[ Editorials | Departments and Series | Index ]

Diabetes Management Series

Fine-tuning Physiologic Insulin Replacement Therapy via Office Downloading of Home Blood Glucose Monitoring Data

Jeff Unger, MD

Eighteen million Americans have diabetes, and projections suggest that an additional 1.3 million people aged 20 years and older will be diagnosed with the disorder annually in this country.¹ Most of these older patients will have type 2 diabetes, but the United Kingdom Prospective Diabetes Study demonstrated that the progressive decline in pancreatic β-cell mass and subsequent reduction in endogenous insulin secretion results in the initiation of exogenous insulin injection therapy in nearly all patients who are originally diagnosed with type 2 diabetes.^2,3 Whereas the short-term goals of diabetes management include avoidance of hyperglycemia and hypoglycemia, physicians must also direct treatment toward surveillance for and avoidance of diabetes-related complications. Microvascular complications (eg, neuropathy, nephropathy, retinopathy) are likely to occur in patients with long-term exposure to hyperglycemia.⁴ The incidence of macrovascular complications such as stroke, myocardial infarction, angina, and peripheral vascular disease can be reduced by improving hemoglobin A1C, blood pressure, and lipid levels; implementing aspirin therapy; and encouraging lifestyle intervention such as smoking cessation, exercise initiation, and weight reduction.⁵

THE ROLE OF HEMOGLOBIN A1C

Glycated hemoglobin A1C refers to hemoglobin components that form slowly and nonenzymatically when hemoglobin is exposed to plasma glucose. The rate of formation of glycated hemoglobin is directly proportional to the ambient glucose concentration. Thus, higher glucose levels result in elevated A1C levels. The American Diabetes Association (ADA) has recommended that patients attain an A1C level below 7%, because the Diabetes Control and Complication Trial⁶ demonstrated that A1C levels above 7% are likely to increase the risk of developing microvascular disease. However, other professional societies such as the American Association of Clinical Endocrinologists have recommend that patients with diabetes attain A1C values below 6.5%.⁷ The rationale for aggressive A1C management in patients requiring exogenous insulin injections is that they should achieve an A1C value that is as close to normal as possible—ie, 4.9% to 5.1%. The lower the A1C value, the lower the risk of microvascular and macrovascular complications. Patients who desire pregnancy should attain the best possible A1C level prior to conception.⁸ Hospitalized patients with near-normal A1C levels also have better outcomes than those who present acutely with hyperglycemia to the intensive-care or cardiac-care settings.⁹

back to top

SELF-MANAGEMENT

Patients can be motivated to become active participants in diabetes self-management programs. All too often, patients' charts are filled with statements about noncompliance, refusal to follow dietary instructions, and poor weight control. In reality, diabetes is a very difficult disease to manage for both patients and physicians. Exogenous insulin is not always absorbed predictably, resulting in erratic blood glucose levels. Insulin absorption is based on the type of insulin, the injection site and volume, and the patient's size and age. In addition, if patients use another insulin injection before the prior one is completely absorbed, hypoglycemia will likely ensue due to a process known as "insulin stacking." Thus, both physicians and patients must understand and follow the rules of insulin pharmacokinetics when prescribing and adjusting physiologic insulin replacement therapy.

When patients are informed that they will require exogenous insulin injections, physicians should advise them that the goal is to achieve blood glucose levels similar to those of nondiabetic individuals. This process is known as physiologic insulin replacement therapy, and can be accomplished using either an insulin pump¹⁰ or a basal-bolus insulin^9,11 regimen. Either regimen requires frequent office visits to fine-tune the patient's insulin regimen. Once the patient has been instructed about factors such as timing the injection in relation to meals and exercise, carbohydrate counting, correction boluses for hyperglycemia, and prevention/management of hypoglycemia, the focus should be shifted toward fine-tuning. These office visits should address scheduling of blood glucose monitoring; identifying the presence, frequency, and timing of hypoglycemia; calculating the percentage of time that the patient attains a predetermined glucose range for optimal A1C control; and detecting preprandial and postprandial glucose trends that require adjustments in bolus and basal insulin dosing. These goals cannot be achieved by reviewing patients' hand-written logs, no matter how detailed they are, as some patients may complete their quarterly logs in the office waiting room.

Self-blood glucose monitoring systems were first marketed in 1978, heralding an era of unprecedented therapeutic innovation in diabetes care.¹² With the development of blood glucose meters, patients could begin to adjust their own insulin dosage to achieve near-physiologic control. In addition, patients with type 2 diabetes could modify carbohydrate intake and exercise to achieve better glycemic control.

However, despite the large number of meters on the market and their ease of use, some large teaching centers still report that 31% of their patients have A1C levels that exceed 9.5%.¹³ One possible way to improve patients' A1C levels is to teach primary care physicians the value of downloading software systems for home blood glucose readings. Almost all blood glucose meters have software systems that can analyze numerous test results, enabling physicians to track and modify glycemic trends. Some meter companies (eg, TheraSense) have a Web-based computer downloading program to allow immediate charting, faxing, and e-mailing of glucose results to physicians.¹⁴ Meters with excellent download software include the OneTouch Ultra, FreeStyle, Accu-Chek meters, Ascensia Breeze, and BD Logic.

Patients must make certain that the date and time on the meter is correct for accurate downloading.

Patients can purchase the software for meter downloading directly from the meter company or download it from the meter Web sites. The data can then be presented at the office at the time of the visit. However, most meter companies will install this software free of charge in physicians' offices and train the staff in data interpretation. Although almost any meter company will do this, it is advisable to find two or three that offer not only software support, but free meters and starter strips for patients as well.

Patients should be instructed to use the same meter for the majority of their blood glucose readings. For patients who use more than one glucose meter, some software programs allow the downloading of data from multiple sources—provided that all of the meters are from the same company that makes the software. There are also more expensive software programs that allow for collation of downloaded glucose data from multiple meters manufactured by different companies.

back to top

INTERPRETATION

Software programs differ in terms of statistical and graphic formats, but electronic logbooks should have certain basic features, including a listing of blood glucose levels by date and time; an average or mean blood glucose (MBG) level per day, as well as a preprandial and bedtime average; the frequency of blood glucose monitoring on any given day and within any given reporting period; the percentage of blood glucose readings that fall within a prescribed target range; and the standard deviations (SDs), which indicate the glycemic variability for specified reporting periods (ie, preprandial, bedtime, nocturnal, and fasting). This information can be formatted in a table, pie chart, histogram, and/or linear chart for a 7- to 90-day period. Patients are encouraged to bring their meters to the office at each visit. The data can be downloaded by the office staff in less than 1 minute. The Accu-Chek "Smart Printer" allows patients to download their own data as they enter the examination room for automatic printing and collection by the medical assistant. The patient's data can be stored on the computer's hard drive, as well as printed out for immediate inspection.

With practice, the physician can interpret the data, discuss the information with the patient, and make any necessary adjustments in medication doses within 5 minutes. Patients often enjoy reviewing these data with the physician. Meter downloads can also motivate patients with poor glycemic control to improve compliance. Unlike written logs, electronic logs will document whether patients are failing to check blood glucose levels before meals (Figure 1). This is critical, because guessing at the insulin dose is totally inappropriate; a patient would require a larger insulin bolus for a blood glucose level of 275 mg/dL than for a level of 65 mg/dL (Figure 1).

Figure is not available.

FIGURE 1. Blood glucose readings for Case #1. This patient is not checking blood glucose levels before each meal and at bedtime despite instructions. She is determining insulin doses by "guessing" at blood glucose levels, a process that is dangerous and not physiologic. The electronic log allows the physician to educate the patient about the importance of frequent blood glucose monitoring and correct insulin dosing. Bef = before; Aft = after; Meds = medications; Carb = carbohydrates; Gluc = glucose.

Physiologic insulin replacement therapy requires patients to test their blood glucose levels frequently. Insulin dosing is based on many factors, including the preprandial glucose level and the amount of carbohydrates in the planned meal. Other relevant factors are the timing of any prior insulin bolus and what, if any, exercise is planned while the next bolus is being absorbed. Thus, frequent testing is critical to ensure good glycemic control. Patients using multiple insulin injections (four per day) should be instructed to perform home blood glucose monitoring before each meal and at bedtime daily. Insulin-pump patients often check blood glucose levels 2 to 3 hours after eating so that a supplemental bolus can be administered to correct hyperglycemia.¹⁵ Patients who fail to check their blood glucose levels at the appropriate times will usually have a corresponding elevation in their A1C values.¹⁶

Patients often ask physicians about the wide range of glucose variability that produces morning levels of 102 mg/dL one day and 165 mg/dL on the next. More important than such individual numbers on an electronic logsheet is the percentage of the patient's blood glucose readings that fall within a given target range. If 49% of the patient's readings over 30 days are in the range of 70 to 180 mg/dL, the patient should have an A1C level of close to 7% (assuming that readings are taken at least three times per day).¹⁷ Thus, the patient should be informed that perfect readings are not necessary to attain A1C levels that will lower the overall risk for developing diabetes-related complications.

The SD of the blood glucose values (ie, the square root of the variance) indicates whether glycemic control is consistent, or undergoing wide swings indicative of suboptimal control. In most cases, twice the SD should ideally be less than the MBG value for that given period; for example, an SD of 30 would be ideal if the fasting MBG value is 80 mg/dL. However, if the SD is 45 and the fasting MBG value is 80 mg/dL, the patient is frequently hypoglycemic. In general, the wider the gap between the SD and the MBG value, the greater the degree of insulin resistance.18 Such patients are usually insulin-deficient, and require significantly higher doses of basal insulin to lower fasting blood glucose levels.

back to top

CASE HISTORIES

Case #1

The patient is a 38-year-old female teacher who has had type 1 diabetes for 16 years. She received an insulin pump to help improve her erratic blood glucose levels, which were occurring while she used multiple daily injections of insulin glargine and preprandial insulin lispro. The patient is very dedicated to diabetes self-management, attending community diabetes classes monthly, implementing carbohydrate counting, and checking her blood glucose levels up to six times daily. Her prepump A1C value was 6.8%, and her meter download data are shown in Figure 2.

Figure is not available.

FIGURE 2. Meter download data for Case #1. The physician should first evaluate the percentage of time the patient achieves blood glucose levels of 70-180 mg/dL. If 49% of the patient's home blood glucose readings are within this range, the predicted A1C level would be < 7%. If 42% of the downloaded values are within this range, the predicted A1C level would be 8.5%. If 45% of the downloaded values are 70-180 mg/dL, the predicted A1C level would be 8%. N= number of hypoglycemic events, blood glucose levels recorded < 60 mg /dL.

In this patient's meter download (pie chart), 47% of the numbers are on target, but she is also experiencing hypoglycemia 28% of the time. In addition, hypoglycemia is occurring before and after meals, as well as after going to bed (nocturnal hypoglycemia). The high frequency of red in the pie charts shows that this patient has hypoglycemic unawareness, and cannot feel the symptoms of low blood glucose levels until they fall below 40 mg/dL. In this case, doubling the SDs at night, before breakfast, after breakfast, before dinner, and at bedtime reveals totals that equal or surpass the MBG levels, suggesting that the patient experiences hypoglycemia frequently during these times.

The data in the trendgraph (Figure 3) show that the patient's overall SD is 80 while her MBG is 133, indicating that she experienced frequent hypoglycemia during this reporting period. Because this patient is using an insulin pump, consideration should be given to lowering her basal rate of insulin delivery. By increasing her basal glucose levels to between 140 and 180 mg/dL for 4 to 6 weeks, the patient may ultimately restore her hypoglycemic awareness.10 Eventually, the basal rate can be increased again to allow for tighter glycemic control. In addition, mealtime boluses should be reduced, and the extended boluses should be changed from 2 to 3 hours to better match carbohydrate absorption from the gut and prevent postprandial hypoglycemia. The patient should be asked to return for another meter download in 3 weeks to make further adjustments in her insulin regimen.


Case #2

The patient is a 29-year-old man who was diagnosed with diabetes 3 months prior to this office visit. On initial presentation, his fasting blood glucose level was 529 mg/dL and his A1C value was 12.4%. The patient had been symptomatic for 5 months prior to his initial office visit with classic diabetes symptoms, including polyuria, polydipsia, weight loss, blurred vision, and polyphagia. Physiologic insulin replacement therapy was prescribed immediately, consisting of bolus insulin aspart and basal insulin glargine. The preprandial boluses were determined by the patient's blood glucose readings, as well as his estimation of the quantity of carbohydrates planned for each meal. His meter download data is shown in Figure 4. This patient is experiencing excellent diabetes control, which is reflected in his 6.5% A1C value.

Figure is not available.

FIGURE 3. Trendgraph for Case #1. The patient is checking her blood glucose frequently—an average of 4.5 times daily. In addition, she has experienced 15 documented blood glucose levels below 60 mg/dL over an 8-day period, and 25% of her total readings are < 70 mg/dL. The green bar in the trendgraph represents the target of 70-170 mg/dL. The solid black line represents the patient's average blood glucose levels on any given day, while the striped lines show the SDs. Each red "X" marks the recorded hypoglycemic readings. BG = blood glucose; N = normal; MBG = mean blood glucose; Hypos = hypoglycemia; Symp = symptoms; Nr=number. [Note: The M(80) and M(120) references are European values used to identify patients in whom blood glucose levels are not adequately controlled. In the United States, these values are not used. Schlichtkrull J, Munck O, Jersild M. The M-value, an index of blood-sugar control in diabetics. Acta Med Scand. 1965;177:95-102.]

Figure is not available.

FIGURE 4. Trendgraph for Case #2. These data were taken over a 7-day period in a newly diagnosed patient with type 1 diabetes. The patient is checking his blood glucose frequently, between six and eight times daily. The SDs, when doubled, are always less than the MBG readings, which means that there is very little variation in his blood glucose levels. The green bar represents the target for the patient's blood glucose readings at 70-170 mg/dL. Each "X" indicates a blood glucose reading. The dark black line represents the MBG level, which runs right across the middle of the green bar. The striped black lines are the SD points, which also lie within the green bar. MBG = mean blood glucose; N = normal.

Case #3

This patient is a 75-year-old woman who has been using an insulin pump for 6 years. She has an A1C level of 5.9%, with a history of hypoglycemic unawareness. The patient lives alone, and she has required emergency treatment for severe hypoglycemia on several occasions. Her meter download data indicate frequent episodes of hypoglycemia (Figure 5). When asked why she was becoming hypoglycemic so often, she explained that she was using repeated corrective insulin boluses in an attempt to avoid hyperglycemia (ie, insulin stacking), which can result in severe hypoglycemia.

Figure 6 shows the absorption curve for any insulin analog (lispro or aspart). Thus, in this patient's case, if she were to administer a 10-U dose at 10:00 am to correct an elevated blood glucose level and then eat lunch at noon, she would have to realize that she still had 6 U of insulin in her system. However, the patient had never been informed about this absorption process. Instead, she would administer her usual insulin dose for meals, ignoring the correction boluses. Thus, she would be using successive boluses, becoming hypoglycemic on a daily basis. Once she understood insulin stacking and knew that she had to wait 6 hours between boluses, the hypoglycemic events were reduced by 90%.

Figure is not available.

FIGURE 5. Meter download data for Case #3. Over a 6-week period, this elderly patient has recorded 24 hypoglycemic events; her meter was programmed to record any blood glucose level < 67 mg/dL as hypoglycemia. The average blood glucose reading over this period is 126 mg/dL, and the SD is 62. Doubling the SD yields a value that is nearly identical to the average blood glucose level—an indication that the patient is frequently hypoglycemic. Adapted from Mudaliar SR, Lindberg FA, Joyce M, et al. Insulin aspart: a fast-acting analog of human insulin: absorption kinetics and action profile compared with regular human insulin in healthy nondiabetic subjects. Diabetes Care. 1999;22(9):1501-1506. Hypo = hypoglycemia.

Figure is not available.

FIGURE 6. Absorption curve for any insulin analog (lispro or aspart). An insulin bolus will take at least 6 hours to absorb completely. At 2 hours, approximately 60% of the initial insulin dose remains to be absorbed; at 3 hours postdose, 40% of the initial dose remains to be absorbed. Adapted from Mudaliar SR, Lindberg FA, Joyce M, et al. Insulin aspart: a fast-acting analog of human insulin: absorption kinetics and action profile compared with regular human insulin in healthy nondiabetic subjects. Diabetes Care.1999;22(9):1501-1506.

Case #4

The patient is a 32-year-old male corrections officer with a 12-year history of poorly controlled type 1 diabetes. He presented on a combination of oral agents and twice-daily, split-dose insulin therapy, but had an A1C value of 13%. The patient also had neuropathy, retinopathy, hypertension, hyperlipidemia, and albuminuria. He reported that his father had died of diabetes-related cardiovascular complications at age 42 years. Basal bolus insulin was prescribed, consisting of insulin glargine at bedtime and insulin lispro prior to each meal. He works irregular shifts, including some night shifts weekly, but was instructed to check his blood glucose levels prior to each meal and at bedtime on a daily basis.

Figure 7 shows that 85% of this patient's readings are above the target range of 70 to 180 mg/dL, predicting an A1C level of more than 8.5%. The SD data indicate persistent insulin resistance throughout the day (Figure 8). The trendgraph in Figure 9 confirms that the patient has persistent hyperglycemia. As there is no evidence of hypoglycemia, higher doses of insulin can be safely initiated.

Figure is not available.

FIGURE 7. Target data for Case #4 shows that 85% of this patient's readings are above the target range of 70-180 mg/dL, which predicts an A1C level > 8.5%. N = normal; Hypo = hypoglycemia.

Figure is not available.

FIGURE 8. The SD data for Case #4. Optimal interpretation of SDs requires at least 20 readings. This patient works varying shifts, so pooling the nighttime readings with the before-breakfast and after-breakfast readings clearly shows severe insulin resistance. There is a wide gap between the doubled SD value and the MBG, indicating that the insulin resistance persists throughout the day. BG = blood glucose; N = normal; MBG = mean blood glucose.

Figure is not available.

FIGURE 9. Trendgraph for Case #4. This shows that the patient's blood glucose levels are always high, so that his current insulin dosage requires drastic adjustments. The data points indicate absolutely no evidence of hypoglycemia, so that higher doses of insulin can be used safely. BG = blood glucose.

back to top

IMPROVING PHYSIOLOGIC INSULIN REPLACEMENT THERAPY

When attempting to improve glycemic control, the ADA recommends targeting the fasting plasma glucose levels initially. When fasting blood glucose levels are at target but A1C levels are elevated, the postprandial levels should be improved.¹⁹ Monnier et al²⁰ determined the contributions of postprandial and fasting hyperglycemia to a patient's overall total A1C levels. If the A1C is < 8.5%, one should focus on lowering postprandial glucose levels to target overall A1C improvement. However, if the A1C is >10%, attempting to improve fasting blood glucose levels initially should result in a lower A1C when re-checked in 3 months.²⁰

back to top

CONCLUSION

Managing patients with diabetes can be very challenging. Physicians must focus on treating patients to target. In addition to achieving A1C levels below 7%, fasting blood glucose levels should be less than 110 mg/dL, and 2-hour postprandial levels should be less than 140 mg/dL.⁷ Physiologic glycemic control cannot be attained without frequent self blood glucose monitoring. Patients must be able to interpret their immediate readings and adjust their insulin dosages appropriately. However, long-term decisions about insulin dosage are the physician's responsibility. Using electronic meter download data provides a reliable, inexpensive means for fine-tuning physiologic insulin replacement therapy in the physician's office. Patients look forward to downloading their data because they too desire the best possible outcome from their diabetes management. Meter data provides invaluable reinforcement. For example, a patient with a download that shows only 42% of readings in the target range would have an A1C value of 8.5%. The patient's insulin regimen would be adjusted, and she would return in 6 weeks. On follow-up, the download should show that 53% of the readings are now in the target range of 70 to 180 mg/dL, which would correspond to a predicted A1C value of 7%. An in-office A1C test can also be performed for immediate confirmation. The patient is proud of her accomplishment, which promotes ongoing frequent blood glucose monitoring. With additional practice, physicians can learn to adjust mealtime boluses, detect missed or delayed insulin injections, and even identify patients who are less than honest when performing home blood glucose monitoring.

back to top

Jeff Unger, MD, is assistant professor of family medicine, Loma Linda University School of Medicine, California; and director, Chino Medical Group Diabetes and Headache Intervention Centers, California.

References

1. Diabetes Association Web site. National diabetes fact sheet. Available at: http://www.diabetes.org/diabetes-statistics/ national-diabetes-fact-sheet.jsp. Accessed November 18, 2004.
2. Tubner RC, Cull CA, Holman RR. U.K. Prospective Diabetes Study. Diabetes Care. 1996;19(2):182-183.
3. United Kingdom Prospective Diabetes Study 24: a 6-year, randomized, controlled trial comparing sulfonylurea, insulin, and metformin therapy in patients with newly diagnosed type 2 diabetes that could not be controlled with diet therapy. United Kingdom Prospective Diabetes Study Group. Ann Intern Med. 1998;128(3):165-175.
4. Unger J. Targeting glycemic control. The Female Patient. 2003;28(9):12-16.
5. Adler AI, Stevens RJ, Neil A, Stratton IM, Boulton AJ, Holman RR. UKPDS 59: hyperglycemia and other potentially modifiable risk factors for peripheral vascular disease in type 2 diabetes. Diabetes Care. 2002;25(5): 894-899.
6. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. The Diabetes Control and Complications Trial Research Group. N Eng J Med. 1993; 329(14): 977-986.
7. American Association of Clinical Endocrinologists. Medical guidelines for the management of diabetes mellitus: the AACE system of intensive diabetes self-management--2002 update. Endocr Pract. 2002; 8(Suppl 1):40-82. Available at: http:// www.AACE. com/clin/guidelines/diabetes_2002.pdf. Accessed November 24, 2004.
8. Unger J. Preconception management of women with type 1 diabetes. The Female Patient. 2001;26(5):40-46.
9. Unger J, Marcus AO. Glucose control in the hospitalized patient. Emerg Med. 2004;36(9):12-18.
10. Unger J, Marcus A. Insulin pump therapy: what you need to know. Emerg Med. 2002;34(9):24-33.
11. Unger J. Intensive management of type 1 diabetes. Emerg Med. 2001;33(9):30-42.
12. Walford S, Gale EA, Allison SP, Tattersall RB. Self-monitoring of blood-glucose: improvement of diabetic control. Lancet. 1978;1(8067):732-735.
13. Suwattee P, Lynch JC, Pendergrass ML. Quality of care for diabetic patients in a large urban public hospital. Diabetes Care. 2003;26(3):563-568.
14. TheraSense Web site. FreeStyle Tracker Diabetes Management System. Available at: http:// www.therasense.com/tracker/index.asp. Accessed November 24, 2004.
15. Unger JR, Fredrickson LP. A primer on intensive diabetes management and insulin pump therapy. Primary Care Reports. 1997;3:9.
16. Evans JM, Newton RW, MacDonald TM, Stevenson RJ, Morris AD. Frequency of blood glucose monitoring in relation to glycaemic control: observational study with diabetes database. BMJ. 1999;319(7202):83-86.
17. Brewer KW, Chase HP, Owen S, Garg SK. Slicing the pie. Correlating HbA values with average blood glucose values in a pie chart form. Diabetes Care. 1998;21(2):209-212.
18. Hirsch IB. Blood glucose monitoring technology: translating data Into practice. Endocr Pract. 2004;10(1):67-76.
19. American Diabetes Association: Standards of medical care in diabetes (position statement). Diabetes Care. 2004;27(Suppl 1):S15-S35.
20. Monnier L, Lapinski H, Colette C. Contributions of fasting and postprandial plasma glucose increments to the overall diurnal hyper-glycemia of type 2 diabetic patients: variations with increasing levels of HbA(1c). Diabetes Care. 2003;26(3):881-885.

back to top