Unlike NPH, Lente, and Ultralente, it is a clear, colorless solution—not cloudy. To aid in distinguishing glargine from other clear insulins (Regular, lispro, and aspart), it is in a different shaped vial with a lavender color code. It also is available in cartridges for a pen delivery system. Glargine cannot be mixed with other insulins (in fact, the package insert states that the syringe must not contain any other medication or residue), because the solution would become cloudy and cause a precipitate, the pharmacokinetics of each insulin would be altered, and most importantly, glargine would not be able to pass through the needle of the syringe.

Most appreciable about clear insulin compared to cloudy insulin is the consistency of absorption. Many factors can influence rate of insulin absorption and alter insulin availability. Study findings show that insulin absorption and action can vary by 50% from patient to patient and by 25% in the same patient from one day to the next.¹⁰ In fact, as much as 80% of the daily variation in the blood glucose of patients taking NPH insulin is attributed to variations in insulin absorption.¹⁰ There is much less variation in absorption of shorter-acting insulin (Regular, lispro, and aspart) and greater variation in absorption of longer-acting insulin (NPH, lente, ultralente), partly because patients often don’t mix cloudy suspensions of insulin enough before administration. It is further speculated that a clear solution shows a more even distribution in the subcutaneous tissue, which reduces the variability in effect. Studies showed that glargine, as compared with NPH, led to a reduced variability of fasting blood glucose and a lower risk of hypoglycemic episodes.¹¹ Furthermore, compared to NPH, glargine is associated with less nocturnal hypoglycemia and better glycemic control after dinner. These data suggest that the target fasting blood glucose can be lower for insulin glargine than for NPH, with less fear of nocturnal hypoglycemia.¹² In addition, in a study of Type 2 patients, treatment with glargine showed significantly less weight gain than treatment with NPH insulin. It was presumed that despite the improved glycemic control with glargine, the decrease in hypoglycemia meant that there was less supplemental calorie intake usually associated with treating hypoglycemia.¹³ Finally, a study with Type 1 diabetes patients showed that those receiving NPH insulin exhibited a characteristic rise in fasting blood glucose between 5 and 8 AM, consistent with the short duration of action and the lack of suppression of the characteristic early morning hyperglycemia (dawn phenomenon). In contrast, the glargine-treated patients had a lower fasting blood glucose, associated with maintained suppression of glucose levels during these morning hours.¹⁴ Figure 4, from the Lantus package insert, shows the difference between NPH and glargine.

Injection sites:  As noted in Chapter 1, NPH and lente insulins vary in their absorption rate depending on the region used (abdomen has the fastest absorption, buttocks the slowest). Just as with rapid-acting insulin analogs, there is little or no difference in the absorption rate of glargine between the main subcutaneous injection sites.¹⁵ Patients have reported a higher incidence of injection-site pain with use of glargine (2.7%) compared to those using NPH (0.7%), but the pain was mild and did not result in any patients discontinuing using glargine.¹⁶

Dosage: Remember that hypoglycemia may occur at different times with different insulin preparations, and when a patient is switching from one type to another, you should recommend careful observation and frequent home blood glucose monitoring. Glargine’s manufacturer recommends that if a patient is changing from a twice-daily NPH schedule to glargine, consider a 20% reduction in the dosage. With a once-a-day NPH schedule, no decrease is indicated.¹⁶ While Glargine is usually given once daily at bedtime, a recent study noted it is safe and effective whether given before breakfast, before dinner, or at bedtime, which may allow for more flexibility.¹⁷ Another study reported that 25% of their study patients with Type 1 diabetes required twice daily doses of Glargine to achieve acceptable glycemic control.¹⁸ As with any insulin recommendation, individual needs may vary.

Detemir (Levemir)

Detemir is the final long-acting basal insulin analog from Novo Nordisk, approved by the FDA in 2005 and launched in 2006. It is indicated for once- or twice-daily injections. It is produced by a process that includes expression of recombinant DNA in Saccharomyces cerevisiae followed by chemical modification. It differs from human insulin in that the amino acid threonine in position B30 has been omitted, and a C14 fatty acid chain has been attached to the amino acid B29. It is a clear neutral solution, meaning that it does not have to be resuspended as does NPH. The mean duration of action of detemir ranged from 5.7 hours at the lowest to 23.2 hours at the highest dose, meaning that the duration of this insulin is dosage-dependent.¹⁹ Detemir should not be diluted or mixed with any other insulin preparation. Further studies and experience will elucidate the differences between glargine and detemir, as well as indications for use of this new, long-acting insulin analog.

Dosage regimens

There are many dosage regimens for insulin. Some of these regimens and their potential problems are discussed below.

Combination therapy

Some patients with Type 2 diabetes do well on a regimen of oral hypoglycemic agents during the day supplemented with one dose of NPH or long-acting insulin at night.

The liver releases glucose during the night to provide the body with glucose when the body is not receiving nourishment and also to provide a surge of glucose in the early morning. Patients with diabetes tend to have an increased outpouring of glucose from the liver (known as increased hepatic glucose output), and an impaired insulin secretion, which results in an inability to control this basal hepatic glucose production.²⁰ Some patients with diabetes can achieve successful control of their blood glucose with only one shot of insulin at bedtime, which suppresses the liver and results in a normal fasting blood glucose.¹⁶ If the patient can wake up with a normal blood glucose level, the oral agent is more likely to work efficiently to maintain normal glucose during the day.

Combination therapy has many advantages. Because patients take only one shot of insulin per day, they accept and adhere to the regimen more readily. It is easy to remember to administer one shot at bedtime, and patients do not have the inconvenience of carrying a syringe during the day.  There also is a reduced possibility of nocturnal hypoglycemia since the bedtime long-acting insulin will peak before breakfast when the dawn phenomenon is most pronounced.

You can instruct your patients to follow a common protocol:

• The patient should be kept on the current oral agent, but start with a low dose of 10 units of NPH or long-acting insulin at HS (at bedtime).¹⁶ For the safety of the patient, especially in the elderly, it is common to start with a dose as low as 5 units. Increase by 1 or 2 units every three or four days until the fasting blood glucose (FBS) is between 90 and 130.
• Another option is to use a premixed insulin (70/30, Novolog mix 70/30, or Humalog Mix 75/25) at dinner, with oral agents at breakfast and lunch.
• An attempt should be made to get the patient’s FBS as normal as possible to start the day off right. This increases the likelihood that the oral agent will take care of the daytime blood glucose levels.
• The larger the dose of NPH or detemir,  the longer it will act.
• Sometimes when a longer-acting oral agent is used, it may be more difficult for the patient to control the prelunch and predinner blood glucose levels. In this case, patients may be switched to faster-onset or shorter-acting oral agents, which can be given before each meal if needed.
• If predinner blood glucose levels remain high, the patient may have to discontinue oral agents and use insulin exclusively.

Dosage regimens

One daily dose of NPH: When insulin was first discovered, people took frequent shots of short-acting Regular insulin, the only insulin on the market. When NPH became available in 1950, it became the preferred regimen because it was easier to take one dose that supposedly would last 24 hours. One potential problem with this regimen is there is no coverage for meals. Moreover, by the next morning, the insulin wanes while the dawn phenomenon takes over. Also, when the insulin peaks in between six and 12 hours, a low blood glucose is likely to occur. A typical blood glucose profile of a patient on one dose of NPH would be: breakfast, 250; lunch, 300; dinner, 60; bedtime, 150; 3 AM, 180; 6 AM, 250.

To counteract this problem, a single mixed dose of NPH and Regular insulin was often recommended. 70/30 insulin, or better yet, a premixed analog (as described in the previous section), achieves an earlier onset for insulin action between breakfast and lunch, but does not solve the problem of the early-morning rise in glucose levels from the dawn phenomenon.¹⁷ In current practice, although this regimen is occasionally seen, it is relatively uncommon and not recommended unless the patient is able to achieve blood glucose levels consistently in the target range.

Two daily doses: One of the more common regimens is two doses (one dose before breakfast, the other before dinner) of NPH and a rapid-acting analog, or NPH and Regular insulin. The rapid-acting analog, or Regular supplies insulin after breakfast and dinner, while NPH provides coverage after lunch and during the night.²¹ As explained previously, the postprandial glucose would be lower, with less hypoglycemia later, with a rapid-acting analog compared to Regular. For this type of regimen, some patients may be able to use Humalog Mix 75/25 or 70/30, or NovoLog Mix 70/30 insulin. Some diabetes specialists tend not to recommend these premixed insulins unless necessary, because there is no way to manipulate the ratio of short-acting insulin to NPH, to compensate for changes in food, stress, or activity. These premixed regimens are rarely effective with Type 1 patients. However, in the elderly patient with mild or stable diabetes or in the stable patient with Type 2 diabetes, premixed insulins have the advantage of allowing the patient to draw up only one type of insulin or to use an insulin pen (a prefilled syringe containing a cartridge of insulin or a prefilled disposable pen, both of which can be “dialed” and delivered). The ease of use with this insulin may lead to better accuracy, improved transportability (if using a pen), and perhaps better patient compliance. A potential problem with the two doses of NPH/Humalog, NPH/NovoLog, NPH/Regular, Humalog Mix 75/25, Novolog Mix 70/30, or 70/30 is hypoglycemia. To counteract this, ask your patients to eat lunch and dinner on time, to monitor blood glucose regularly, and to eat an afternoon snack if increased activity is planned or if dinner is delayed.

Three daily doses: Two shots of insulin a day has been a common regimen, but it is difficult to maintain target glucose levels with two daily doses. Sometimes, increasing the PM NPH makes blood glucose drop too low during the night, yet lowering the PM NPH makes the fasting glucose too high. In other words, the PM NPH peaks in the middle of the night rather than in the early morning. A logical solution is for the patient to take three doses of insulin a day. In this regimen, the PM dose is divided by giving only a rapid-acting analog or Regular insulin before dinner and NPH at bedtime.¹⁷ This way, the bedtime insulin peaks when you want it to (before breakfast), which prevents both the dawn phenomenon and the Somogyi phenomenon. However, the patient usually must have a bedtime snack to prevent nocturnal hypoglycemia. As described previously, nocturnal hypoglycemia occurs less with a rapid-acting analog given at dinner compared to Regular given at dinner.

Four doses a day:  As mentioned, the action of cloudy insulin is more erratic than that of clear insulin. Patients have been known to complain that they would eat the same food, take the same dose, but have widely fluctuating blood sugars. After eliminating certain variables (injection technique, timing of insulin, assessment of food intake), often the most contributing factor was the cloudy insulin, which can have widely varying activity and absorption compared to that of clear insulin.⁴ As a result, long-acting insulin analogs have become more popular while Lente and Ultralente insulins have been discontinued. Furthermore, to remedy the above-mentioned problem, various insulin regimens are available that use rapid-acting analog or Regular with each meal and some type of background insulin at bedtime (NPH, glargine, or detemir), or background insulin at breakfast and dinner or breakfast and bed, along with a rapid-acting analog or Regular with each meal. These regimens allow patients with hectic lifestyles and erratic meal or activity patterns to maintain their lifestyles and maintain good glycemic control. But except for the newest insulin glargine, the disadvantage of many of the multiple-shot regimens is that the background/basal insulin would not be available at a consistent, peakless level all day. If too many hours elapse between meals without adequate insulin on board, the liver releases glucose, and  not enough insulin is available to suppress the liver. Moreover, if patients skip even one shot of meal insulin during the day, and definitely if they eat without taking rapid- or short-acting insulin, they may have rapid glucose elevation.

Table 4

On the next page is a summary of various insulin regimens. Remember that no two patients are alike—their situations and needs vary, along with duration of diabetes and residual beta-cell function. Furthermore, 15 to 20 years after the onset of Type 2 diabetes, many patients will have significantly reduced insulin secretion, which will require the use of multiple shot regimens similar to those used for Type 1 patients.²² Meticulous record keeping, analysis, and trial and error will help determine which regimen works best.

Insulin pumps

In Type 1 patients of longer duration who have no residual beta-cell function and thus no basal insulin available, or for some Type 2 patients who have declining beta-cell residual, an insulin pump may be the choice since it easily provides the basal and bolus insulin needed.

An insulin pump is about the size of a beeper and contains a motor and a special syringe. The motor causes the syringe to push out tiny drops of insulin every few minutes on a continuous basis. The patient presses a button to deliver the desired dose of bolus insulin.

Insulin pumps attempt to mimic the action of the pancreas by using only short-acting Regular or rapid-acting lispro or aspart insulin. Currently, Humalog, Novolog, and Apidra all have FDA approval for use in pumps. The pump supplies a low background dose or basal dose of insulin. An additional amount of insulin (bolus dose) also is delivered before each meal or snack. The basal dose usually is 40% to 50% of the total amount of insulin, while the boluses make up the remaining 50% to 60%.²³

Insulin pumps cost about $6,000, plus about $175 a month for pump syringes, infusion sets with either a needle or Teflon catheter, batteries, and tape. Pumps have existed for more than 20 years, and their popularity is rapidly increasing because patients want improved control. In addition, not only has pump technology and pump appearance improved and the device shrunk, but there has been a significant increase in technical support and training provided by insulin pump companies and contract pump trainers.

The pump provides maximum freedom and flexibility in terms of meal times and activity. If the basal dose is correct, the patient can skip a meal and not get hypoglycemic as long as he or she does not perform any strenuous activity. Also, the likelihood of hypoglycemia decreases since there never is a large amount of insulin in the patient’s system, and there is less variability of absorption. Moreover, the pumps have pre-programmable basal rates that can be adjusted to solve the dawn phenomenon or any other variations in glucose levels. Thus, an insulin pump may be an excellent option to achieve target glucose levels.

The disadvantages of the insulin pump include an increased (although small) chance of infection at the site since the needle remains under the skin for two to three days and the possibility of diabetic ketoacidosis. If the needle dislodged or the pump malfunctioned, insulin delivery would abruptly stop. Because the patient would have only a small amount of short-acting insulin on board, the insulin effect would be gone in two hours. Without the protective effect of a longer-acting insulin in the bloodstream, blood glucose could rise rapidly to dangerous levels. However, with proper patient selection, education, and frequent glucose monitoring, this potentially serious complication is uncommon.

It must be emphasized that not all patients are candidates for insulin pump therapy. Generally, an insulin pump is considered for the following conditions: persistently poor glucose control with wide fluctuations in blood glucose; persistent dawn phenomenon; day-to-day variations in school or work schedule, mealtimes, or activity levels, which confound the degree of regimentation required to manage glucose with multiple injections; and preconception or pregnancy with a history of poor glycemic control. A potential insulin pump candidate must be highly motivated to achieve and maintain good glycemic control and commit to self-monitoring of blood glucose at least four times a day. Finally, insulin pump therapy requires a qualified and accessible health care team to assess, train, and maintain ongoing management of the insulin pump patient.

It is beyond the scope of this course to provide the details of managing patients on insulin pumps. Recommended reading includes workbooks from insulin pump companies; Pumping Insulin, Everything You Need for Success With an Insulin Pump, 3rd edition, by John Walsh, PA, CDE, and Ruth Roberts, MA, published in 2000 by Torrey Pines Press in San Diego; and Smart Pumping: A Practical Approach to the Insulin Pump, by Howard A. Wolpert, MD, 2002, the American Diabetes Association.

Conclusion

Newer insulin analogs more closely mimic the body’s normal response to insulin and allow patients to fit diabetes into their lives; they can “dose and eat,” simplifying their scheduling of injections to cover meals. Patients have experienced improved fasting and postmeal glucose levels without increased incidence of hypoglycemia.

Numerous options for insulin regimens exist. Each patient is different, and meticulous record keeping, analysis, and trial and error will help determine which regimen works best. Further studies and experience will help us use these new insulins most effectively.

Chapter Three
Adjusting Insulin: General Guidelines and the Retrospective Approach

Chapter 1 explained the actions and types of insulin and common problems of blood glucose levels. Chapter 2 described the newer insulin analogs and dosage regimens available. This chapter provides basic guidelines for adjusting insulin doses and explains the retrospective approach to changing insulin.

Presented here are basic principles of dosage adjustment assuming a target goal of 90-130 mg/dl before meals, but patients’ needs and responses can vary. For that reason, if you are involved in adjusting insulin doses or in teaching patients how to adjust doses, you must develop a detailed protocol approved by your medical director or your institution.

General dosage guidelines

The following are basic guidelines for adjusting insulin. 

• It is important to keep pattern control in mind: Too often, decisions to change dosage are based on one day or even on one abnormal glucose level; you must look at the pattern of the blood glucose levels over at least three to five consecutive days.¹ Reassure patients not to worry about isolated, unexplained abnormal blood glucose levels.
• “Fix the fasting first” is another general rule. High blood glucose confers some insulin resistance; thus, an elevated fasting blood glucose requires a larger dose of insulin to lower it. Generally, if patients start with a normal blood glucose level in the morning, they can more easily control glucose levels for the rest of the day.²
• Because of the dawn phenomenon (see Figure 2, page 5), glucose released from the liver in the early morning needs to be controlled; patients often may need a bedtime dose of insulin.
• Only one insulin should be adjusted at a time. Insulin action seems to have a ripple effect: If blood glucose is higher at lunch, for example, it will stay higher at dinner and at bedtime. In this instance, solve the first problem that you encounter in the day’s pattern of blood glucose levels. After you solve that, the rest of the day’s pattern may fall into range.
• Remember that with Regular, NPH, and detemir, insulin action is dose-dependent:  The more you give, the longer it lasts.^3,4
• The dose should be changed by 10% at a time.⁵ The general rule is that if blood glucose is lower than target level, cut the insulin by two to four units; if blood glucose is higher than target level, raise the insulin by only one or two units at a time.

n  In general, you should wait three to five days to get the full effect of the dose before making further adjustments; however, there are exceptions to this rule. The above guidelines apply to nonemergency adjustments. If glucose is very high (over 300 mg/dl), small supplements of short- or rapid-acting insulin may be needed to quickly bring down the glucose levels.⁵ If blood glucose is low, you may have to cut the dose daily to eliminate this problem.

Algorithms

Algorithms are guidelines that tell you how much insulin to give at each injection.  Algorithms help you understand when insulin peaks and which dose should be changed. Algorithms are the basis of self-adjustment for insulin; they allow the patient to make daily insulin choices based on blood glucose levels, food intake, exercise, and variations in normal routine.

There are two approaches to adjusting insulin. In the prospective approach, the patient can make daily adjustments in the short-acting or rapid-acting insulin dosage based on anticipated food intake and exercise and the prevailing blood glucose level.⁵ In the retrospective approach, insulin doses are changed in response to levels of control occurring throughout the previous three to five days.⁵

This approach obviously requires a relatively consistent diet and exercise pattern. The following algorithms are based on a retrospective approach (see Table 1, page 4 for actions of insulin); the target level of blood glucose used for these examples is 90-130 mg/dl before meals. Section 1 of the following two sections describes algorithms for adjusting insulin when using two or three shots a day. Section 2 describes algorithms when using four shots a day.

Section 1

The following is an algorithm for adjusting insulin using a split and mixed dosage for basal insulin and short-acting or rapid-acting insulin. It is based on four premises.¹

1. The morning dose of short-acting or rapid-acting insulin has its major action between breakfast and lunch, and its effect is reflected in the blood test results after breakfast and before lunch.
2. The morning basal insulin has its major action between lunch and dinner, and its effect is reflected in the blood glucose results after lunch and before dinner.
3. The evening dose of short-acting or rapid-acting insulin has its major action between dinner and HS, and its effect is reflected in the blood test results after dinner and before bed.

The evening or HS basal has its major action overnight, and its effect is reflected in the blood glucose result taken during the middle of the night and the next morning.

Algorithm for hyperglycemia not explained by unusual diet or exercise, appearing at the same time for three to five days in a row:¹

• If prebreakfast glucose is over 130 mg/ dl, and the 3 AM glucose level is not low, increase the dinner or HS dose of basal by one or two units.
• If prelunch glucose is over 130 mg/dl, increase the morning short-acting or rapid-acting insulin dose by one or two units.
• If the predinner glucose is over 130 mg/dl, increase the morning basal by one or two units.
• If HS glucose is over 130 mg/dl, increase the predinner short-acting or rapid-acting insulin dose by one or two units.

Algorithm for hypoglycemia not explained by unusual diet or exercise:¹

• If prebreakfast blood glucose is less than 90 mg/dl or there is evidence of hypoglycemia during the night, decrease the dinner or at bedtime basal insulin dose by one or two units.
• If the prelunch test is less than 90 mg/dl, or if there is a hypoglycemic reaction between breakfast and lunch, decrease the prebreakfast short-acting or rapid-acting insulin by one or two units.
• If the predinner test is less than 90 mg/dl, or if there is a hypoglycemic reaction between lunch and dinner, decrease the morning basal by one or two units.
• If the bedtime test is less than 90 mg/dl, or if there is a hypoglycemic reaction between dinner and bedtime, decrease the predinner short-acting or rapid-acting insulin by one or two units.

Section 2

The following are algorithms for adjusting insulin using three shots of short-acting or rapid-acting insulin (one shot before every meal) with basal insulin at bedtime, and they are based on these four premises:¹

1. The bedtime basal has its major action during the night and on arising, and its effect is reflected in the blood glucose levels during the night and before breakfast.
2. The prebreakfast short-acting or rapid-acting insulin has its major action between breakfast and lunch, and its effects are reflected in the prelunch blood glucose level.
3. The prelunch short-acting or rapid-acting insulin has its major action between lunch and dinner, and its effects are reflected in the blood glucose level before dinner.
4. The predinner short-acting or rapid-acting insulin has its major action between dinner and bedtime, and its effects are reflected in the blood glucose results before bedtime.

Algorithm for hyperglycemia not explained by unusual diet or exercise, occurring at the same time for three to five days in a row:¹

• If the prebreakfast glucose is over 130 mg/dl (and if the blood glucose level at 3 AM is not less than 70 mg/dl), increase the bedtime basal by one or two units.
• If prelunch glucose is over 130 mg/dl, increase prebreakfast short-acting or rapid-acting insulin by one or two units.
• If predinner glucose is over 130 mg/dl, increase the prelunch short-acting or rapid-acting dose by one or two units.
• If bedtime glucose is over 130 mg/dl, increase predinner short-acting or rapid-acting insulin by one or two units.

Algorithm for hypoglycemia not explained by unusual diet or exercise:¹

• If prebreakfast glucose is under 90 mg/dl or there is evidence of hypoglycemia occurring during the night, reduce the bedtime basal by one or two units.
• If prelunch glucose is under 90 mg/dl, or hypoglycemia occurs between breakfast and lunch, reduce the morning short-acting or rapid-acting insulin by one or two units.
• If predinner glucose is under 90 mg/dl, or hypoglycemia occurs between lunch and dinner, reduce the prelunch short-acting or rapid-acting insulin by one or two units.
• If bedtime glucose is under 90 mg/dl, or hypoglycemia occurs between dinner and bedtime, reduce predinner short-acting or rapid-acting insulin by one or two units.

Conclusion

Algorithms provide general guidelines for adjusting insulin. To help your patients successfully control diabetes with insulin management, you should develop a detailed protocol for insulin adjustment and verify target goals, approved by your medical director or your institution.

Chapter Four
Adjusting Insulin: The Prospective Approach and Practical Applications

Chapter 1 explained the actions and types of insulin and established the ground rules for assessing a patient’s knowledge before beginning insulin adjustment principles. Chapter 2 described the new insulin analogs and various dosage regimens used for Type 1 and Type 2 patients. Chapter 3 provided basic guidelines for adjusting insulin using the retrospective approach. This chapter focuses on the prospective approach and explains how to change the insulin dose based on compensatory and anticipatory doses. It concludes with practical examples of the principles learned in this course.

There are basic principles of dosage adjustment, but patient responses can vary. If you are involved in adjusting insulin doses or in teaching patients how to adjust doses, you must develop a detailed protocol approved by your medical director or institution.

For the purpose of this chapter, you can assume the patient has had a thorough dietary assessment and is following a consistent diet.

Algorithms

Algorithms are guidelines that tell you how much insulin to give at each injection and are the basis for the patient’s self-adjustment of insulin. They allow the patient to make daily insulin choices based on blood glucose levels, food intake, exercise, and variations in normal routine such as emotional stress or illness.

Chapter 3 described the retrospective approach in which insulin doses are changed in response to the pattern or trend of blood glucose levels during the previous three to five days. Other guidelines included the “fix the fasting first” rule, which essentially means that starting with a normal blood glucose in the morning makes it easier to control glucose levels for the rest of the day. It has been noted that generally you should adjust only one insulin dose at a time and begin by solving the first problem you encounter in the day’s pattern of blood sugars. After you solve that, the rest of the day’s pattern may fall into range. Finally, the dose should not be changed by more than 10% at a time, and there should be a waiting period of three to five days to evaluate the effect.

Table 5

Prospective approach

In the prospective approach, the patient can make daily adjustments in the dose of short-acting or rapid-acting insulin based on anticipated food intake, exercise, and the prevailing blood glucose level. This type of algorithm addresses the dynamic nature of insulin requirements in relation to blood glucose and allows the patient to compensate for fluctuations above or below normal values. The dose of basal insulin still is determined by the previously explained retrospective method.

Table 6

The three components to the prospective approach of using short-acting or rapid-acting insulin are the basic dose, the compensatory dose, and the anticipatory dose.¹

Basic dose: The basic dose is the dose the patient takes routinely, that is, the dosage required to take care of normal food intake and usual activity. This dose is adjusted only according to the retrospective approach of looking at the trend of blood glucose levels for the previous three to five days. For example, if pre-lunch glucose levels were consistently above the target level and you knew no cause for the elevated glucose levels, the morning dose of short-acting or rapid-acting insulin would be raised by one or two units at a time, every three to five days, until normal glucose levels were achieved at lunch.

Compensatory dose. The compensatory dose is the “quick fix” dose based on the prevailing blood glucose just before the insulin injection. Table 5 shows a common regimen for using short-acting or rapid-acting insulin in this approach, in which the compensatory dose is added to the already established basic dose of short-acting or rapid-acting insulin.¹ This scale is a common starting point, and it is one that patients can remember. However, insulin effect is determined by individual patient need and the total amount of insulin a patient requires. Paul C. Davidson, MD, from Atlanta developed an “insulin sensitivity factor,” also called the “correction factor,” that is defined as the number of mg/dl the blood glucose level will drop over a two-to four-hour period after the administration of one unit of insulin. The sensitivity factor can be estimated by dividing the patient’s total daily insulin dose into 1,500.² Using this formula, the scale in Table 5 would have been derived from a total daily dose of 30 units of insulin. That is, 1,500 divided by 30 equals 50; therefore, one unit of short-acting or rapid-acting insulin will drop the blood glucose by 50 points. Because of the more rapid onset and peak of rapid-acting analogs, John Walsh suggests in his book Pumping Insulin that you might want to use a 1,800 rule rather than a 1,500 rule when calculating insulin sensitivity.² You would divide 1,800 by your total dose, and that answer would tell you how many mg/dl one unit of Humalog, NovoLog, or Apidra would drop the blood glucose. In other words, you might need less rapid-acting than Regular to lower a high blood glucose.

This principle explains why some patients are very sensitive to the addition of extra short-acting or rapid-acting insulin. For example, a small, thin woman with Type 2 diabetes taking 20 units of insulin daily and using Regular insulin would have a sensitivity factor of 75 (1,500 divided by 20 = 75). So in this situation, she would add one unit of short-acting insulin for a blood glucose in the range of 151-225, 2 units for 226-300, three units if over 300. Likewise, a person using a rapid-acting insulin analog and requiring 60 units of insulin daily would need one unit for every 30 points above the target level (1,800 divided by 60 = 30) and would follow the advice in Table 6. Furthermore, many patients discover they need more short-acting or rapid-acting insulin to bring down higher glucose levels since high blood glucose confers some insulin resistance; thus, an elevated blood glucose requires a larger dose of insulin to lower it. In this case, using Table 5, if the glucose level is 251 mg/dl to 300 mg/dl, the patient may need four to five extra units of short-acting or rapid-acting insulin, and if over 300 mg/dl, five to six units or more.

With the nighttime use of short-acting or rapid-acting, you must be careful in advising patients to take extra insulin at bedtime to treat an elevated glucose since there is an increased risk of nocturnal hypoglycemia as the short-acting or rapid-acting insulin peaks while the nighttime long-acting insulin is just beginning to act. However, after looking at the pattern of fasting glucose levels and having the patient check the blood glucose at 3 AM to ensure there has been no nocturnal hypoglycemia, you usually can determine when bedtime short-acting or rapid-acting insulin is necessary. One possible regimen is as follows:

• If bedtime glucose level is less than 150 mg/dl, have a snack.
• If bedtime glucose level is 150 to 200 mg/dl, omit the snack.
• If bedtime glucose level is 201 to 300 mg/dl, omit the snack and add one unit of short-acting or rapid-acting insulin.
• If bedtime blood glucose is over 300 mg/dl, omit snack and add two units of short-acting or rapid-acting insulin.

You may need to lower or raise these doses according to individual needs. Again, it is important to clarify with the physician the target goal for each patient since this target will dictate how you develop your scale. For example, if the target is 90 to 150 mg/dl, your scale will follow Table 5 or Table 6. If the target is 70 to 130 mg/dl, then Table 5 would list the blood glucose range starting below 70, then 70 to 130, 131 to 180, 181 to 230, 231 to 280, and 281 to 330.

You may wonder why it is necessary to subtract a unit of insulin if the blood glucose is below 90 mg/dl (see tables 5 and 6). Many patients will find this step unnecessary since if their blood glucose is below 70 mg/dl, they will ingest a fast-acting sugar food before the meal, and if their blood glucose level is 70 to 100 mg/dl, they may be fine with their usual meal and usual basic dose of insulin. But some patients’ blood glucose levels will continue to drop. For example, if the glucose level is 70 mg/dl and they take their usual dose of insulin, their blood glucose level will be 60 mg/dl by the next premeal test. Only by looking retrospectively at patients’ patterns can you determine when this is the case. For patient safety, it is generally recommended that you reduce the dosage if the blood glucose is lower than usual; if you later find this step unnecessary, you can eliminate that component of the scale.

As an alternative to using the previously mentioned tables (examples in tables 5 and 6), some patients prefer a more precise method, especially if they are using an insulin pump and can deliver insulin boluses in tenths of units. This method would use the following formula:³

Current glucose minus target glucose divided by correction factor = correction dose. For example: Assume the patient hasn’t eaten or taken insulin in four hours, the target glucose is 100, and the patient’s correction factor is 50. Using the above formula, if the current glucose were 280, then 280 minus 100 = 180, and 180 divided by 50 = 3.6 units. So the patient would take 3.6 units to lower the glucose to the target goal.

Take note when using the above method that patients must be cautioned to avoid overlapping their insulin. They must be instructed to take into account the type of insulin they use, its duration, and how long it has been since their previous injection and meal. Specific and individual rules for considering “insulin on board” must be given to each patient, which is beyond the scope of this course.⁴

A word of caution when using these tables: Too often, the patient is given a scale for short-acting or rapid-acting insulin without the proper explanation that this is only a temporary fix. Patients who use these tables without looking at the trend or pattern of blood glucose level over three to five days may not see the whole picture. A common problem is that when the morning fasting glucose level is too high, the patient adds more breakfast short-acting or rapid-acting insulin according to the scale and may continue to do so daily, even for months at a time until the next office visit. Instead, after three days of high fasting blood glucose levels (with no known cause such as increased food intake or illness), the patient should look at the pattern retrospectively and increase the evening dose of basal insulin to prevent the morning glucose from being too high. Moreover, although it takes at least three days to see a pattern, it is more common to look at the trend over a week’s time. Weekdays frequently differ from weekends in terms of food intake, activity, and stress. A patient may need one regimen of insulin for weekdays, a lower morning dose of basal/short-acting or rapid-acting on Saturday because of increased activity, and a higher dose for a less-active Sunday.

Note an important difference between using rapid-acting or short-acting insulin. Rapid-acting analogs have the advantage in their rapid action and shorter duration of more quickly and safely reducing a temporarily elevated blood glucose with less fear of delayed hypoglycemia previously caused by the prolonged action of Regular.⁵ As previously explained, the more Regular insulin you give, the longer it lasts, and 10 units could last six to 16 hours, depending on the patient. This phenomenon is commonly described as the “tail of Regular.” Frequently patients complain that it takes too long for Regular insulin to drop their blood glucose. When Regular insulin finally brings the blood glucose down, it’s “after the fact,” the meal content has been digested, and the patient develops hypoglycemia. In contrast, 10 units of rapid-acting analog would begin to affect the blood glucose in five to 15 minutes, with its most significant effect within one hour, thus reducing the chance of delayed hypoglycemia.

Anticipatory dose: The anticipatory dose is based on the amount of food (primarily carbohydrates) planned for the upcoming meal and the level of activity planned over the next four to six hours. A general rule is to start conservatively although the dose probably will not be sufficient. After you determine the individual patient response, you can titrate the dose accordingly.

Anticipatory dose for food

• Start by recommending adding 1 to 2 units of short-acting or rapid-acting insulin for increased intake of carbohydrate food. Although this seems too simple and general, it is an easy way to help the patient become comfortable changing the dose. Usually it is not necessary to raise the dose for increased fat or protein intake, and it should never be raised for alcohol. Alcohol cannot be converted to glucose; it inhibits gluconeogenesis and interferes with the counter-regulatory response to insulin-induced hypoglycemia. Through these mechanisms, alcohol may contribute to hypoglycemia.
• Many patients prefer to learn carbohydrate counting principles, which allow greater flexibility in eating with better control of blood glucose. Initially, carbohydrate counting is implemented as a stable meal plan in which a specific amount of carbohydrate is consumed at each meal and snack. The patient weighs and measures food to gain skill in establishing portion sizes. Then, the insulin-carbohydrate ratio can be identified, which is the ratio between grams of carbohydrate eaten and the number of units of short-acting or rapid-acting insulin required at each meal. For example, if 75 grams of carbohydrate are typically consumed at dinner and the usual dose of short-acting or rapid-acting insulin is five units, the patient’s dinner insulin-carbohydrate ratio is 75 divided by 5 equals 15, which means that one unit of short-acting or rapid-acting insulin covers approximately 15 grams of carbohydrate. The ratio may vary at each meal, but once established, it can be used to calculate the meal dose, based on the amount of carbohydrate. Commonly, patients require one unit of short-acting insulin for every 10 to 15 grams of carbohydrate.⁵ This generalization is not adequate to achieve optimal control since the previous ratio example was derived through trial and error and by reviewing food records and glucose monitoring results.
• This explanation of carbohydrate counting only partially describes the process.  Patients need an individualized consultation with a registered dietitian, along with carbohydrate counting tools and references. A suggested reference is the American Diabetes Association’s Complete Guide to Carb Counting by Hope S. Warshaw and Karmeen Kulkarni.
• Patients taking an intermediate or long-acting insulin or basal insulin in the morning should be careful when adding short-acting or rapid-acting insulin before lunch for increased food since both the morning basal insulin and the lunchtime short-acting or rapid-acting insulin could cause hypoglycemia by midafternoon. Advise patients to analyze their records carefully and add short-acting or rapid-acting insulin at lunch sparingly, only after their records show it is necessary.

Anticipatory dose for activity

It is common to advise patients to add a fruit, or a fruit plus starch and protein, before an anticipated increase in activity.⁵ This may be fine for an active child or a thin or normal weight adult, but undesirable for a person trying to lose weight through exercise, in which case eating before exercise is counterproductive. Furthermore, many patients complain of indigestion if they eat before strenuous activity. The following recommendation is based on the premise that the patient desires advice on how to change the insulin dose rather than on how to increase food intake.

n  If a patient plans to reduce activity, add one to two units of whichever insulin is peaking at the time of the inactivity.⁶ For example, raise the morning short-acting insulin if inactivity is expected in the morning, but raise the morning basal insulin (or lunch short-acting or rapid-acting if applicable) if inactivity is planned for the afternoon.

n  For increased activity, reduce the appropriate dose by at least one to two units to start, but more as needed depending on the level of activity.⁶ It is common to start by decreasing by 10% of the total dose. Sometimes both short-acting or rapid-acting and longer-acting insulin will need to be decreased for prolonged or heavy exercise, which can lower the blood glucose for up to 18 hours because of the increased uptake of glucose by the muscle. Strenuous exercise such as skiing or hiking may require a decrease of up to 50% of the intermediate-acting insulin and up to 100% of the short-acting or rapid-acting insulin.⁶

n  A special note about rapid-acting analogs: Before rapid-acting insulin was available, patients using Regular insulin (short-acting) were more likely to experience hypoglycemia several hours after exercise or before the next meal because of the prolonged effect of Regular insulin (in addition to the prolonged effect of exercise). In fact, patients often complained that in order to exercise before a meal, a snack was necessary to prevent hypoglycemia. Patients trying to lose weight did not appreciate this, as noted. In a study using Humalog, hypoglycemia was twice as likely during exercise one to 1½ hours after a meal, but 46% less likely when done three to 4½ hours after the meal.⁷ Since it is more feasible to exercise at least two to three hours after a meal to avoid interfering with digestion, rapid-acting insulin may provide some benefit for patients in this regard. Likewise, if the patient chooses to exercise soon after an injection of rapid-acting analog, the dose should be reduced, a snack should be added, or both.

Practical applications

The following examples of adjusting insulin are based on principles discussed throughout this course. In these examples, assume that the target blood glucose range is 90 to 130 mg/dl before meals, that the patient is following a reasonable diet, and that the blood glucose examples are the average pattern over the past three to five days.

Example 1

Insulin dose: 24 units of NPH insulin plus two units of Humalog insulin in the morning, 12 units of NPH insulin plus 2 units of Humalog insulin before dinner.

Blood glucose levels:
Breakfast  110 mg/dl
Lunch  200 mg/dl
Dinner  200 mg/dl
Bedtime  130 mg/dl

Solution:  Increase the morning Humalog insulin by one to two units only. Solve the problem of the lunch glucose level first (before fixing the dinner glucose level) because after you correct the lunch level, the dinner glucose level may fall into normal range. If it doesn’t, you can then increase the morning NPH insulin dose by one to two units.

Example 2

Insulin dose: 24 units of NPH insulin plus

4 units of Humalog insulin in the morning, 12 units of NPH insulin plus 4 units of Humalog insulin before dinner.

Blood glucose levels:

Breakfast  200 mg/dl
Lunch  150 mg/dl
Dinner  120 mg/dl
Bedtime  120 mg/dl
3 AM  130 mg/dl

Problem: When you tried raising the dinner dose of NPH insulin to 13, the 3 AM glucose level was 50 mg/dl.

Solution: Switch to three shots of insulin by “putting the NPH to bed” and giving only rapid-acting insulin before dinner; that is, try the following routine: 24 units of NPH plus 3 units of Humalog insulin in the morning, 5 units of Humalog insulin before dinner, and eight units of NPH insulin at bedtime.

There are several reasons for that solution:

• “Putting NPH to bed” will make it peak more efficiently before breakfast, rather than in the middle of the night.
• The more shots you give, the more effective the insulin, so generally the total dose should be decreased by approximately 10%.
• Since only Humalog insulin will be given before dinner without NPH, more Humalog insulin will be needed.
• Giving NPH insulin at bedtime will make it overlap somewhat in the morning with the morning Humalog insulin, so to be safe, decrease the morning Humalog insulin.

Example 3

Based on the following blood glucose levels what insulin dose would you advise a patient to take before dinner if the patient is planning to overeat at dinner and is new to experimenting with changing the dose? (Refer to Table 5 algorithm.)

Insulin dose: 6 units of Humalog insulin in the morning, 4 units of Humalog insulin before lunch, 8 units of Humalog insulin before dinner, and 12 units of glargine insulin at bedtime.

Blood glucose levels:
Breakfast  110 mg/dl
Lunch  110 mg/dl
Dinner  220 mg/dl

Solution: Add 2 units of Humalog insulin for the glucose level of 220 mg/dl, and add 2 units of Humalog insulin for the anticipated increased food intake; that is, add 4 units of Humalog insulin to the basic dinner dose of 8 units, which means 12 units of Humalog insulin should be given before this meal.

Also, if the dinner glucose levels remain high over the next three to five days, use the retrospective approach and raise the lunch Humalog insulin to prevent the dinner glucose level from being so high.

Example 4

A patient plans one hour of strenuous aerobics after dinner.

Insulin dose: 15 units of NPH insulin plus 5 units of NovoLog insulin in the morning and 10 units of NPH insulin plus 5 units of NovoLog insulin before dinner.

Solution:  Decrease the dinner dose to 8 units of NPH plus 3 units of NovoLog  insulin since the peak time of NovoLog  insulin and the time of onset of NPH insulin will overlap.  Also, the exercise effect may carry over several hours.

Conclusion

Algorithms provide guidelines for adjusting insulin and allow a patient to make daily insulin choices based on blood glucose levels, food intake, exercise, and variations in normal routine. The patient’s action depends on the answers to several questions he or she asks at the time of any premeal insulin injection.⁸

• What is my blood glucose now?
• What do I plan to eat now, i.e., a usual size meal or a large or small one, and how much carbohydrate?
• What do I plan to do after eating, i.e., usual, increased, or decreased activity?
• What has happened under these circumstances previously?

The principles presented in this series are only basic guidelines; patients vary in their needs and responses to insulin. To help your patients achieve diabetes control with insulin management, you should develop a detailed protocol for insulin adjustment approved by your medical director or institution.