• Open Access

Continuous Glucose Monitoring in Dogs and Cats


  • Work Performed at the College of Veterinary Medicine Teaching Hospital, University of Missouri—Columbia, Columbia, MO.

Corresponding author: Charles E. Wiedmeyer, DVM, PhD, DACVP, 900 E. Campus Dr., A345 Clydesdale Hall, Columbia, MO 65211; e-mail: wiedmeyerc@missouri.edu.


Use of continuous glucose monitoring in veterinary medicine is gaining popularity. Through use of a commercially available continuous glucose monitor system, insights into daily glucose changes in dogs and cats are achievable. The continuous glucose monitoring system measures glucose concentrations in the interstitial fluid of the subcutaneous space by use of a small, flexible probe. When placed in the subcutaneous tissue, the probe is connected to a recording device that is attached to the animal and records the interstitial fluid glucose concentration every 5 minutes (288 readings per 24 hours). Once attached and properly calibrated, the instrument can remain in place for several days, hospitalization of the patient is not necessary, and the normal daily routine of the animal can be maintained. The data from the recording device are then downloaded and a very detailed picture of the interstitial fluid glucose concentration over that time period can be obtained. Subcutaneous interstitial fluid glucose concentrations have a good correlation to blood glucose concentrations within a defined range. The continuous glucose monitoring system has distinct advantages over traditional blood glucose curves and is a valuable tool for managing diabetic dogs and cats. In addition, other clinical uses for continuous glucose monitoring are being developed. This review is designed to outline the technology behind the continuous glucose monitoring system, describe the clinical use of the instrument, provide clinical examples in which it may be useful, and discuss future directions for continuous glucose monitoring in dogs and cats.

A continuous glucose monitoring system (CGMS)a has recently been introduced in veterinary medicine as a method for monitoring glucose concentrations in dogs and cats.1,2 This system was originally designed for human diabetic patients as an at-home method of glucose concentration monitoring that circumvented the need for repeated capillary blood sampling. Sensors that continuously measure glucose have been so as an alteDepartment of Veterinary Pathob z ood glucose measurement in order to achieve better overall blood glucose control. The technology of the CGMS is designed to measure subcutaneous interstitial fluid (ISF) rather than blood glucose concentrations.3 The subcutaneous ISF is ideal for glucose concentration measurement because it is easily and safely accessed, has rapid equilibration with the blood, and has a good correlation with blood glucose concentration. The CGMS measures ISF glucose tions with a small, flexible sensor inserted through the skin into the subcutaneous space, secured to the skin, and attached to a recording device. The ISF glucose concentration is recorded and stored every 5 minutes (288 readings per 24 hours). After the CGMS is removed, the data are downloaded to a computer for analysis. The CGMS avoids several limitations of traditional blood glucose curves including intermittent assessment of blood glucose concentration, hospitalization, patient restraint, and repeated phlebotomy. Because the CGMS provides 288 data points per 24 hours, it provides a detailed insight into daily changes in ISF glucose concentration. Once the CGMS is placed, 3 blood glucose samples are required for calibration each calendar day of monitoring, hospitalization is not necessary, and the patient can remain in his or her home environment. These advantages can minimize confusion caused by stress hyperglycemia often observed in cats. In addition, because the animal can be kept in its home environment and the normal daily routine can be maintained, a more natural state of ISF glucose can be assessed.

Evaluation of dogs and cats undergoing insulin therapy for diabetes mellitus is the most common use of the CGMS.1,4 Because the CGMS is capable of providing a detailed glucose profile, a more precise assessment of glycemic control can be made. Since being validated as a method for monitoring ISF glucose in dogs and cats, the CGMS has gained popularity for use by veterinarians. The purpose of this review is to provide information about the CGMS technology, describe the clinical use of the CGMS, offer examples of its use in the clinical setting, introduce the latest advances in CGMS technology, and outline the potential uses of CGMS. This information will allow the veterinarian to make informed decisions about the use of CGMS.

CGMS Technology

A complete CGMS systema consists of a single-use sensor, a recording monitor, a docking station, a computer, and the requisite software (Fig 1). The sensor is an amperometric device designed to measure glucose in the subcutaneous ISF. The sensor consists of an electroenzymatic 3-electrode cell by which a constant potential is maintained between a working electrode and a reference electrode.3 The glucose concentration is detected by the sensor based on the generation of hydrogen peroxide from the reaction of glucose and oxygen with the enzyme glucose oxidase (Fig 2).3,5–7 Oxidation of hydrogen peroxide results in the generation of an electrical signal that is recorded by a recording device.3,7 The sensor is housed in flexible tubing with a side window enclosed by a polyurethane membrane that allows the active electrode to interact with the interstitial fluid.3 This polyurethane membrane is glucose diffusion limited, and therefore a linear relationship between the ISF glucose concentration and the sensor current is maintained. The sensor current is detected by the recording device and translated into a glucose concentration. The tubing containing the sensor is contained within a 22-G needle for introduction through the skin into the subcutaneous tissue. Once the flexible sensor is in place and the 22-G needle is removed, the sensor is secured to the skin and attached to a small cable leading to the monitoring device.

Figure 1.

 Continuous glucose monitoring system (recording device and docking station for downloading data). Inset: interstitial fluid glucose probe.

Figure 2.

 Representation of flow of glucose from capillary through the interstitial space to cells. Inset: Chemical reaction of glucose with glucose oxidase to form an electrical signal that occurs in the continuous glucose monitoring system sensor.

The dynamic relationships between glucose concentrations in the plasma and ISF have been extensively studied.3,8,9 Most minimally or noninvasive methods for detecting glucose focus on estimating its concentration in the ISF rather than in blood.8 Despite the number of technologies used to measure glucose in the ISF, concerns arise regarding potential differences between blood glucose and ISF glucose concentrations. Direct measures of ISF glucose cannot be readily obtained, thus creating potential problems with proper calibration of devices that measure ISF glucose. To address this issue, the dynamic relationship between blood glucose and ISF glucose concentration must be understood. Fundamentally, blood glucose and ISF glucose are separated by the capillary wall. Changes in ISF glucose are related to the rate of diffusion of glucose across the capillary wall and rate of glucose clearance from the ISF (Fig 2). Glucose enters the ISF by diffusion across the capillary wall and is then irreversibly cleared from the tissue bed in a concentration-dependent manner. Because the rate of glucose uptake in the tissues immediately surrounding the probe is negligible, a steady state between blood glucose and ISF glucose is achieved. The dynamic relationship between these compartments has been described mathematically and by experimental models.8–11

Theoretically, the steady state between blood and ISF glucose could be altered by rapid changes in blood glucose concentrations, by the influence of insulin, or by both.3,8,9 These factors may cause a time delay in the equilibrium between the ISF and vascular compartments. Studies performed in dogs have examined sensor responses to stepwise increases in blood glucose concentration with and without changes in endogenous insulin.3,8 In these models, rapid changes in blood glucose concentrations resulted in corresponding rapid increases in the electrical signal produced by the sensor without interference from exogenous insulin.3,8 However, changes in ISF glucose concentrations lagged behind changes in blood glucose concentrations by 5–12 minutes.3,8 This time interval may be due to delays in glucose diffusion across the capillary membrane or the sensor polyurethane membrane. Vacillations in capillary wall or sensor membrane diffusion can introduce errors in the calculated sensor glucose results.3 To avoid this pitfall, a digital filter designed to correct for consistent delays in diffusion is built into the recording device.

Clinical Use of the CGMS

Placement of the CGMS in a dog or a cat has been described previously.1,2,4,12,13 In short, a small patch of hair (approximately 3 cm × 3 cm) is shaved on the lateral thoracic region. The skin is cleaned of excess hair and wiped with an alcohol swab. The sensor and attached stylet are inserted through the skin and the stylet is removed. The sensor is fastened to the skin by use of a cyanoacrylate adhesive. Once the sensor is securely placed, the cord from the recording device is connected to the sensor and attached to the skin with adhesive. By means of an elastic bandage and tape harness or a monitoring vest, the recording device is secured to the animal's dorsum (Fig 3). Any bandage material covering the unit display screen is cut away and cloth tape is placed around the formed edges (Fig 3). This procedure provides a window in which to visualize the recording device display. After the system is attached, a 1-hour initialization period is required. The monitor is then calibrated by entering 3 separate blood glucose concentrations into the monitor for each calendar day the instrument is attached. It is imperative that a method for obtaining a rapid blood glucose concentration be used for calibration. Because the unit records an ISF result every 5 minutes, delay of data entry into the recording device may result in improper calibration of the CGMS. A reliable hand-held blood glucose monitor is sufficient for this procedure. Comparisons between hand-held glucose monitors and more standardized methods have been described.14 Also, it is essential that the 1st blood glucose calibration point be entered as soon as possible after the 1-hour initialization period. Delay with this procedure may also result in improper ISF glucose readings. For proper calibration of the CGMS, at least 3 blood glucose concentrations between 40 and 400 mg/dL must be entered into the monitor each calendar day. The manufacturer has set the working range of ISF glucose for the CGMS between 40 and 400 mg/dL. Because the instrument is incapable of recording a result beyond this range, blood glucose concentrations used for calibration must fall within this range.

Figure 3.

 Attachment of the continuous glucose monitoring system (CGMS) to a dog. Inset: Bandage cut away to view CGMS monitor window.

As stated, for proper calibration of the CGMS, at least 3 blood samples must be analyzed for glucose concentration over a 24-hour period. Most routine assays and measuring devices use a 1-point calibration for calculation of an accurate value. With a 1-point calibration, a calibration constant is determined for the calibration point and subsequent reported results are calculated by use of that constant until the next calibration is determined. The ISF glucose concentration is determined by an enzymatic reaction that converts glucose into an electrical signal. The electrical signal is applied to an algorithm that is used to convert the signal (nanoamperes) into a glucose concentration. The sensor's electrical signal is calibrated to blood glucose concentrations by use of linear regression with a fixed intercept. This calibration process uses all of the blood glucose readings within the dynamic range entered into the instrument with a 24-hour moving window to generate an accurate calibration factor, as opposed to a single-point calibration. Independent glucose meter readings not used for calibration have been evaluated to validate this calibration approach.8,15 Blood glucose and ISF glucose concentrations have shown good correlation in multiple species within the range of the instrument.3,8,16

After attachment, initialization, and calibration of the instrument, dogs and cats can be returned to their home environment. Once home, the owners are instructed to follow the animal's normal routine of insulin administration, feeding, and exercise. This practice helps to decrease episodes of stress-induced hyperglycemia and maintains the animal's normal day-to-day routine. If multiple days of monitoring are required, the animal can be returned to the clinic for a brief period (≥90 minutes) each day for the 3 blood glucose samples required for calibration. It is not necessary to evenly space the timing of the blood samplings used for calibration.a Alternatively, the owners may calibrate the instrument at home. Techniques have been described for sampling of blood from the ear of dogs and cats and use of a portable glucose meter.17,18 This technique may reduce stress-induced hyperglycemia associated with travel to the clinic or excessive handling of the patient. Events such as feeding, insulin administration, and trips to the hospital can be recorded into the CGMS, allowing the clinician to evaluate the glucose curve generated accurately.

After the monitoring period (24–72 hours), the instrument is removed and the site is examined for any signs of redness or irritation. Owners of dogs and cats should be instructed to check periodically for signs of discomfort or irritation at the site of sensor placement or for abnormal behavior (eg rolling, rubbing, or biting at the area of sensor attachment). In the authors' experience, the CGMS is well tolerated with no adverse effects. Because the CGMS recording device weighs approximately 30 g and is the size of a standard deck of cards, even small patients maintain normal mobility. Also, the instrument is quite durable and is capable of withstanding moderate insults.

After removal from the animal, the data are downloaded from the recording unit. The recording device is set into a docking station connected to a computer and the data are downloaded by the manufacturer's software. The software is user-friendly and provides a wealth of data for each monitoring. The data can be viewed as a multiday composite curve for the duration of the recording, as a separate curve for each calendar day, or in a spreadsheet format. Data in the recording device are stored and can be downloaded multiple times if necessary. However, data from each monitoring session must be erased from the recording device before the next monitoring. Data downloaded to the computer can be stored electronically indefinitely.

Use of the CGMS for veterinary species does have limitations. If the sensor detaches from the skin, the system can no longer record glucose concentrations and provides incomplete data. Also, if an inaccurate blood glucose concentration is entered into the CGMS for calibration purposes, a calibration error may occur and may end the monitoring. As stated previously, the system has a working ISF glucose range of 40–400 mg/dL. Interstitial fluid glucose values >400 or <40 mg/dL are not recorded at their true concentration. Therefore, data recorded by the CGMS may not accurately reflect the glucose profile in animals with marked hyper- or hypoglycemia.

Clinical Examples

The following are examples of clinical cases demonstrating the use of the CGMS. These are presented to illustrate the graphical data generated by the CGMS and how these data were used to manage the animals.

Example 1

The CGMS was attached to a clinically normal adult dog and cat (Fig 4A,B) for demonstration purposes. The instrument was initialized, calibrated, and allowed to measure ISF glucose concentration for approximately 24 hours. Note the ISF glucose line tracing typically remains within the normal blood glucose reference interval established by our laboratory. There are times throughout the 24-hour period during which the ISF glucose concentration is above or below the reference interval. These periods of time are transient and appear to be clinically insignificant to the overall health of the animal.

Figure 4.

 Interstitial fluid glucose concentration recorded for approximately 24 hours in a clinically normal dog (A) and cat (B). Open arrows indicate the beginning of glucose monitoring. Normal reference interval for blood glucose level for each species is noted on the graph next to the y-axis and denoted by a hashed line. The start of the monitoring is indicated by an arrow, and the end of monitoring is denoted by a solid arrowhead.

Example 2

A 9-year-old, male castrated, domestic longhair cat diagnosed with diabetes mellitus 6 months before presentation was presented for routine diabetic monitoring with the CGMS. The cat was being managed with insulin therapy (NPH, 4 units q12h, SQ). The cat was allowed to return to its home environment during monitoring. The owners were instructed to follow the cat's routine insulin and feeding schedule. The owners returned the cat to the hospital for calibration of the CGMS once daily. Fluctuations in ISF glucose concentrations over the entire monitoring period were noted on the composite graph (Fig 5). At the start of monitoring, the ISF glucose concentration decreases over a 2-hour period and then gradually increases. Shortly after the evening insulin administration (7:00 pm), the ISF glucose concentration declines dramatically over an 8-hour period. The ISF glucose concentration reaches a nadir of 70 mg/dL at 2:35 am (taken from sensor data provided by the CGMS software). The ISF glucose concentration then increases steadily over the next 5 hours and exceeds the upper limit of detection for the GCMS (400 mg/dL) for another 5 hours. Insulin was administered at approximately 1:00 pm, after which the ISF glucose concentration then steadily decreases to a nadir of 71 mg/dL at 6:30 pm. The owners administered the next insulin dose at approximately 12:30 am, resulting in the morning decline in ISF glucose concentration to a nadir below the lower limit of detection (40 mg/dL) at 3:50 am. Finally, after remaining at 40 mg/dL or less for approximately 30 minutes, the ISF glucose concentration steadily increases. Owing to the dramatic fluctuations in ISF glucose concentration and altered insulin schedule, the cat was switched to longer-acting insulin and the owners were advised on proper care.

Figure 5.

 Composite graph of interstitial fluid glucose levels over an approximately 48-hour monitoring period. The open arrow indicates the beginning of the monitoring period and the solid arrowhead denotes the end of the monitoring period. Times of insulin administration are denoted by “I.” Normal reference interval for blood glucose level for each species is noted on the graph next to the y-axis and denoted by a hashed line. The start of the monitoring is indicated by an arrow, and the end of monitoring is denoted by a solid arrowhead.

Example 3

A 2-year-old, female spayed, yellow Labrador Retriever diagnosed with diabetes mellitus as a puppy presented for polyuria, polydipsia, and weight loss. The dog currently was being managed with insulin therapy (NPH, 15 units q12h, SQ) and a high-fiber diet. After attachment and calibration of the CGMS in the hospital, the dog was returned to her home environment and maintained on her routine schedule of insulin administration, feeding, and exercise. The owner calibrated the instrument daily at home to avoid returning the dog to the hospital. The composite graph of ISF glucose concentrations over the 72-hour monitoring reveals an erratic pattern with little day-to-day continuity (Fig 6A). The average ISF glucose concentration over the entire monitoring period was 253 mg/dL (51–372 mg/dL, SD 68) (data taken from analysis performed by CGMS software). The ISF glucose concentration decreases dramatically to below the normal reference interval on the 3rd day of monitoring. This finding may have been missed on a traditional 12- or 24-hour glucose curve. The owner was counseled on proper care and 2 weeks later the monitoring was repeated. Interestingly, the composite graph of the ISF glucose concentrations revealed a much different pattern (Fig 6B). The overall baseline average interstitial glucose concentration is lower (average 198 mg/dL [118–282 mg/dL, SD 26]) and there appears to be daily continuity. As a result of these findings, it was determined that owner management may be the cause for the erratic ISF glucose concentrations. After proper owner education, the clinical signs resolved.

Figure 6.

 Composite graph of the 1st interstitial fluid (ISF) glucose monitoring period (A) and the 2nd monitoring period 2 weeks later (B). Each ISF monitoring period was for approximately 72 hours. Normal reference interval for blood glucose level for each species is noted on the graph next to the y-axis and is denoted by a hashed line. The start of the monitoring is indicated by an arrow and the end of monitoring is denoted by a solid arrowhead. Note the inconsistent daily continuity between the graphs represented in A and B. Also note the decline in A to hypoglycemia in one of the recording days.

Example 4

A 3-year-old, female spayed, Cavalier King Charles Spaniel previously diagnosed with diabetes mellitus and treated with insulin therapy (NPH, 5 units q12h, SQ) presented for recent onset of polyuria and polydipsia. CGMS assessment was recommended to determine the dog's response to current therapy (Fig 7A–C). After CGMS attachment and calibration, the dog remained in the hospital the 1st night of monitoring. The next afternoon, the dog was sent home and the owners were instructed to follow the dog's routine insulin, feeding, and exercise schedule. There is a clear difference in ISF glucose concentration tracings the afternoon of day 1 (hospital environment with prescribed therapy) and days 2 and 3 (home environment, shaded area on graph). The ISF glucose concentrations are erratic during the hospital stay, but they do not exceed 400 mg/dL. But in the dog's home environment (days 2 and 3), the ISF glucose concentrations exceed the upper limit of detection (400 mg/dL) for an extended period of time. This difference between the ISF glucose concentrations on day 1 and days 2 and 3 was attributed to owner noncompliance with strict feeding instructions. Upon returning to the home environment, the dog was fed excessive additional treats by the owner, which resulted in the spike in ISF glucose concentrations. This problem with compliance may have been missed had a traditional in-hospital blood glucose curve been utilized to assess response to therapy. The owner was advised of proper diet and the insulin dose was adjusted accordingly. The dog's clinical signs resolved after initiation of these management strategies.

Figure 7.

 Consecutive calendar days of approximately 48 hours of interstitial fluid glucose monitoring. A—1st calendar day, B—2nd calendar day, C—3rd calendar day. The start of the monitoring is indicated by an arrow and the end of monitoring is denoted by a solid arrowhead. Stars indicate blood glucose concentration entered into the instrument for calibration. Normal reference interval for blood glucose level for each species is noted on the graph next to the y-axis and denoted by a hashed line. The shaded area indicates the time the dog spent in the home environment.

Additional Uses for the CGMS

In addition to routine evaluation of insulin or dietary therapy response in dogs and cats with diabetes mellitus, the development of the real-time continuous glucose monitorb has widened the application of this technology. The real-time CGMS gives a continuous display of the ISF glucose concentration. The wireless nature of the system minimizes the impact on the patient and necessity for handling. The real-time monitor allows the clinician to detect trends and rapidly identify glucose fluctuations without repeated phlebotomies and blood loss for the patient.

Additional Uses for Diabetic Patients

The real-time CGMS allows continuous monitoring of dogs and cats with diabetic ketoacidosis while avoiding multiple phlebotomies or necessitating the placement of a central venous catheter for blood sampling purposes. Because glucose readings from the CGMS have a good correlation with blood glucose concentrations, blood sampling is limited to that required for the CGMS calibration.1 Finetuning of regular insulin administration either via continuous IV infusion or intermittent IM injection can be performed by simply observing the display reading. Furthermore, the real-time system is wireless, and hence the monitor can be attached to the outside of the patient's cage and readings can be recorded, minimizing patient stress. In this capacity, the CGMS may be a more cost-effective way to closely monitor blood glucose (every 30 minutes to 2 hours).

Intra- and Postoperative Monitoring

The real-time CGMS may be used to monitor diabetics or animals at risk for either hyper- or hypoglycemia during anesthesia. Glucose homeostatic abnormalities during anesthesia pose a substantial risk. For example, global brain damage and postoperative seizures have been associated with intraoperative hypoglycemia in dogs undergoing surgical ligation of congenital portosystemic shunts.c In these patients, frequent (≤ every 30 minutes) blood glucose monitoring may be warranted. Other conditions associated with abnormal glucose homeostasis, such as sepsis, insulinoma, neoplasia, or diabetes mellitus, may also benefit from CGMS monitoring during anesthesia. Because real-time output is provided, results can be recorded every 5 minutes during anesthesia along with routine vital parameter assessment. This frequent assessment allows for rapid identification of glucose alterations and allows the anesthetist and surgeon to make timely decisions about the use of dextrose or insulin in the patient. Additionally, the same CGMS probe can be used post-operatively for 2–3 days to monitor fluctuations in blood glucose concentration.

Critical Care

In recent years, glucose homeostasis has become a focus in critical care. There is a growing body of evidence in human medicine and in animal experimental models indicating that tight glycemic control during critical illnesses such as sepsis leads to a markedly improved outcome. In fact, normalization of blood glucose concentrations may decrease mortality up to 50% in human patients with sepsis.19 Altered mortality is due to the influence of glucose and insulin on the production of inflammatory mediators.20 In veterinary medicine, hyperglycemia is associated with poor outcome in dogs with congestive heart failure.21 Based on information in humans, even mild alternations in blood glucose concentration may alter the inflammatory response and frequent (q15–30min) monitoring may be necessary to allow early intervention. Because monitoring of this nature is time consuming, stressful for the patient, and may be associated with iatrogenic anemia, glucose sampling is limited to every 2–6 hours for most veterinary patients. Substantial alterations in blood glucose concentration can be missed during this long time interval and associated inflammatory, metabolic, neurologic, and osmotic derangements may develop. The real-time CGMS circumvents these shortcomings and allows aggressive glucose monitoring to become a reality in veterinary medicine. Other conditions such as pancreatitis, insulinoma, head trauma, severe babesiosis, xylitol toxicosis, neoplasia, and severe liver dysfunction may also benefit from aggressive glucose monitoring. Additionally, rapid assessment of blood glucose concentrations for animals on partial or total parenteral nutrition, dextrose-containing fluids, or insulin therapy can be made. Continuous glucose monitoring allows rapid assessment of fluctuations in blood glucose concentrations and timely institution of appropriate treatment.

Pediatric Evaluation

Hypoglycemia in pediatric patients can be challenging to monitor. Often, the size of the patient limits blood sampling because of a lack of vascular access or concerns about phlebotomy-induced anemia or hypovolemia. The real-time CGMS can be placed in these patients and used to quickly assess the need for dextrose therapy after initial stabilization. The real-time CGMS will alert the clinician to a pending hypoglycemic crisis and also help avoid complications of dextrose over-supplementation.

Future Directions

With the introduction of new technology comes advanced clinical applications. As mentioned previously, the manufacturer has released a similar CGMS instrument that is wireless and is capable of providing a real-time ISF glucose monitoring. The advantages of this wireless, real-time CGMS are obvious and likely will expand the use of continuous glucose monitoring in veterinary medicine. The commercial CGMS currently available are not designed specifically for veterinary patients. With continued utilization of this technology in veterinary medicine, additional applications will enhance our ability to manage patients with glucose-related disorders.


aCGMS Gold, Medtronic Minimed, Northridge, CA

bCGMS Guardian, Medtronic Minimed

c Torisu S, Washizu M, Hasegawa D, et al. Sustained Severe Hypoglycemia during Surgery as a Genesis of Global Brain Damage in Post Ligation Seizure of Congenital Portosystemic Shunts Dogs. J Vet Intern Med 2006;20:753