensor and a dysfunctional sensor. In addition, two dogs were excluded from the study due to sensor detachment < 24 h of placement. Potential causes for early detachment include patient factors (self-removal secondary to scratching, rubbing on furniture/carpet, strenuous activity, housemate interactions) and ineffective placement (inadequate shaving or cleaning, patient movement during placement). Although early detachment and a dysfunctional sensor are not typically detrimental to patient care, owners should be made aware of the potential complication in order to appropriately establish expectations. This could also be a discussion with clients before performing ultrasound or radiographs in patients with a sensor in place to warn of potential complications.

There were numerous limitations to the present study. The major limitation was the retrospective nature and small sample size. Data for the study were collected over a year-long period resulting in altering personnel. All technicians performing sensor placement were trained to do so, but there is variability in technique or tendency to use cyanoacrylate tissue glue during placement. There was no standardized protocol for recording complications at time of sensor detachment. Due to early detachment or at-home removal of the FGMS without timely images obtained, minor complications might not have been deemed problematic enough to report.

In conclusion, the examined FGMS appears to be relatively safe and well-tolerated. Most of the devices remained functional for a full 14 d and were not associated with complications. Early sensor detachment and mild dermatologic changes (erythema, crusts) are relatively common complications that should be discussed with dog owners prior to placement.