Insulin Dosing and Monitoring Protocols

Managing diabetic cats with insulin with the objective of diabetic remission is a balancing act requiring sufficient insulin for glycemic control without causing hypoglycemia. The recommended starting dose for most insulins is 0.25 U/kg to 0.5 U/kg, with higher doses preferred for those cats with higher BG (more than 20 mmol/L [360mg/dL]) recognized at diagnosis.^71,104,106 Frequent re-evaluations are required during the initial stabilization period, as outlined later and in Box 24-1. Typically, most cats first go through a phase where insulin dose is increasing, then stabilize when insulin dose is consistent, and then for cats achieving remission, a phase of decreasing insulin dose.

In human medicine, specialized educators provide most of the information on disease management, insulin injection and handling, BG monitoring, and so forth. In the veterinary setting this function is performed by the veterinarian or often by the veterinary nurse/technician. Owners must be properly trained to give insulin injections and educated on the important aspects of the disease in cats. For example, if some of the insulin dose is spilled during injection, the owner should be warned not to give additional insulin to avoid overdose. Whenever there is uncertainty about whether a dose has been administered, it is safest to wait until the next scheduled dose, because the consequences of missing one dose are negligible. Close contact with owners during the initial weeks of insulin therapy can help identify and correct any problems or misconceptions.

The owner and all members of the household should be aware of the clinical signs of hypoglycemia, which include lethargy, trembling, ataxia, altered mentation, seizures, and coma. If signs suspected to be caused by hypoglycemia are noted at home, the owner should be instructed to give high glucose syrup orally. Suitable products are marketed for use by human diabetics. If the cat is unable to swallow, the syrup can be rubbed on the oral mucosa. When an episode of hypoglycemia occurs or is suspected, the owner should also seek veterinary care and discontinue insulin dosing in the meantime.

It is the authors’ experience that hypoglycemia is rare initially, and if it is recognized biochemically within the first few days of therapy, there is a very low probability of clinical signs. Starting most diabetic cats at 0.5 U/kg increases the chance of establishing timely glycemic control. However, starting a cat on this dose of insulin requires close monitoring from day 1. When glargine was first introduced to management of feline diabetes, initial protocols called for BG testing every 4 hours for each of the first 3 days and then weekly.⁷¹ It is usually more practical to perform BG curves on day 1 and day 3. Long-acting insulins are marketed in human medicine as being peakless; this is not the case (at the very least in cats),⁷² and “spot” checks of BG are not adequate or appropriate to maintain glycemic control of diabetic cats. Serial BG measurements (BG curves) are the most effective monitoring technique to establish what is happening with the cat’s glucose homeostasis. Fructosamine assays are a crude measure reflecting the mean BG concentration for approximately the past week (compared with longer time periods in other species)⁶⁴ and are not an effective measure for tight management of glycemic control. High serum fructosamine concentrations indicate poor glycemic control but give no information about the nadir BG concentration and cannot be used to determine whether the insulin dose must be increased or decreased. In addition, some disease states, such as hyperthyroidism, affect fructosamine concentrations. Cats with overt, uncontrolled hyperthyroidism may have fructosamine concentrations below the normal range, probably as a result of metabolic changes.¹⁰¹

There is no uniformly recognized BG monitoring protocol for diabetic cats. High remission rates have been achieved with weekly in-hospital monitoring of BG curves (stretching to every 2 weeks after 4 weeks)⁷³ as well as with home monitoring up to daily and insulin dose changes instituted as often as every 3 days.^104,106 The following schedule is suggested for BG monitoring either at home or in the hospital (serial measurements are taken on each day):

BG can be evaluated every 3 to 4 hours. After an owner obtains a BG curve at home, especially during the stabilization period, the results should be reported to the clinician or veterinary nurse/technician. An appropriate BG curve for a stable diabetic cat being treating with long-acting insulin (e.g., glargine or detemir) is shown in Figure 24-5. A cat with a curve similar to the one illustrated needs no dose adjustment. However, cats can have such a curve yet need a dosage reduction (or increase) at a subsequent evaluation. In the early stages of treatment with long-acting insulins, the BG curve may resemble a curve obtained with lente insulin (Figure 24-6) with more distinct peaks but will usually flatten within 2 to 3 weeks. The slow absorption of glargine and detemir can result in atypical shaped curves with elevations of BG in the middle of the day (but this usually only occurs when there is minimal change in BG). Any dose adjustment decisions should be based on the lowest BG level of the day and should always take into account the patient’s clinical assessment (e.g., body weight, body condition score, appetite, water intake, and urine output).

FIGURE 24-5 Ideal blood glucose curve for stable diabetic cat using long-acting insulin (e.g., glargine, detemir). Note that there typically is a nadir, but the variation from preinsulin blood glucose to nadir is not great.

FIGURE 24-6 Ideal blood glucose curve for stable diabetic cat using intermediate-acting insulin (e.g., lente). Note the extended periods of time that blood glucose remains above the renal threshold.

An appropriate curve for a stable diabetic cat being treating with an intermediate-acting insulin (e.g., Vetsulin/Caninsulin or Humulin-N) is shown in Figure 24-6. Note the extended periods of time that BG remains above 15 mmol/L (above 260 mg/dL).

Common sense should be used when making any insulin dose adjustments. Suggested guidelines for dosage adjustments are shown in Table 24-3. Some guidelines for long-acting insulin dose adjustments also include recommendations based on the pre-insulin BG. The reason for monitoring is to assess the effect that particular dose of insulin has had, so changing the dose on the morning of testing defeats that purpose. The exception to this rule is when BG concentration is low before administration of insulin, in which case that dose should be skipped and the dose administered 12 hours later should be lower. As for long-acting insulins, dose adjustments should always take into account the patient’s clinical assessment (e.g., body weight, body condition score, appetite, water intake, and urine output).

TABLE 24-3 Suggested Guidelines for Insulin Dose Adjustment Based on Blood Glucose Curve Results

Once a cat achieves remission, ongoing BG monitoring should continue for the first month. Over the long term, urine glucose, body condition, appetite, and water intake can be monitored to detect loss of euglycemia. If hyperglycemia is detected, insulin therapy should be restarted promptly to avoid further damage to pancreatic beta cells. A low-carbohydrate, calorie-controlled diet often helps prevent obesity, minimize the demand on beta cells, and lower the risk of relapse.

Changing Insulins

No washout period is needed to change insulin type. The new insulin can be substituted at the next scheduled dose. Recommendations for changing from one product to another for well-regulated cats include the following examples:

In all cases, if a patient is currently not well regulated, the new insulin should be started at the recommended starting dose for the product and regulated as for a newly diagnosed patient.

Glucometer Choice

Handheld glucometers to assess BG at the point of care have become standard in the management of DM in animals as well as humans. Only a limited number of glucometers have been critically assessed in the veterinary literature, with varying results.^{15,22,65,125,136} Accuracy is assessed by precision (variation of the individual glucometer for a particular sample) and bias (variation from a reference laboratory measurement). Accuracy is most often poor when very low or very high BG concentrations are measured.

Most glucometers are calibrated for human use. The distribution of glucose between plasma and red blood cells is different in feline blood compared with human blood. Glucometers for human use generally read lower (approximately 1 to 2 mmol/L [18 to 36 mg/dL]) than measurements from an automated chemistry analyzer. To further confuse matters, whole BG (measured with a glucometer) is approximately 10% lower than glycolysis-inhibited plasma glucose (“gray-top” tube samples); also compared with whole BG measurements, serum glucose reduces by approximately 5% to 7% per hour as a result of continued metabolism of glucose by the red blood cells.⁹⁷ Even glycolysis inhibitors contained in gray-top blood tubes (e.g., sodium fluoride/potassium oxalate) result in reduced plasma glucose readings when compared with promptly spun serum samples.¹²² Additionally, capillary BG measurements can be between 10% and 24% higher than venous whole BG in a nonfasted state.⁵⁸ Despite these multiple variations, studies in cats have shown minimal variation between samples obtained from the pinna versus peripheral vein,^117,126 and a further study assessed glucometers by directly comparing capillary pinna samples to venipuncture collected samples.¹³⁶

The highest degree of accuracy among the glucometers tested to date has been shown with the AlphaTRAK glucometer (Abbott Laboratories), one of the few machines calibrated for use in veterinary patients. Although the AlphaTRAK is not as easy to use as human-medicine glucometers, it requires a very small sample volume (0.3 µL) and the test time is only 15 seconds. However, care must be taken with this glucometer insofar as it may be calibrated for canine or feline blood and accuracy is greatly diminished if it is set incorrectly.¹³⁶ The Ascencia Elite (Bayer Healthcare) glucometer has also been demonstrated to be accurate^125,136 but reads lower than the AlphaTRAK and reference laboratory samples,¹³⁶ which potentially means that lower nadirs will be recognized.^104,106 Other veterinary glucometers include GlucoPet and GlucoVet (Animal Diabetes Management, Janesville, Wisconsin), although they have not been critically assessed as of this writing. Glucometers should be selected carefully for use in practice and owners’ use at home, with the clinician and owner having an awareness of the bias of the glucometer compared with laboratory measurements of glucose. Ideally, the clinician would calibrate each glucometer used at home against the hospital’s own blood chemistry analyzer or calibrated glucometer.

Home Monitoring

Home monitoring is often used to refer to monitoring of BG values, but all owners also need to assess how well their cat is eating (either increased or decreased appetite) and drinking and whether its general behavior is normal. Ideally, all diabetic cats would receive daily monitoring at a minimum. The ability to record body weight at home is a very valuable monitoring tool. Owners should be encouraged to buy inexpensive scales suitable for measuring small patients (e.g., scales used to weigh human infants) and to keep a weekly log of body weight.

Most owners of diabetic cats are initially very concerned about their cat’s diagnosis, and many are concerned about their ability to use syringes to inject insulin (although most soon find it is easier than giving oral medications to their cats), let alone take BG measurements. Home monitoring of BG is the goal, but clinicians should introduce these concepts to their clients and demonstrate how to obtain blood samples in a stepwise manner during visits in the stabilization period.

Perceived benefits of home BG curves include obtaining samples in a less stressful environment where the cat’s food intake should be normal and the ability to perform more frequent BG monitoring. Many owners have the dedication and the ability to perform BG curves at home, but many do not realize this at first and can be intimidated by this prospect. As already noted, success is most likely when additional tasks for the owner of a newly diagnosed diabetic cat are sequentially introduced. Showing owners how to take a BG sample at the same time they are shown how to give an insulin injection for the first time is likely to reduce compliance. Some owners are happy to proceed with sampling their cat’s BG after being shown only once; others need reassurance and multiple demonstrations. Common reasons for their initial reluctance to attempt home BG monitoring include fear of hurting the cat, fear of taking a blood sample, and concern over cost and time commitment. Watching an owner perform the procedure allows the clinician to identify and correct problems. Providing reliable and easy to understand educational resources as well as ready access to telephone support is also important.¹¹⁹ Long-term compliance with home BG monitoring is reportedly good, with 65% of owners performing it regularly in one study. Most owners in that study performed home BG curves every 2 to 4 weeks. Home monitoring provides owners with more confidence in their ability to manage their cat’s disease.⁵⁴

Alternatively, some owners become overconfident in their interpretive abilities and will not seek continued veterinary care until dramatic signs arise. It is important for owners to understand that information about protocols used to treat human diabetics cannot be applied to cats. In veterinary medicine, unlike human medicine, the goal in treating diabetes is not normoglycemia, because even some well-controlled diabetics can be at least slightly hyperglycemic at some time during the day.

In one study 12% of owners made changes to their cat’s insulin dose without consulting the clinician. Owners should be cautioned against the temptation to make frequent changes to insulin dose and should make changes only after consultation with a veterinarian.⁵⁴ Another study found that approximately one fourth of cats had higher BG concentrations for 2 to 3 days after a glargine insulin dose increase.¹⁰⁴ The reasons for this are not known but may be associated with a lag time before feedback mechanisms adjust glucose homeostasis. The authors’ experience is that the instinct of most owners is to further increase the insulin in this circumstance, sometimes on three consecutive days, which leads to an increased risk of hypoglycemia.

It must be emphasized that management of the diabetic cat is a team effort that includes the owner, the clinician, and often support staff such as veterinary technicians and nurses. Regular re-evaluations should be established to monitor weight and body condition score, look for underlying issues (e.g., dental infection), and confirm BG findings and discuss the owner’s findings at home (see Box 24-1). One study has shown that the number of re-evaluations did not differ significantly between cats managed with and without home monitoring.⁵⁴ Once glycemic control has been achieved, diabetic cats should be re-evaluated approximately every 3 months.

BG concentrations measured at home have been compared with those measured in hospital in one study; surprisingly, hospital measurements were lower than those obtained at home.¹³ This may have been due to inappetence in hospital or stress from the less experienced owner taking the sample at home. In about 40% of cats, the hypothetical treatment decisions based on the two methods of generating BG curves did not agree. Alternatively, it has been demonstrated that BG readings can be significantly higher in hospital (Figure 24-7).¹⁰⁰

FIGURE 24-7 Blood glucose curve (BGC) at home and in the hospital 3 days later in a cat with diabetes mellitus (spayed female domestic shorthair, 4.7 kg). The cat was diabetic for 1 year and received 0.5 U/cat Caninsulin/Vetsulin every 12 hours at the time when the two curves were generated. According to the owner, the cat was doing well and had no signs of diabetes mellitus. The high blood glucose levels measured in the hospital were assumed to be stress-induced. Insulin dosage was not changed and the cat continued to do well. Blood glucose is noted in mmol/L; to convert to mg/dL, multiply by 18.

(From Reusch C, Kley S, Casella M: Home monitoring of the diabetic cat, J Feline Med Surg 8:119, 2006.)

Another study compared paired BG curves generated at home on 2 consecutive days in seven diabetic cats. In the second part of the study, two BG curves generated at home and one BG curve generated in hospital were compared. The results demonstrated considerable day-to-day variation in BG curves, even those obtained at home. The day-to-day variability between the home and hospital BG curves was not larger than between the paired home BG curves. There was less variability in the home BG curves obtained from cats with good glycemic control than from cats with moderate to poor glycemic control. These findings would indicate that a single BG curve may not always be reliable for making treatment decisions, whether performed in the hospital or at home.¹

In humans variations in BG concentrations occur within a 24-hour period as well as day to day and can be associated with activity level, meal size and type, stress, and some medications. Even when those factors are controlled, other factors, such as varying rates of insulin absorption from different injection sites and variation in length of insulin activity, cause day-to-day variability in BG concentrations. Many of these factors probably play a role in the variability of BG curves in cats as well. In particular, the role of absorption from different injection sites has not been investigated in cats. As more about day-to-day variability in BG measurements in cats becomes known, it is important not to rely solely on BG values but also to take into account the physical examination and other laboratory findings (e.g., weight loss or gain, urine specific gravity) and the owner’s observations (e.g., appetite, water intake, activity level) when considering insulin dose adjustments.

Sites that may be used for home BG sampling are the ear^54,100,126 and the foot pads.¹³⁴ The pinnae (either the haired or nonhaired surface) are most commonly used to obtain samples (Figure 24-8). The ear should be warmed by rubbing or holding a heated cotton ball against the ear (the cotton ball can be heated by placing in a standard microwave for 10 seconds). The ear should not be cleaned or disinfected with alcohol. In longhaired cats, it may be helpful to shave a small portion of the pinna to improve visibility or to apply a thin film of petroleum jelly (petrolatum) to prevent the blood from dissipating into the fur. The opposite side of the ear can be stabilized with a cylindrical object, such as a roll of bandage tape. Then a lancet device or a 25-g hypodermic needle (without a syringe attached) can be used just near (but not directly over) the marginal ear vein to produce a blood drop. The test strip of the glucometer is then touched to the drop to absorb blood. While the machine is processing the sample, a piece of gauze or cotton ball can be used to apply pressure to the site to stop bleeding.

FIGURE 24-8 A, The pinnae are commonly used to obtain blood glucose samples. The opposite side of the ear can be stabilized with a cylindrical object, such as a roll of bandage tape. B, Then, a lancet device or a 25-g hypodermic needle (without a syringe attached) can be used just near (but not directly over) the marginal ear vein to produce a blood drop. C, The test strip of the glucometer is then touched to the drop and to absorb blood.

The main carpal (or tarsal) foot pads are usually easiest to use for sampling because they are the largest. Again, the pad should be warmed before a lancing device or hypodermic needle is used to prick the pad. Regardless of the site used for sampling, it can be helpful to place the cat in a favorite and comfortable spot for blood collection. The puncture site is usually not painful or visible.

Lancet devices are readily available from most pharmacists, and it is appropriate for the clinicians to have several brands in stock to try on individual cats because cats react differently. Two methods of capillary sampling from the ear have been described.¹²⁶ One method, using a conventional lancing device, was evaluated only in dogs but has also been used in cats. Alternatively, a lancing device that uses negative pressure (Ascensia Microlet Vaculance, Bayer Diagnostics) to help ensure a blood drop of adequate size can be used.⁹⁹

As may be expected, certain difficulties are commonly encountered when owners initially begin home BG monitoring, such as the need for assistance to restrain the cat, numerous attempts to produce a blood drop of sufficient size, and resistance from the cat. These difficulties tend to decrease over time as the owner gains experience with the technique.¹¹⁹

Continuous glucose monitoring systems have been designed to help human diabetics improve glycemic control and have recently been introduced in veterinary medicine. These devices measure interstitial glucose concentrations through a subcutaneously placed sensor. The sensor and transmitter are secured to the patient’s back or side by bandaging. The data are relayed to a monitor that must be placed within several feet of the patient to record readings. Glucose readings are taken every 5 minutes, allowing for a large number of readings in a 24-hour period. Validation and use of these devices in cats have been described for both in-hospital and at-home use. Currently, drawbacks include the need for calibration every 12 hours and the need to remove the sensor after 72 hours, as well as the cost.*

Home monitoring urine for glucose and ketones may be an appropriate additional form of monitoring for some patients, particularly management of fractious cats. The main purpose of once- or twice-weekly urine monitoring is detection of uncontrolled hyperglycemia through persistently high urine glucose readings or, conversely, the detection of hypoglycemia through persistently negative urine glucose readings for cats receiving PZI or lente insulin. Persistently negative urine glucose readings may also indicate that diabetic remission is imminent. Negative urine glucose readings are commonly seen in well-controlled cats on glargine or detemir insulin. Owners should be cautioned never to change insulin dose solely on the basis of urine glucose readings.

Glucotest granules (Purina Veterinary Diagnostics) are a litter additive designed to detect glucosuria through a color change. In one study of the Glucotest granules, the glucose concentration was measured accurately in 29 of 48 feline urine samples. Of the inaccurate measurements, most were overestimates in the 2.8 to 16.6 mmol/L (50 to 300 mg/dL) range. Measurements taken at 8 hours were more accurate than immediate measurements. However, the researchers used frozen urine with added glucose rather than urine from diabetic cats, which may affect results.³⁰

Urine dipsticks are commonly used to measure glucose as well as other urine chemistries and may be used by owners at home. The accuracy of these sticks for feline urine is not well investigated. The ability of one brand (Bayer Multistix) to detect glucosuria in feline urine was determined by comparison to a chemistry analyzer. Sensitivity (73%), specificity (97%), and positive (73%) and negative (97%) predictive values were calculated. Overall, the accuracy of the test strip for classifying the urine glucose concentration in the correct interval was 91% (compared with 59% in dogs). Inaccuracies tended to be underestimates. The study was performed on urine samples submitted to a laboratory, however, whereas most owners are testing urine voided into a litter box, which may affect results.⁶ Well-regulated diabetic cats receiving PZI or lente insulin should have urine glucose readings between trace and 1+, although well-regulated cats on glargine or detemir insulin may have negative or trace readings. Persistent values outside this range should prompt the owner to seek veterinary advice.

Monitoring the fractious diabetic cat is a problem well known to most veterinarians. Fortunately, home BG monitoring can often be performed for these patients, even if they require sedation to obtain BG samples in the hospital. Some difficult-to-handle cats can be restrained enough for safe access to the pinna or the saphenous vein or for placement of a heparinized catheter for repeat blood sampling. Results should be interpreted in light of stress, but it is the author’s (RMB) experience that results consistent with clinical signs (and at-home results) can be achieved in most cases in these circumstances. Other tools owners can use for monitoring of these difficult patients include measuring water consumption, monitoring body weight and condition, and measuring urine glucose. Close attention should be paid to appetite, behavior, and so on. Periodic re-evaluation in the hospital, even if sedation is required, should include a full physical examination, serum fructosamine, and minimum database testing (complete blood count, serum chemistries, urinalysis, urine culture, total T₄) as required.