Blood Glucose Monitoring: a Practical Guide for Use in the Office and Clinic Setting

Home , Blood glucose monitoring

In Brief Interpretation of self-monitoring of blood glucose (SMBG) results is not routinely practiced in the office and clinic setting. However, such interpretation in the presence of patients may facilitate improved patient-provider discussion, clinical decisions, and ability to manage glycemic patterns. This review outlines key steps in a systematic review of SMBG data, including 1) identifying the degree of blood glucose control using mean and standard deviation or variance, 2) identifying patient safety concerns with regard to hypoglycemia, and 3) understanding the factors influencing blood glucose control by noting trends and patterns. Case studies are provided to illustrate the process for using and interpreting electronic SMBG downloads. Blood Glucose Monitoring: A Practical Guide for Use in the Office and Clinic Setting

In 1978, the introduction of home In an uncontrolled study conducted blood glucose meters for self-moni- at two community pharmacies in Vir- Peggy Soule Odegard, PharmD, toring of blood glucose (SMBG) pro- ginia, blood glucose values decreased vided a new window through which significantly at 12 months compared BCPS, CDE, and Jennifer Rose people with diabetes could observe to baseline (P < 0.05), while medica- Beach, PharmD, CDE and respond to the daily influences tion adherence was maintained at on blood glucose affecting their dia- 90% for those receiving pharmacist betes.1,2 This new vantage point on review and consultation about their blood glucose also provided health SMBG results.5 care providers with a novel and effec- Some third-party insurers offer free tive way to provide counseling, target blood glucose meters through their therapy, and empower people with drug benefit. However, personalized diabetes to have more control over training and instruction of patients or management of their diabetes. Since regular monitoring of results does not that time, the technology of blood appear to be linked to these programs. glucose monitoring has developed Instead, clients are referred to the meter significantly to facilitate convenience use instructions available through the of meter use and interpretation of blood glucose meter manufacturer’s results.3 Specific developments that websites.6,7 The consensus statement have driven the market include developed by people with diabetes at substantial reduction in meter size, the seventh annual “Taking Control speed of testing (5 seconds is com- of Your Diabetes” conference indicates mon now), automated controls and that patients desire to have increased strip coding, minimal need for blood participation in decisions and under- (as low as 0.3 µl), enhanced memory, standing of test results. Systematic and the ability to download results. review and discussion with patients Table 1 displays the key features of about SMBG results makes this desire several popular meters. a possibility.8 Although technology has ad- Whether in a clinic, office, or vanced to facilitate increased use of pharmacy setting, key challenges home blood glucose meters, published to systematic and ongoing SMBG reports describing systematic review consultation and review include poor and interpretation of this information provider familiarity with meters by health care providers are limited. because of a lack of training and A major diabetes trial conducted in workload limitations precluding community pharmacies that involved more extensive consultation with free distribution of blood glucose patients. These limitations in the monitors, education on use, and translation of meter technology into ongoing monitoring and consultation daily practice may negatively affect with pharmacists on results, demon- initial meter education and follow-up strated improved diabetes control.4 monitoring of results. In a clinic or 100 Diabetes Spectrum Volume 21, Number 2, 2008 Table 1. Selected Downloadable Blood Glucose Meters and Their Features 10 From Research to Practice / Diabetes Technology Update Name Coding Blood Testing Time Test Range Hematocrit Test Memory Special Volume (µl) (seconds) (mg/dl) Limits Features and (%) Notes Accu-Chek No coding 1.5 5 10–600 25–65 300 glucose Strips are Compact required readings with housed inside Plus* time and meter. Do date; 7-, 14-, not use if on and 30-day peritoneal averages dialysis. Accu-Chek Snap-in code 4 26 10–600 20–65 480 glucose Strips are Advantage* key readings with curved for time and ease of use. date; 7-, 14-, and 30-day averages Accu-Chek Snap-in code 0.6 5 10–600 Not reported 500 glucose Up to four Aviva* key in package readings with alarms can insert time and be set in the date; 7-, 14-, meter. Do and 30-day not use if on averages peritoneal dialysis. Ascencia Automatic 0.6 5 10–600 0–70 in 480 glucose Optional Contour (new new meters, readings with meal mark- 2007 ver- 20–60 in old time and ers with sion)† meter version date; 14-day postmeal average reminder alarms Ascencia Automatic 1 5 20–600 20–55 420 glucose Strips housed Breeze2† readingswith in meter. May time and be better for date; 1-, 7-, arthritis suf- and 30-day ferers. averages OneTouch Manual 1 5 20–600 30–55 150 glucose Ultra‡ readings with time and — date; 14- and 30-day averages OneTouch Manual 1 5 20–600 30–55 500 glucose Meal markers Ultra2‡ readings with and other time and comments date; 7-, 14-, available. and 30-day averages OneTouch Manual 1 5 20–600 30–55 500 glucose UltraMini readings with — (new version)‡ time and date OneTouch Manual 1 5 20–600 30–55 More than Electronic UltraSmart‡ 3,000 records logbook op- with 7-, 14-, tions includ- 30-, 60-, and ing insulin, 90-day aver- exercise, ages health data, medication, and food continued on p. 102 Diabetes Spectrum Volume 21, Number 2, 2008 101 Table 1. Selected Downloadable Blood Glucose Meters and Their Features ,10 continued Name Coding Blood Testing Time Test Range Hematocrit Test Memory Special Volume (µl) (seconds) (mg/dl) Limits Features and (%) Notes Precision Xtra Calibrator in 0.6 5 20–500 30–60 450 glucose Tests blood (new version)§ each test strip readings with ketones box time and (strips use 1.5 date; 7-, 14-, µl; 10-second and 30-day results) averages Therasense Manual 0.3 7–15 20–500 0–60 250 glucose Freestyle§ readings with time and — date; 14-day average Therasense Manual 0.3 7 20–500 0–60 250 glucose Up to four Freestyle readings with daily alarms Flash§ time and can be set date; 14-day in meter. Do average not use if on peritoneal dialysis. Therasense Manual 0.3 5 20–500 0–60 250 glucose Up to four Freestyle Free- readings with daily alarms dom§ time and can be set date; 14-day in meter. Do average not use if on peritoneal dialysis. Therasense Automatic 0.3 5 20–500 15–65 400 glucose Up to four Freestyle Lite§ readings with daily alarms time and can be set date; 7-, 14-, in meter. Do and 30-day not use if on averages peritoneal dialysis. Truetrack|| Code chip 1 10 20–600 30–55 365 glucose Value meter readings with option 14- and 30- day averages Prodigy Automatic 0.6 6 20–600 20–60 450 glucose Talking Autocode¶ readings with meter; value time and meter option date; 7-, 14-, 21-, 28-, 60-, and 90-day averages *Roche (1-800-858-8072, www.accu-chek.com) †Bayer (1-800-248-8100, www.BayerDiabetes.com) ‡Lifescan (1-800-227-8862, www.lifescan.com) §Therasense (1-888-522-5226, www.AbbottDiabetesCare.com) ||Home Diagnostics (1-800-803-6025, www.homediagnostics.com) ¶Diagnostic Devices, Inc. (1-800-243-2636, www.prodigymeter.com)

provider office setting, interpreta- function or downloading results to providers in rapid interpretation of tion of meter data is often limited evaluate the mean blood glucose and SMBG information in the clinic and to either reviewing patient logbooks standard deviation (SD). office setting. In this review, we will or records with infrequent use of the Software technology that is read- discuss, through the presentation of 7- or 14-day blood glucose averaging ily available and easy to use can assist two case studies, these technologies

102 Diabetes Spectrum Volume 21, Number 2, 2008 method of record keeping in diabetes. From Research to Practice / Diabetes Technology Update , continued However, interpretation of SMBG based on handwritten logbook records is dependent on patients’ record-keeping intensity and organization, and these records can vary dramatically in terms of helpfulness in making management decisions. Some patients who bring in a handwritten logbook are very careful to note the time of day, how much insulin or medication was given, what was eaten, and the corre- sponding blood glucose result. Other patients provide a collection of results, in random order, without indicating the date, time, or reason for taking the blood glucose measurement. When relying on handwritten SMBG results, it is essential that results are organized, timed, and provide a sense of the influence of Figure 1. Sample SMBG logbook. medications and lifestyle on blood glucose to allow for meaningful and suggest a step-by-step method for use of meter memory histories and diabetes care decision making. If log- same-visit interpretation of SMBG computation, and electronic meter books are lacking sufficient informa- results in the clinic or office setting. downloads with external software tion to provide feedback on diabetes analysis. Regardless of method, there self-management, patients should Case Study 1 are five standard goals for a system- be specifically instructed to increase B.T. is an 80-year-old woman with atic review: SMBG, including the time of day to type 2 diabetes, who has come to the 1. Estimate the degree of blood test and timing with regard to meals clinic for a checkup of her diabetes. glucose control and variation or medications. Follow-up should She also has congestive heart failure, throughout the day. then be arranged to assess the SMBG hypertension, and hyperlipidemia. She 2. Identify patient safety concerns information after 1–2 weeks. If there is confined to a wheelchair and there- with regard to hypoglycemic are concerns about the reliability of fore gets very little physical activity. trends or events. results, a hemoglobin A1c (A1C) test Her most recent health crisis includes 3. Understand the factors influenc- or random finger-stick check can be hospitalization for pneumonia. Dur- ing blood glucose control. performed in the office or clinic to ing that time, her blood glucose 4. Suggest strategies for achieving validate the information. levels were consistently 300–400 improved blood glucose control. Important considerations when mg/dl. Before this illness, her blood 5. Provide reinforcement to patients reviewing handwritten SMBG log- glucose levels were more or less under that this information is valuable books are to identify hypoglycemic control. She had been increasing her and useful in their care. events (safety concerns) and glucose insulin doses to cover the higher patterns (efficacy of therapy).9 Table blood glucose levels related to her In general, the average or mean 2 reviews the steps for manual illness, but she is now complaining of blood glucose (using at least a 3- interpretation of SMBG results. Log- symptoms of hypoglycemia. day history) can provide an idea of books are commonly organized with Her usual home regimen is 10 units recent diabetes control and allow for columns for time of day and rows for of glargine insulin twice daily, 10 units counseling on factors that may have day of the week (Figure 1), allowing of lispro before breakfast, 8 units of affected blood glucose, including diet, the eye to travel down columns to lispro before lunch, and 10 units of physical activity, and medications. identify trends at certain times of day lispro before dinner. She does not have and across rows for trends occur- a logbook or written record of blood Patient Records and Logbook ring certain days of the week. After glucose results. However, she says that Reviews reviewing a patient’s logbook, wor- she checks her blood glucose a few Ideally, interpretation of SMBG sho- risome hypoglycemic events should times each day. How might a record uld be based on a systematic review be immediately discussed with the of her SMBG assist in her care today if or download of meter results because patient and a plan should be created you could review it quickly and easily this provides the most objective me- to avoid future events. Trends in in the exam room with her? thod for reviewing information and glucose values also should be dis- providing feedback to patients. In cussed to provide encouragement for Methods for Interpreting SMBG reality, however, many patients bring positive trends and identify strategies Results only their handwritten records or for reducing negative trends (e.g., There are several methods for evaluat- logbook to office visits. weekend high readings because of ing SMBG results, including manual Logbooks (Figure 1) are distributed eating at restaurants). The final step review of patient diaries or logbooks, with each new meter and are a common is to provide insight to the patient Diabetes Spectrum Volume 21, Number 2, 2008 103 Table 2. Interpretation of SMBG Results Step Action Manual Review of Electronic Meter Logbook Download 1 Connect download cable to meter and cue computer to 3 begin downloading at least 1–2 weeks of data. Select desired report and print. 2 Review at least 1–2 weeks of SMBG results. 3 3 3 Identify and circle worrisome hypoglycemic events (e.g., 3 3 SMBG < 70 mg/dl). Ask whether these required the assistance of another individual (i.e., “serious” events) or were manageable by the patient with carbohydrate treatment. Also ask whether they can be explained (e.g., missed a meal or over-treatment of hyperglycemia) or were unexpected. 4 Review consistency and timing of SMBG. 3 3 5 Assess the overall (aggregate) SMBG average and vari- 3 Variability must be 3 Variability is ability throughout the day. visually observed as a reported as an SD. • Compare average seen to previous review and to A1C range. values. • The goal for an acceptable SD is less than half of the average, in the absence of predominant hypoglycemia. • Time frames with a higher SD should be evaluated for reasons of variability. 6 Assess the time-specific SMBG averages. 3 This would have 3 • Comparing each time frame average can help identify to be done manually the lowest and highest average, which may help to or by quick visual determine problem areas. glance down logbook • Verify patient’s schedule as part of evaluation because columns. shift workers or others with abnormal schedules may not fit the meter’s or book’s preset times. 7 Evaluate trends in basal and prandial control. Note posi- 3 3 tive and negative trends. • Basal trends: Evaluate fasting blood glucose by comparing bedtime to morning blood glucose values. • Prandial trends: Evaluate postmeal blood glucose; establish patient’s eating time and routines. • Other trends: Review blood glucose records for evidence of insulin stacking, problems with correction doses, or over-treatment of hypoglycemia. • Consider adjusting insulin doses by 10–20%. 8 Discuss your assessment with patients to get their input 3 Difficult to perform 3 Automated report- on trends and to improve efficacy and safety of therapy. full review of trends ing facilitates the • Set realistic testing goals for the next visit. at one visit given the ability to have this • Changes that are made may require special testing sce- need for calculation discussion within narios (such as recommending a few 3:00 a.m. blood and manual data minutes of a down- glucose measurements to follow up on a basal insulin analysis. load. Practitioners adjustment). Discuss these expectations. Decide on can review current follow-up plans. average tests per day on some download programs to help set future goals.

on the overall glucose average and Case Study 1 Update maintaining her logbook has been degree of variability and to roughly B.T. reveals that she cannot write any- difficult. She is cared for by her son equate these to the expected A1C more because she has severe tremors, and daughter-in-law, who are busy (Table 3). especially on her right side, which is during the day and unable to help her her dominant hand. She confides that with this task.

104 Diabetes Spectrum Volume 21, Number 2, 2008 Table 3. A1C Correlation to hypoglycemia symptoms. Without between food absorption and insulin From Research to Practice / Diabetes Technology Update the information mapped out more action onset. Average Plasma Glucose succinctly, it is difficult to interpret Average Plasma A1C trends and patterns. Case Study 1 Update Glucose (%) B.T.’s. meter is downloaded, and her (mg/dl) Meter Downloads and Analysis aggregate mean for the past month Downloading SMBG results to an is 245 mg/dl, reflecting her earlier 135 6 office or clinic computer, secured hospital admission and illness-related 170 7 website, or even a patient’s computer hyperglycemia. She has an SD of 131 for analysis is a fast, simple, and mg/dl, which is higher than desired. 205 8 effective method for evaluating glu- This could reflect that, although she is often hyperglycemic, she is having 240 9 cose information. Major advantages of electronic SMBG download and episodes of hypoglycemia, contribut- 275 10 analysis include speed of review; ing to a large degree of variability. In her case, it could also be the result 310 11 instant and accurate calculation of results (versus estimation often used of the change in her usual level of 345 12 when looking at logbooks); ability control, with better blood glucose to report data in many ways using levels before pneumonia and higher Meter Memories and Computation levels afterward. Many patients rely on their meter’s tables, graphs, charts, or statistics; memory to act as the record rather and ability to show these reports immediately to patients for use in Exploring Problems With Basal than take the extra step to write Glucose Control planning care. down their results in a logbook. Evaluating fasting blood glucose Both specialty software and These patients find entering the large is the best way to diagnose needs native software offered by the meter amounts of data they generate into a for improved basal control. For the company are readily available for logbook to be tedious and unneces- majority of individuals, the first downloading. Several programs are sary.10 With the growing prevalence morning blood glucose (premeal) is a able to interface with the popularly of diabetes and the increasing use of good proxy for fasting blood glucose used meters (Table 4). One example computers by patients and health care (8 hours since last meal). For those of specialty software is Clinipro,11 providers, the use of meter technology taking insulin, preprandial glucose and downloading should become the which can download most commer- monitoring provides information standard in the future. As an interim cially available meters into one soft- regarding the efficacy of basal insulin step, the use of meter memories can ware program. After using software doses (subcutaneous injection or add to a patient’s evaluation and can to download meter readings, a pro- pump basal infusion rate) and can help validate handwritten results. cess similar to that described above also provide insight into the effective- The meter memory can be for manual interpretation of logbook ness of previously injected prandial accessed by turning on the meter results can be used to interpret the coverage.12 For those taking oral and scrolling through the available data (Table 2). medications, such as sulfonylureas results manually. The number of The overall mean and SD are key or metformin, fasting blood glucose readings stored varies based on factors to evaluate when using soft- can provide a sense of basal control meter type (Table 1), and results are ware analysis. The aggregate mean with regard to influence on insulin usually listed in chronological order can be a good quick comparison production (sulfonylureas) or inhibi- from most to least recent. It can help from the last visit to the current visit tion of nighttime glucose output to write down values from the past in terms of diabetes control, assum- (metformin). 1–2 weeks and then evaluate them ing the amount of hypoglycemia is Elevated morning blood glucose based on time of day and result. approximately equal. The average levels can also be the result of night- Many meters may have the capacity can be used to compare to recent time snacking that presents as morn- to display an average, but this, again, A1C values so that discrepancies can ing hyperglycemia, so it is important depends on meter brand. The most be identified that may need further to explore reasons for high readings common meters will give 7-day, 14- investigation (Table 3). before making a change. Other night- day, or 28-day averages. Review of The SD, which is the square root time routines, such as exercise, medi- the meter memory is also helpful to of the variance, indicates the amount cation dosing, and sleep patterns, are verify handwritten logbook results, of variability among all the meter also relevant topics to review with clarify testing times (if the date and readings. It has been suggested that patients and can lend evidence or time have been set properly), and cal- the SD value doubled should be doubt to a given presumption about culate SMBG averages, if the meter less than the aggregate mean as a the patients’ blood glucose patterns. is capable. way to assess safety and efficacy of The blood glucose trend overnight treatment regimen.10 Certainly, the is especially important, and bedtime Case Study 1 Update higher the SD, the more variability blood glucose results can help to B.T. has brought in her meter, and a patient is experiencing, with many evaluate this trend. For example, if scrolling through her values reveals possible causes, such as erratic insulin a patient goes to sleep with a blood higher-than-normal blood glucose or medication absorption, improper glucose level of 120 mg/dl (> 3–4 levels. She is having hypoglycemia timing of meals with insulin onset, hours after dinner) and has no bed- as well, confirming her reports of or gastroparesis creating a mismatch time snack, a blood glucose at or just Diabetes Spectrum Volume 21, Number 2, 2008 105 Table 4. Selected Diabetes Management Software Programs10 Name Meters Supported Special Features System Costs (Manufacturer) Requirements Accu-Chek Compass Accu-Chek Logbook, trend Pentium 90 or Accu-Chek Compass (Roche) Advantage, Com- graph, average day, higher, Windows 98, software = $29.99; pact, Compact Plus, average week, target Windows NT 4.0, cable for Aviva/Com- Active, Complete, range, hypoglycemia Windows ME, Win- pact/Active = $15; Aviva measurement, 1- dows 2000, or Win- cables for Advantage page summary dows XP Home or or Complete = $10 XP Professional, 32 MB of RAM, 50 MB of available hard disk space, graphics card that supports a resolution of 800 × 600 pixels and 256 colors, serial/COM port, MS Internet Explorer 5.0 or higher OneTouch Diabetes All OneTouch me- Logbook, 14-day Windows-compat- Software kit includ- Management Soft- ters with data ports, summary, glucose ible computer with ing cable and CD = ware version 2.3.1 including Ultra, trend, pie chart, av- 266 MHz, Intel Pen- $29.99, suggested (Lifescan) Ultra2, New Ultra- erage reading, stan- tium 2 or equivalent, retail. Visit website Mini, UltraSmart dard day, data list, minimum 128 MB www.OneTouchDia- histogram, insulin of RAM, minimum betes Software.com log, health checks free hard disk space for full details on 100–200 MB during where to purchase installation/100 MB and for download of after installation, drivers for UltraMini Microsoft Windows download 98 Second Edition, Windows NT 4.0 Workstation (SP6 or above), Windows 2000 Professional (SP4 or above), Windows ME, and Windows XP (SP2 or above) Home and Professional (SP2 or above), 9-pin/25-pin COM or USB port, video monitor at least 800 × 600 pixels and 256 colors, CD-ROM drive TrackRecord Data TrueTrack, Summary, logbook, Microsoft Windows Download latest Management Soft- TrueRead, Prestige conformance, XP Professional with version for free ware IQ glucose trend, and SP2 or Vista, Adobe at website www. (Home Diagnostics) extended logbook Acrobat Reader, thesmartchoice.com Intel Pentium III or or call 1-800-342- higher, USB connec- 7226 for free CD; tivity cable $19.99 continued on p. 107

106 Diabetes Spectrum Volume 21, Number 2, 2008 Table 4. Selected Diabetes Management Software Programs10, continued From Research to Practice / Diabetes Technology Update Name Meters Supported Special Features System Costs (Manufacturer) Requirements CoPilot Health FreeStyle, Freestyle Diary list, logbook, PC with 400 mega- Download free from Management System Flash, FreeStyle glucose modal day, hertz (MHz) or website www.abbott v3.1 (Abbott) Freedom, Preci- lab and exam, glu- more, Internet diabetescare.com; sion Xtra, Freestyle cose line, statistics, connection or CD- find details there Tracker, Cozmore glucose average, ROM drive, RAM about getting cables daily combination, of 64 MB or more, (purchased separate- weekly pump, glu- Microsoft Windows ly, ~ $29.95). cose pie graph; can 98 SE, 2000 or XP print, fax, or e-mail operating systems, reports monitor with 1024 × 768 or higher resolution, serial port, available 9-pin EIA- 232 or appropriate adapter for a USB for glucose meter connection Precision Link Di- FreeStyle, FreeStyle Data points, target PC with 400 mega- Download free from rect Diabetes Data Flash, FreeStyle ranges, standard hertz (MHz) or website www.abbott Management System Freedom, Precision day, log book, date more, Internet diabetescare.com; v2.6 (Abbott) Xtra, Precision QID, and time trends, sta- connection or CD- find details there Precision Xceed, tistics, data listings ROM drive, RAM about getting cables Optium Xceed, of 64 MB or more, (purchased separate- MediSense Optium, Microsoft Windows ly, ~ $29.95). Precision Sof-Tact 98 SE, 2000, or XP operating systems, monitor with 1024 × 768 or higher resolution, serial port, available 9-pin EIA- 232 or appropriate adapter for a USB for glucose meter connection WINGlucofacts 3.03 Contour, Breeze2, Global comments Windows 95, 98, Software = $19.95; (Bayer) Dex2, Elite XL and logbook, modal NT, 4.0, ME or XP, cable = $16.95, pur- day, modal week, 486 microprocessor chase online at www. glucose trend, two- higher, 32 MB, 21 ascencia store.com period comparison, MB available hard insulin dosage and disk space, CD- effects, hypo- and ROM drive, VGA hyperglycemia epi- or higher resolution, sodes, rapid swings network option

< 120 mg/dl would be expected the nomenon is a natural increase in not seem related to waking or fasting next morning. If the patient wakes blood glucose between about 4:00 blood glucose levels. up with blood glucose levels higher and 8:00 a.m., presumably as the than at bedtime, it may be related result of counter-regulatory hormone Exploring Problems With Prandial to inadequate basal medication, increases or the release of growth hor- Glucose Control reactive hyperglycemia in response mone in the early morning hours.13 Higher blood glucose attributed to to low blood glucose, or the dawn prandial inadequacies can be seen phenomenon. Case Study 1 Update when evaluating peak postprandial Reactive hyperglycemia and the B.T.’s fasting average in the past has blood glucose (determined 1–2 hours dawn phenomenon can be difficult to usually been quite good. She main- after the start of a meal). Because diagnose without the benefit of con- tains that higher readings now are many patients do not test in the period tinuous glucose monitoring (CGM) the result of her illness. She is having after eating, a good idea of postpran- or multiple nighttime SMBG values hypoglycemia at times, but this does dial coverage can be obtained from to observe trends. The dawn phe- looking at the preprandial value for

Diabetes Spectrum Volume 21, Number 2, 2008 107 already administered by the patient. There is generally a 4-hour window after the initial injection of rapid- act- ing insulin when this is most commonly seen, and it can be resolved by reducing the amount of additive correction insulin given when there is already substantial insulin remaining “on board.” For patients who take an oral sulfonylurea or use NPH insulin, hypoglycemia may be related to delayed or skipped meals.

Case Study 1 Update Figure 2 shows B.T.’s download. Despite higher blood glucose levels from her previous illness, her meter download shows that her blood glucose control is improving, and the temporary increase in insulin is now too much. Review of her SMBG results before the illness also reveals that often her blood glucose levels were lower than ideal. A change is necessary in her prandial insulin regimen. Instead of 10 units before breakfast, 8 units before lunch, and 10 units before dinner, she is asked to use 7 units before each meal, Figure 2. B.T.’s meter download. representing a 25% reduction in her the next meal. If patients can recall breakfast. She gave 15 units of aspart daily prandial insulin, and a 15% the past 2 days of meals, activity, for breakfast, and the low glucose reduction of her overall daily insulin and medications, their blood glucose level occurred 2 hours afterward. dose. control can be evaluated based on She also had another hypoglycemic the most recent data collection. reaction after dinner, and although Frequency of Testing and Related This can be important not only to she was unable to test because of the Testing Challenges interpret the data, but also to obtain severity of the tremor with the reac- The American Diabetes Association a sense of the day-to-day influences tion, she treated with candy and the (ADA) Standards of Medical Care on glucose variability. The SMBG feeling resolved. She had taken 15 in Diabetes recommends that SMBG averages on the glucose meter can be units of aspart for dinner. should be carried out three or helpful to get a global sense of how a more times daily for patients using patient is doing. For example, if the Other Common Issues Seen When multiple daily insulin injections or aggregate mean blood glucose level Evaluating SMBG Data insulin pump therapy. The ADA also for all predinner blood glucose values There are more aspects to evaluat- states that SMBG may be useful for is higher than any other time of the ing glucose control than basal- and patients using less frequent insulin day, information about what typi- prandial-related concerns. A common injections, noninsulin therapies, or cally happens at lunch and afterward issue for those using insulin is related medical nutrition therapy alone.14 would be useful. If there is no appar- to their correction algorithm, an insulin This recommendation likely stems ent reason for the hyperglycemia, dose used to normalize blood glucose from the discrepancy of data support- such as snacking or over-treatment above the preprandial target. If too ing the benefit of frequent SMBG on of hypoglycemia, then increasing much insulin is given for correction of glycemic control for those who are prandial coverage at lunch would hyperglycemia, it may be reflected in not using intensive insulin regimens. make sense. the download by hypoglycemia after However, from a practical stand- the correction dose. If an insufficient point, it can be difficult to make Case Study 1 Update correction insulin dose is given, the effective recommendations for change B.T. has begun experiencing hypo- reverse would be true, with continued without some idea of how a patient’s glycemia after meals. It is discovered hyperglycemia after treating. blood glucose manifests on a routine that she has been increasing her Another cause of hypoglycemia basis. Certainly cost considerations prandial insulin recently to cover the after a treated hyperglycemia is are a factor, although most insurance high blood glucose levels related to what is sometimes termed “insulin companies will cover their preferred her illness. She has since been treated stacking.” This occurs when insulin meter strips, and Medicare covers with antibiotics, and her pneumonia correction (using rapid- or short- three times daily testing for insulin- is resolving. She had one blood glu- acting insulin) is layered too heavily treated patients or once-daily testing cose reading of 40 mg/dl shortly after on top of the prandial dose of insulin for patients who do not use insulin.

108 Diabetes Spectrum Volume 21, Number 2, 2008 From Research to Practice / Diabetes Technology Update

Figure 3. B.T.’s follow-up. From the patients’ perspective, one 3. Additional SMBG should be per- CGM reason testing may not be a prior- formed in certain situations, such CGM devices are now available for ity is health care providers’ lack of as acute illness, intercurrent illness, home use for patients with diabetes. enthusiasm at reviewing the data. If changes in medications, for patients These devices offer a large amount there is no feedback to be gained, it is with impaired awareness of hypo- of blood glucose information, with difficult to expect adherence. glycemia, and during pregnancy. readings displayed every 5 min- An extensive review of blood 4. Postprandial SMBG testing utes. Although CGM devices offer glucose monitoring data, part of a should be used by all patients with another great way for clinicians to global consensus conference report diabetes to minimize postprandial help patients with disease manage- on SMBG,15 included the following excursions and guide lifestyle ment and for patients to have much five recommendations: changes. more data to evaluate daily decision 1. SMBG should be initiated in all 5. SMBG should be viewed as an making, they are still not the norm. patients with diabetes as an inte- education tool to inform patients With many practices still resistant gral part of an overall diabetes about the effects of lifestyle and to downloading meters, the time management program. behavioral changes. and resources needed to begin CGM 2. SMBG should be performed downloading and review on a routine ≥ 3–4 times daily for patients Certainly, barriers to the above basis make it not yet practical for treated with multiple daily insulin recommendations are well recog- most offices. This emerging technol- injections or using an insulin nized, including high out-of-pocket ogy is described in full elsewhere in pump; ≥ 2 times daily for patients costs or no insurance, language this issue (p. 112). above their glycemic target who barriers, and lower education lev- are managed with oral agents els.16,17 However, it is prudent to help Case Study 2 and once daily insulin; ≥ 1 time patients where possible to overcome P.Q. is a 64-year-old woman who per day with a weekly profile (to these barriers, given the importance presents for a review of glycemic include pre- and postprandial glu- of the information. control. She has type 2 diabetes, cose readings) for patients at their weight management issues, hyperglycemic target who are managed Case Study 1 Update tension, and dyslipidemia. She is with either once-daily insulin or B.T. comes back to the clinic con- now retired from her profession as a oral agents alone; ≥ 1 time per vinced that her meter is broken. She school teacher, and she feels that this day with more frequent weekly is getting such good results from has allowed her to start paying much profiles for patients at their gly- the recent change that she requests more attention to her personal health cemic target who are managed validation of accuracy with our clinic needs. She states that she is disap- with oral agents plus once-daily meter. The two meters show identi- pointed that today’s weight does not insulin; and ≥ 1 weekly profile for cal results. Her download summary reflect the 10-lb loss she achieved patients managed nonpharmaco- appears in Figure 3 and shows an this past fall, but also acknowledges logically, whether they are at or improved aggregate mean of 122 that she was not as careful during above their glycemic target. mg/dl with an SD of 34 mg/dl. the holiday season and therefore has

Diabetes Spectrum Volume 21, Number 2, 2008 109 Figure 4. P.Q.’s average download. 237 mg/dl, LDL cholesterol of 140 mg/dl, HDL cholesterol of 82 mg/dl, triglycerides of 75 mg/dl, and all other tests within normal limits. P.Q.’s meter download chart is shown in Figure 4, and a graph of her blood glucose readings is shown in Figure 5.

Discussion To interpret the data gathered from P.Q.’s meter, the clinician: 1. Reviews the report, highlighting key data. • What is the time period? (It should be at least 2 weeks.): The time period is 1 month Figure 5. P.Q.’s blood glucose graph. with 102 readings. • Are there any concerning regained some of the weight she had On physical examination, her hypoglycemic events (e.g., lost. She feels she is ready to renew weight is 225 lb (down 3 lb from her SMBG < 70 mg/dl)? her dietary efforts and become much weight 9 months ago). Her blood Yes, there have been eight more physically active. She has been pressure is 152/70 mmHg, but she events < 60 mg/dl. This is quite careful with food choices for has a considerable office component, discussed with P.Q., who the past week. which we have demonstrated with a reveals that these occurred Her medications include: 24-hour ambulatory blood pressure after exercising, although the monitor. Examination of her lower • Glargine, 20 units in the morning one episode that occurred in extremities shows good dorsalis • NPH, 8 units at bedtime the nighttime is a mystery. She pedis pulses. She has a rather signifi- suggests that maybe she gave • Lispro, 6–8 units at breakfast, cant fifth metatarsophalangeal on the her NPH shot twice the night 8 units at lunch, and 8 units right foot callus. This is not thick, but before. at dinner (She does not vary it is an area to be monitored, and the • What are the trends? these doses with a correction clinician draws her attention to it. On The highest mean blood algorithm.) a 10-g monofilament challenge, she is glucose average (189 mg/dl) • Losartan, 100 mg daily able to appreciate the monofilament is at night. The highest SD • Ezetimibe, 10 mg daily over all areas checked. (120.65 mg/dl) is occurring at • Niacin, 1,000 mg at bedtime Laboratory values include an bedtime, with a mean of 131 • Aspirin, 81 mg daily A1C of 8.8%, total cholesterol of mg/dl (based on four values). 110 Diabetes Spectrum Volume 21, Number 2, 2008 pharmacy/blood_glucose.html. Accessed 21 The best control is at lunch care and promoting self-manage- From Research to Practice / Diabetes Technology Update (132 mg/dl, SD 50.54 mg/dl). ment for individuals with diabetes. February 2008 • What are the average blood According to surveys, people with 7Aetna: Aetna enhances its complimentary glucose readings? diabetes would like increased ability blood glucose monitor program with Spanish language tools [article online]. Available at Morning fasting: 161 mg/dl to participate in the decision mak- www.aetna.com/news/2005/pr_20050202. (SD 71.47 mg/dl); mid-morning for their diabetes care. Because htm. Accessed 21 February 2008 ing 137 mg/dl (SD 68.89 blood glucose interpretation requires 8Edelman SV, Britton K: What people with mg/dl); lunchtime 132 mg/dl discussion with patients to identify diabetes want their caregivers to know: de- (SD 50.54 mg/dl); mid-after- reasons for variability and patterns velopment of the TCOYD patient consensus noon 164 mg/dl (SD 96.81 in control, it is a natural way to statement: Insulin 2:146–147, 2007 mg/dl); dinner 128 mg/dl (SD facilitate integration of the patients’ 9Centers for Disease Control and Prevention: 77.16 mg/dl); mid-evening perspective, needs, lifestyle, and abili- Take charge of your diabetes [article online]. 154 mg/dl (SD 72.71 mg/dl); ties into office-based decision making Available at http://www.cdc.gov/diabetes/ and bedtime 131 mg/dl (SD pubs/tcyd/ktrack.htm. Accessed 8 February for diabetes care. 2008 120.65 mg/dl). The goals of a review of SMBG 10 • What is the aggregate mean Hirsch IB: Blood glucose monitoring tech- results should include: 1) estimating nology: translating data into practice. Endocr blood glucose? the degree of blood glucose control Pract10:67–76, 2004 154 mg/dl (SD 80.71 mg/dl). and variation throughout the day; 2) 11NuMedics, Inc. CliniPro (software for 2. Interpret the data. identifying patient safety concerns care providers) [article online]. Available at • The aggregate blood glucose with regard to hypoglycemic trends www.numedics.com/products. Accessed 17 is acceptable. However, the or events; 3) understanding the February 2008 SD reveals high glucose vari- factors influencing blood glucose 12Renard E: Monitoring glycemic control: ability throughout the day. control; 4) suggesting strategies for the importance of self-monitoring of blood P.Q.’s SD needs to be lower achieving improved blood glucose glucose. Am J Med 118 (Suppl. 9A):12S–19S, 2005 to reduce risk of serious hypo- control; and 5) providing reinforce- 13 glycemia and hyperglycemic ment to patients with diabetes that Carroll MF, Schade DS: The dawn phenom- damage. enon revisited: implications for diabetes ther- this information is valuable and use- apy. Endocr Pract 11:55–64, 2005 3. Plans for care. ful in their care. Meter technology • Increase glucose management 14American Diabetes Association: Standards should be used to provide the best of medical care in diabetes—2008 [Position related to exercise to prevent care possible for patients and should hypoglycemia. Statement]. Diabetes Care 31 (Suppl. 1):S12– become the minimum requirement S54, 2008 • Because P.Q. maintains that for caring for patients with diabetes. 15 she is careful with her diet Bergenstal RM, Gavin JR III: The role of Incorporation of this technology into self-monitoring of blood glucose in the care and has not been snacking at clinical practice should be simple, of people with diabetes: report of a global night, increase her bedtime practical, and, above all, considered consensus conference. Am J Med 118 (Suppl. long-acting insulin to 10 units 9A):1S–6S, 2005 necessary for patient care. to manage her nighttime and 16 Davidson J: Strategies for improving gly- fasting hyperglycemia. Con- cemic control: effective use of glucose moni- solidate the insulin to glargine References toring. Am J Med 118 (Suppl. 9A):27S–32S, 2005 (basal) and lispro (prandial), 1Sonksen PH, Judd SL, Lowy C: Home mon- 17 to simplify her regimen. itoring of blood-glucose: method for improv- Karter AJ, Ferrara A, Darbinian J, Ackerson LM, Selby JV: Self-monitoring of blood glu- • Review new orders: glargine, ing diabetic control. Lancet 1:729–732, 1978 cose: language and financial barriers in a 10 units at bedtime and 20 2Walford S, Gale EA, Allison SP, Tattersall managed care population with diabetes. units in the morning. Lispro, RB: Self-monitoring of blood glucose: im- Diabetes Care 23:477–483, 2000 6–8 units before meals (use ~ provement of diabetic control. Lancet 1:732– 2 units per 15 g of carbohy- 735, 1978 drate in each meal). 3Mazze RS: Making sense of glucose moni- Peggy Soule Odegard, PharmD, • Instruct P.Q. to use a correc- toring technologies: from SMBG to CGM. BCPS, CDE, is an associate profes- tion dose of 1 unit of lispro Diabetes Technol Ther 7:784–787, 2005 sor of pharmacy in the School of for every 50 mg/dl that her 4Cranor CW, Christensen DB: The Asheville Pharmacy and a diabetes educator blood glucose rises above Project: short-term outcomes of a commu- and pharmacist in the Department nity pharmacy diabetes care program. J Am of Medicine at the University of 150 mg/dl before meals. This Pharm Assoc (Wash) 42:149–159, 2003 should be given in addition to Washington in Seattle. Jennifer Rose 5 her usual prandial dose. Beringer R, Shibley MC, Cary CC, Pugh CB, Beach, PharmD, CDE, is a clinical Powers PA, Rafi JA: Outcomes of a commu- assistant professor of pharmacy in nity pharmacy-based diabetes monitoring Summary program. J Am Pharm Assoc (Wash) 39:791– the School of Pharmacy and a dia- Systematic interpretation of blood 797, 1999 betes educator and pharmacist in the glucose data in the office or clinic can 6Amerihealth: Blood glucose meter pro- Department of Endocrinology and provide clinicians with invaluable gram [article online]. Available at www. Metabolism Diabetes Care Center at information for improving diabetes amerihealth.com/providers/resources/ the same institution.

Diabetes Spectrum Volume 21, Number 2, 2008 111