Phase I Final Report (Part E): Effects of Advanced Medical Technologies – Renal Diseases
Medical Technology Assessment Working Group:
Assessing the Impact of Medical Technology Innovations on Human Capital
Duke University Center for Demographic Studies
Prepared for the Institute for Medical Technology Innovation
January 31, 2006
Click here to download printable version.
Diabetes is both a serious and costly disease whose prevalence increased by 40% among U.S. adults between 1990 and 1999 and is expected to increase by 165% between 2000 and 2050, with the fastest increases occurring in older and minority subpopulations (Narayan et al., 2003). It is an emerging epidemic closely related to an increase in obesity, lack of physical activity, the advancing age of the population, and social inequalities (Nathan, 2002; Cooper and deAngelis, 2001). On a global scale, it affects nearly 200 million people. Without extensive intervention, this number is expected to exceed 300 million by 2025 (Blonde and Karter, 2005). Persons with diabetes have twice the risk of heart disease and stroke as those without diabetes (CDC Diabetes Cost- Effectiveness Group, 2002), and diabetic cardiovascular disease has become a worldwide burden (Saudek, 2002). Diabetes-related retinopathy continues to be the most frequent cause of blindness among adults aged 20 to 74 years; diabetic nephropathy is the most common cause of end-stage renal disease; and diabetes is the major cause of lower extremity amputation (Blonde and Karter, 2005). Diabetes-related mortality is increasing as well; it is the sixth leading cause of death in the U.S. and the fourth leading cause worldwide (Blonde and Karter, 2005; IDF, 2005; Cooper and deAngelis, 2001). It also causes substantial reductions in quality of life. According to Narayan and colleagues (2003), a U.S. male diagnosed with diabetes at age 40 will lose almost 12 life-years and 19 quality-adjusted life-years; a U.S. female diagnosed at the same age will lose about 14 life-years and 22 QALYs. The total cost of diabetes in the U.S. in 2002 was $132 billion, $92 billion in direct medical costs and $40 billion in indirect costs representing disability, inability to work, and premature mortality (Hogan et al., 2003). Diabetes is estimated to account for 5% to 10% of all health care expenditures worldwide (IDF, 2005).
The Diabetes Control and Complications Trial (DCCT), whose results were first published in 1993, and the United Kingdom Prospective Diabetes Study (UKPDS), published in 1998, have provided indisputable evidence that tight blood glucose control can prevent or delay complications and provide enhanced quality of life for all persons with diabetes (DCCT Research Group, 1993; UKPDS Research Group, 1998a). Our review focuses on two aspects of maintaining strict glucose control: self-monitoring of blood glucose (SMBG) and alternatives to insulin delivery via multiple daily injections. Approximately twenty-five glucose monitoring devices are on the market, and minimally- and non-invasive techniques for blood glucose monitoring are in development, in testing, or nearing FDA approval. Continuous glucose sensor (CGS) technology, combined with insulin pump therapy, will allow the first “closed-loop” systems for glucose management; i.e., an artificial pancreas. Insulin pumps, or systems for continuous subcutaneous insulin infusion (CSII), have been in existence for more than 25 years; over 200,000 are in use worldwide, about 70% of them in the U.S. (Colquitt et al., 2004). Several non-invasive methods of insulin delivery also are being investigated. In this review, we focused on the efficacy of self-monitoring of blood glucose, primarily for type 2 diabetics (i.e., the form associated with older age, obesity, a family history of diabetes, impaired glucose metabolism, and other factors; 90% to 95% of diabetics are type 2) and the efficacy of the insulin pump (CSII) compared with multiple daily injections of insulin (MDI). In addition, we reviewed briefly the status of innovations in monitoring and delivery; e.g., minimally- and non-invasive methods of blood glucose monitoring and inhaled insulin (of which at least one form is close to FDA approval).
We found the quality of evidence to be low, as confirmed by other reviewers (see e.g., Davidson J, 2005; Colquitt et al., 2004). This is in many cases due to small sample sizes, lack of comparison groups, non-randomization, and other flaws in methodological design; however, it is also due to the complexity of the disease trajectory, the large number of intervention combinations that make establishing comparability in groups difficult, and the degree to which confounding factors can contribute to results. Costeffectiveness data, while well developed for intensive therapy as a whole (e.g., DCCT Research Group, 1993), are lacking or inconclusive for individual components of it, in large part due to the paucity of data linking interventions to long-term complications and mortality (Colquitt et al., 2004).
Although self-monitoring of blood glucose is recommended at least 3 to 4 times a day for type 1 diabetics (ADA, 2005b) – those in whom the beta cells of the pancreas have been destroyed due to autoimmune, genetic, and environmental factors – and its efficacy is well established (e.g., DCCT Research Group, 1993; NIDDKD, 2005), the scientific evidence for the efficacy of self-monitoring in type 2 diabetics is tepid at best (see e.g., Sarol et al., 2005; Welschen et al., 2005). This is due both to the quality of the evidence (Blonde and Karter, 2005) and to the inconsistency of patient education surrounding self-monitoring – both in use of glucose monitors (Briggs and Cornell, 2004) and in the kinds of actions that need to be taken in response to blood glucose readings (see e.g., Davidson MB, 2005). Compliance with a self-monitoring regimen and degree of baseline glucose control are also key factors. Murata et al. (2005) found that SMBG benefited only subjects whose testing compliance exceeded 75% or who had a baseline level of glycated hemoglobin (A1C), the primary outcome measure for glucose control, greater than 8% (< 7% is recommended; ADA, 2005b). Although the health care savings achieved by preventing complications likely outweigh the short-term costs of selfmonitoring (see e.g., Karter et al., 2003; 2001), econometric modeling to establish that linkage is absent (Davidson MB, 2005).
Continuous subcutaneous insulin infusion (CSII), or insulin pump therapy, is used by only an estimated 8% of diabetics in North America. Although it is thought to offer advantages over multiple daily injections (e.g., reduction in the frequency or duration of hypoglycemic events, improvement in glycemic control, and enhancement of quality of life), results of studies have been largely inconclusive. This is due both to the quality of the evidence (see e.g., Colquitt et al., 2004) and to a concentration on narrow outcome measures (for the most part, A1C and hypoglycemia). CSII offers both advantages and disadvantages in quality of life. For example, although it provides more flexibility in mealtimes and obviates the necessity for giving oneself insulin injections on a regular basis, it also can bring significant demands (e.g., for self-care) and may have psychosocial drawbacks (e.g., CSII may affect an individual’s body image and desire to engage in sexual intimacy; Weissberg-Benchell et al., 2003). Studies conclude that CSII provides better glycemic control, as measured by reduction in baseline AlC (glycated hemoglobin), in patients with higher baseline levels (i.e., those with poorer glycemic control at the outset), but there may be no advantage of CSII over MDII for those with lower baseline levels. This suggests that CSII may be more appropriate for higher-risk patients. Unfortunately, studies consistently have been unable to differentiate between CSII and MDII in terms of complications; e.g., frequency of hypoglycemic events or weight gain. Thus, there is a lack of compelling evidence to persuade a greater proportion of diabetics to change from MDI to CSII. In fact, if type 1 diabetics, who comprise only 5% to 10% of all diabetics, represent the primary market for CSII, the market could already be saturated. CSII is more costly than MDI, and there are insufficient data to judge whether the incremental cost is offset by corresponding benefits. For type 2 diabetics who use insulin, CSII is unlikely to be cost-effective, even in the short run (see e.g., Colquitt et al., 2004).
Although there are exciting innovations on the horizon (e.g., the artificial pancreas), there seems to be a fundamental disconnect between practical needs and scientific evidence. The latter is narrowly-focused, and patients are left largely to fend for themselves when it comes to understanding the hows and whys of glucose control. Meaningful patient education is in short supply, largely because there is inadequate funding for it. This creates a vicious cycle. For example, even patients who self-monitor at an adequate level (only 33%-40%, according to Davidson J, 2005) often do not use feedback from self-monitoring to change behavior and medication (whether oral agents or insulin). As a result, changes in glucose control are marginal at best – and evidence supporting self-monitoring is weak. If meaningful advances in diabetes are to be forthcoming, education is a primary problem that will have to be addressed.
The lack of strong scientific evidence – combined with inadequate, and perhaps outmoded, mechanisms for educating both physicians and their patients – provide significant obstacles to the diffusion of new technologies and contribute to suboptimal use patterns. Higher costs for new technologies are also a factor. These are important for two reasons. Even though Medicare may reimburse a new intervention, a physician still may not prescribe it if he or she perceives that the cost is not justified. Cost is also important because of the demographics of diabetes; i.e., its disproportionate prevalence among racial and ethnic minorities, persons of low socioeconomic status, and the elderly. Cost-effectiveness studies are urgently needed that tie the efficacy of interventions to long-term outcomes, including meaningful measures of quality of life.
Diabetes is both a serious and costly disease whose prevalence increased by 40% among U.S. adults between 1990 and 1999. Further, it is estimated that the number of individuals in the U.S. with diagnosed diabetes will increase by 165% between 2000 and 2050, with the fastest increases occurring in older and minority subpopulations (Narayan et al., 2003). Diabetes is an emerging epidemic closely related to an increase in obesity, lack of physical activity, the advancing age of the population, and social inequalities (Nathan, 2002; Cooper and deAngelis, 2001). On a global scale, it affects nearly 200 million people. Without extensive intervention, this number is expected to exceed 300 million by 2025 (Blonde and Karter, 2005). For the U.S. population, the lifetime risk of acquiring diabetes is estimated to be 1 in 3 for males and 2 in 5 for females. This is considerably higher than the risk of many other major diseases; for example, the widelypublicized risk of breast cancer is 1 in 8 for U.S. women (Narayan et al., 2003).
The population burden of diabetes complications is enormous in terms of mortality, morbidity, and loss of quality of life. Persons with diabetes have twice the risk of heart disease and stroke as those without diabetes (CDC Diabetes Cost-Effectiveness Group, 2002), and diabetic cardiovascular disease has become a worldwide burden (Saudek, 2002). Diabetes-related retinopathy continues to be the most frequent cause of blindness among adults aged 20 to 74 years; diabetic nephropathy is the most common cause of end-stage renal disease; and diabetes is the major cause of lower extremity amputation. Diabetes-related mortality is increasing as well (Cooper and deAngelis, 2001), and reductions in quality of life are equally disturbing. According to Narayan and colleagues (2003), a U.S. male diagnosed with diabetes at age 40 will lose almost 12 life9 years and 19 quality-adjusted life-years; a U.S. female diagnosed at the same age will lose about 14 life-years and 22 QALYs. These are many of the reasons that diabetes consumes a disproportionate share of total health care expenditures in the United States – estimated at upwards of $100 billion (Nathan, 2002; Cooper and deAngelis, 2001).
At the same time, advances in the management of diabetes hold promise. The Diabetes Control and Complications Trial (DCCT), whose results were first published in 1993, and the United Kingdom Prospective Diabetes Study (UKPDS), published in 1998, provided indisputable evidence that tight blood glucose control can prevent or delay complications and provide enhanced quality of life for persons with diabetes (DCCT Research Group, 1993; UKPDS Research Group, 1998a). The monitoring of blood glucose is one critical component of glucose control; another – particularly for those diabetics who depend on insulin for survival, but increasingly also for those who develop diabetes later in life – is delivery of insulin. These were the primary foci of our literature review. In examining monitoring (in particular, self-monitoring) of blood glucose we devoted primary attention to the 90-95% of diabetics who develop diabetes later in life (CDC, 2005a); the necessity of self-monitoring for those who are insulin-dependent is undisputed (ADA, 2005b; DCCT Research Group, 1993). Discussion of monitoring devices applies equally to all diabetics who self-monitor blood glucose.
Procedures and Methods
To select specific topics for subsequent meta-analyses, we first performed a literature review. The results of this review are the subject of this report. We placed most emphasis on scientific findings reported in peer-reviewed journals. We supplemented published studies with judicious use of web-based information. For published research, we identified keywords for both diabetes and the major associated devices/procedures; we then searched multiple established and well-respected search engines for all relevant publications (e.g., Medline, PUBMED, Embase, the Cochrane database, Biological Abstracts, Public Health Abstracts, and the Health Periodicals Database). We also searched medical journals with high impact (e.g., the Journal of the American Medical Association, the New England Journal of Medicine, and The Lancet) and major journals for diabetes and metabolic disorders (e.g., Diabetes Care, Diabetic Medicine, the Journal of Endocrinology and Metabolism, Diabetes and Metabolism, Diabetes Research and Clinical Practice, Diabetes and its Complications, Clinical Endocrinology, Diabetologia, and Diabetes Metabolism Research and Reviews). Using multiple sources increased our confidence that we identified the appropriate “pool” of publications.
Keywords selected for the searches focused on the disease, relevant devices/ procedures for treating the disease, and the outcomes of interest, primarily reduction in morbidity and mortality and enhancement of quality of life.
Our search of web sites was targeted at sources likely to have high-quality, up-todate information. Selected web sites hosted by federal agencies and by professional organizations were our primary sources. These were predominantly the Centers for Disease Control, the American Diabetes Association, the International Diabetes Foundation, the National Institute of Diabetes and Digestive and Kidney Disorders (NIDDKD), the National Diabetes Clearinghouse, the National Diabetes Education Program, the Veterans Health Administration Diabetes Program, and Clinical Trials.gov (for ongoing trials). After careful screening, we developed a large database of publications and web sites most relevant to our research. The scientific findings reported here are based on careful assessment of the most relevant publications in this database.
Overview of the Report
We first examine epidemiologic trends in diabetes. Next we present an overview of strategies for glucose control as they relate to self-monitoring of blood glucose and insulin delivery. Finally, we summarize our findings and suggest directions for future research and analysis.
Background and Epidemiological Trends
There are two main types of diabetes mellitus. Type 1 diabetes involves a process that destroys the beta cells of the pancreas. Because the beta cells make the hormone insulin, which regulates blood glucose, this leads to severe insulin deficiency. Insulin treatment is required for survival. Onset of type 1 diabetes usually occurs in childhood or early adulthood, although it can strike at any age. It accounts for about 5% to 10% of all diagnosed cases of diabetes in the U.S. Risk factors for type 1 diabetes include autoimmune, genetic, and environmental factors (CDC, 2005a).
Type 2 diabetes was previously called “adult onset diabetes.” It usually begins as insulin resistance, a disorder in which the cells do not use insulin properly and, as the need for insulin rises, the pancreas gradually loses its ability to produce insulin. It accounts for approximately 90% to 95% of all diagnosed cases in the U.S. Type 2 diabetes is associated with older age, obesity, family history of diabetes, history of gestational diabetes (a form of glucose intolerance occurring in some women during pregnancy), impaired glucose metabolism, physical inactivity, and race/ethnicity. Ethnic groups at particularly high risk are African-Americans, Hispanic- and Latino-Americans, American Indians, and some Asian Americans and Native Hawaiians or other Pacific Islanders. Although type 2 diabetes is more common in adults, it is increasingly being diagnosed in children and adolescents (CDC, 2005a).
In 2002, approximately 18.2 million people in the U.S. had diabetes: 13 million diagnosed and 5.2 million undiagnosed (CDC, 2005a). In addition, it is estimated that 200 million people worldwide have diabetes and that by 2030 that number will rise to 350 million (Bergenstal et al., 2005). The vast majority of U.S. diabetics (18.0 million) are 20 years of age or older. Of the total population 20+ in 2002, 8.7% had diabetes. Of persons 60+, 18.3% had diabetes. The total prevalence by age group is shown in Figure 1.
Total prevalence of diabetes in people aged 20 years or older by age group in the United States, 2002
Source: 1999-2001 National Health Interview Survey and 1999-2000 National Health and Nutrition Examination Survey estimates projected to year 2002 (CDC, 2005).
Total U.S. prevalence by race and ethnicity is shown in Figure 2.
Age-adjusted total prevalence of diabetes in people aged 20 years or older by race and ethnicity in the United States, 2002
Source: 1999-2001 National Health Interview Survey and 1999-2000 National Health and Nutrition Examination Survey estimates projected to year 2002. 2000 outpatient database of Indian Health Service (CDC, 2005).
The worldwide prevalence of diabetes increased by one-third during the 1990s, due to increasing obesity and an aging population (IDF, 2005).
In 2002, 1.3 million new cases of diabetes were diagnosed in the U.S. in people 20 years of age and older. These are broken down by age group in Figure 3
Number of new cases of diagnosed diabetes in people age 20 years and older by age group in the United States, 2002
Source: 1999-2001 National Health Interview Survey estimates projected to year 2002 (CDC, 2005).
Diabetes was the sixth leading cause of death listed on U.S. death certificates in 2000. In addition, it contributed to an estimated 213,062 deaths. Worldwide, diabetes is the fourth leading cause of death (Blonde and Karter, 2005; IDF, 2005; Bansod and Shrivastava, 2004). Overall, the risk of death among those with diabetes is about twice that for people without diabetes. Diabetes may be underreported as a cause of death; e.g., studies have found that only about 35% to 40% of decedents with diabetes have diabetes listed anywhere on the death certificate, and only about 10% to 15% have it listed as the underlying cause of death (CDC, 2005a). Heart disease is the leading cause of diabetesrelated deaths, and adults with diabetes have heart disease rates two to four times higher than adults without diabetes. The risk of stroke is also two to four times higher for persons with diabetes. About 65% of deaths among people with diabetes are due to heart disease and stroke (CDC, 2005a). Mortality from diabetes also has been increasing (Figure 4). The substantial increase in diabetes mortality compared to other major disease groups (e.g., cancer) is at least partially, however, attributable to a higher rate of diagnosis of diabetes in later years than in earlier ones; i.e., increasing awareness of diabetes and its effects has led to its being recognized earlier and cited more often as a source of mortality.
Mortality from Major Diseases 1980-1995
Source: Diabetes Research Working Group (1999); Olefsky, 2001.
Other complications of diabetes include hypertension (about 73% of adults with diabetes) and blindness (diabetic retinopathy causes 12,000 to 24,000 new cases of blindness each year). Diabetes accounts for 44% of new cases of end-stage renal disease in the U.S. It also leads to nervous system disease, amputations, dental disease, pregnancy complications, biochemical imbalances, and increased susceptibility to other illnesses such as pneumonia and influenza (CDC, 2005a). Type 2 diabetics may be expected to accrue $47,240 per patient over a 30-year period for management of complications (Caro et al., 2002). Annual medical costs for treatment of type 2 diabetes increase dramatically as complications arise, and the cost of treating a patient with complications can be up to 90% greater than the cost of treating one whose diabetes is well controlled (Dixon et al., 2004; Brandle et al., 2003). Twenty to fifty percent of patients who have type 2 diabetes are not diagnosed until complications have already developed (Nathan, 2002; Cooper and DeAngelis, 2001; Harris and Eastman, 2000). Diabetes is estimated to account for 5% to 10% of all health care expenditures worldwide (IDF, 2005).
In 2002, the total cost of diabetes in the U.S. was $132 billion, $92 billion in direct medical costs and $40 billion in indirect costs representing disability, inability to work, and premature mortality (Hogan et al., 2003). These figures substantially underestimate the total cost of diabetes to the extent that they do not include the cost of diabetes-related heart disease, stroke, and other complications. A number of cost models have suggested that better glycemic control, particularly in type 2 diabetics, will lead to significant longer-run reductions in the economic burden of diabetes by reducing complications (e.g., Eastman et al., 1997a, b; DCCT Research Group, 1996). More recent studies (e.g., Wagner et al., 2001; Testa and Simonson, 1998; Gilmer et al., 1997) suggest that cost savings may be evident within a shorter time span (e.g., within three years).
Self-Monitoring of Blood Glucose
Glycemic control is fundamental to the management of diabetes (see ADA, 2005b). Prospective randomized clinical trials such as the DCCT (DCCT Research Group, 1993) and the U.K. Prospective Diabetes Study (UKPDS Research Group, 1998a; 1998b) have shown that improved glycemic control is associated with sustained decreased rates of retinopathy, nephrophy, and neuropathy (DCCT-EDIC Research Group, 2000). The American Diabetes Association recommends A1C as the best test for glycemic control (ADA, 2005a). A1C is also known as glycated hemoglobin or glycosylated hemoglobin and indicates the level of blood glucose over the last 2-3 months.12 Treatment regimens that reduced average glycated hemoglobin (A1C) to ~7% (~1% above normal limits; ADA, 2005b) were associated with fewer long-term microvascular complications despite an increased risk of severe hypoglycemia and weight gain (Lawson et al., 1999; Stratton et al., 2000). Improved glycemic control promotes weight gain by reducing both basal metabolic rate (BMR) and glucosuria, the presence of abnormally high levels of glucose in the urine (see e.g., Mäkimattila et al., 1999). Hypoglycemia results if blood glucose levels fall too low and is always a risk, albeit a manageable one, in any form of intensive treatment regimen (see Colquitt et al., 2004).
body of evidence for the efficacy of self-monitoring of blood glucose has been emerging more recently (for a review, see Blonde and Karter, 2005). Although we now know that effective control of blood glucose is critical to the management of diabetes, data from the third National Health and Nutrition Examination Survey (NHANES III) conducted 1988 to 1994 and NHANES 1999 to 2000 show that glycemic control did not improve between the two survey periods; in fact, it actually worsened among individuals with type 2 diabetes (Davidson J, 2005). UK, Swedish, and Dutch studies have shown comparable results (Ubink-Veltmaat et al., 2003; Berger et al., 1999; Gatling et al.,
1A1C is formed when glucose in the blood binds irreversibly to hemoglobin to form a stable glycated hemoglobin complex. Since the normal life span of red blood cells is 90-120 days, the A1C will only be eliminated when the red cells are replaced; A1C values are directly proportional to the concentration of glucose in the blood over the full life span of the red blood cells. The A1C value is an index of mean blood glucose over the past 2-3 months but is weighted to the most recent glucose values. Values show the past 30 days as ~50% of the A1C, the preceding 60 days giving ~25% of the value and the preceding 90 days giving ~25% of the value. This bias is due to the body's natural destruction and replacement of red blood cells. Because red cells are constantly being destroyed and replaced, it does not take 120 days to detect a clinically meaningful change in A1C following a significant change in mean blood glucose.
1998). Increased use of self-monitoring of blood glucose (SMBG) is among the many proposed strategies to address the problem. Measurement of glycated hemoglobin (A1C) provides an overall, long-term assessment of glucose control and its potential effects on the kidneys and other organs but does not provide information on daily fluctuations that can be used in treatment decisions to optimize blood glucose control and minimize complications. SMBG, on the other hand, provides feedback on the impact of nutrition, physical activity, therapy with oral antidiabetic agents, and insulin, and allows design, implementation, and adjustment of physiologic insulin replacement programs (Blonde and Carter, 2005). In addition, almost all guidelines for diabetes management support the integral role of SBMG in overall treatment programs (see e.g., Owens et al., 2004).
At the same time, self-monitoring of blood glucose poses challenges for both the patient and the health care provider. Health providers must familiarize themselves with the value, techniques, and objectives of self-monitoring, educate their patients, and provide follow-up. Patients must understand what they are trying to accomplish, learn to use glucose meters correctly, and respond appropriately to blood glucose values in “real time;” e.g., to adjust insulin or oral medication dosage, change food intake or physical activity levels, and make other lifestyle modifications aimed at achieving optimum glucose control (Bergenstal et al., 2005; Davidson J, 2005; Skelly et al., 2005; Peel et al., 2004).23 Some findings suggest that patients who would benefit the most from intensive
2 Failure to educate, and the resulting failure to act on results of self-monitoring, may be a primary reason that results of many studies designed to measure the effect of SMBG have been inconclusive (Ipp et al., 2005; Davidson MB, 2005; Davidson et al., 2004). One-on-one patient education is time-intensive. Unfortunately, despite its promise, current evidence does not seem to support the efficacy and costeffectiveness of telemedicine interventions (e.g., telephone support, computerized educational programs on disk, computer logging of results) to support SMBG; for a meta-analysis, see Farmer et al., 2005. For an overview of initiatives to improve education in diabetes management, see NDEP, 2005 and NDIC, 2005
diabetes management may be the least likely to monitor blood glucose. Those may include the elderly, minorities, those with lower socioeconomic status, the obese, those with significant comorbidities, and those with poorer glycemic control (Briggs and Cornell, 2004; Adams et al., 2003). Frequency of self-monitoring is consistently shown to fall below recommended levels (see e.g., Davidson J, 2005; Gulliford and Latinovic, 2004; Vincze et al., 2004; Harris, 2001; Karter et al., 2000). Karter and colleagues (2000) found that 60% of patients with type 1 diabetes and 67% of those with type 2 diabetes failed to monitor at the frequencies recommended by the American Diabetes Association (ADA, 2005b). Those percentages may be even lower in Europe (see e.g., Yki-Järvinen et al., 1999). Urine testing has been advocated for those not being treated with insulin, particularly since it is less costly (see e.g., Miles et al., 1997; Gallichan, 1994), but its efficacy is unproven (Coster et al., 2000), and it may be associated with patient dissatisfaction or misunderstanding of implications (e.g., “since I only have to test urine rather than use a blood glucose meter, I must not be as seriously ill”) or results (see Lawton et al., 2004).
Blood Glucose Self-Monitoring in Type 2 Diabetic Patients
The Diabetes Control and Complications Trial, which enrolled only type 1 patients, demonstrated clearly that those individuals performing frequent self-monitoring of blood glucose as part of intensive insulin therapy showed improved glycemic control and decreased diabetic complications (DCCT Research Group, 1993a). The importance of SMBG is reflected in clinical practice guidelines, in which SMBG is recommended three or more times a day for most patients with type 1 diabetes (ADA, 2005b).
The evidence for SMBG in type 2 diabetics, particularly in those not using insulin, has not been as strong – and the ADA recommends a frequency “sufficient to facilitate reaching glucose goals” (ADA, 2005b). A number of randomized controlled trials (e.g., Davidson et al., 2004; Kwon et al., 2004; Guerci et al., 2003; Schwedes et al., 2002; Brown et al., 2002) and other cross-sectional or non-randomized studies (e.g., Karter et al., 2005; Karter et al., 2001; Chan et al., 2000) have shown that SMBG, if done frequently enough, is associated with improved glycemia; however, others have failed to show a relationship between SMBG and reduced A1C values (e.g., Davidson et al., 2005; Franciosi et al., 2005; Wen et al., 2004; Coster et al., 2000 [a meta-analysis]; Meier et al., 2002; Harris 2001). The reasons for this inconsistency are varied. Primary among them are wide differences in the frequency of self-monitoring; compliance rates with the prescribed SMBG regimen ranging from 25% to 100%; and study design (for example, the degree to which patients received feedback on the results of self-monitoring and education about actions they should take in response to the readings).
One of the largest observational studies of type 2 diabetics has been the Diabetes Outcomes in Veterans Study (DOVES); see e.g., VHA, 2005; Murata et al., 2005; Murata et al., 2004a,b,c,d,e,f; Murata et al., 2003a,b,c; Hoffman et al., 2003. The results from DOVES and from two recent meta-analyses (Sarol et al., 2005; Welschen et al., 2005), the former a meta-analysis of the literature from 1966 to 2004, have been favorable but tepid; i.e., either the magnitude of the difference was small or the reported value of selfmonitoring had significant qualifiers associated with it.
In general, the quality of evidence is low (see e.g., Blonde and Karter, 2005). Sarol and colleagues, for example, excluded 291 of 296 studies retrieved, citing among other problems non-randomized controlled trials, inclusion of non-type 2 patients, absence of comparison groups, absence of relevant outcomes (e.g., A1C concentrations), non-comparability of treatment groups (e.g., significant differences in patient demographics), non-specification of co-interventions, and other confounding factors. Data analysis using a random-effects model showed a mean A1C reduction for therapies that included SMBG testing as part of a multicomponent management strategy compared with therapies that did not use self-monitoring; however, the authors concluded only that “in the short-term, and when integrated with educational advice, SMBG as an adjunct to standard therapy may contribute to improving glycemic control among type 2 diabetic patients.” Welschen et al. (2005) came to similar conclusions in a meta-analysis of 5 studies representing 1,159 non-insulin-treated patients. They reported that “selfmonitoring of blood glucose might be effective in improving glycemic control in patients with type 2 diabetes who are not using insulin.”
Two DOVES studies report on the effects of SMBG in patients with stable, insulin-treated type 2 diabetes (the majority of studies have been of non-insulin-treated type 2 diabetics). The earlier study (Hoffman et al., 2002) assessed once- and twice-daily SMBG testing in 150 subjects over an 8-week period. It found that the overall correlation of glucose testing and A1C was 0.79 (p < .0001). Mean blood glucose values for each of four once-daily testing strategies were significantly correlated with A1C (r = 0.65-0.70, p < .0001), as were mean blood glucose values for each of six twice-daily testing strategies (r = 0.73-0.75, p < .0001). A more recent study (Murata et al., 2005) examined a fourtimes daily testing strategy over an 8-week period in 201 subjects. SMBG benefited only subjects whose testing compliance exceeded 75% or who had a baseline A1C > 8.0%. Results from DOVE, however, show that self-monitoring of blood glucose is superior to A1C measurement in predicting long-term hypoglycemia with type 2 diabetes (Murata et al., 2004c). The reason for this is that A1C is an average measure of blood glucose over a 2-3 month period but does not show variability of blood glucose levels within that time period; i.e., a person with widely fluctuating glucose levels on a daily basis can report the same A1C value as a person for whom blood glucose remains relatively stable; yet the former is at a much higher risk for hypoglycemia.
Blood Glucose Meters
Clearly, compliance is a key factor in the success of self-monitoring of blood glucose. It is dependent upon many factors, the most important among them patient education (Briggs and Cornell, 2004). Another component is ease of use – a factor that has been addressed by advances in blood glucose measurement devices. More than 25 blood glucose meters are commercially available. According to Briggs and Cornell, the best meter for a patient with diabetes is the one he or she will use correctly. The diabetes self-testing market is doubling every five years (Bansod and Shrivastava, 2004). Most devices are invasive in nature and consist of glucose test strips and a photometric or electrochemical mechanism to determine glucose concentrations in whole blood. Meters differ in the amount of blood needed for each test, testing speed, overall size, ability to store test results in memory, and cost of test strips used (for a review, see Bansod and Shrivastava, 2004). They are generally user-friendly, requiring two or three steps to operate and with results displayed in a minute or less. User error is the primary cause of inaccurate blood glucose readings (see e.g., Nettles, 1993), again emphasizing the need for education in a meter’s use (for example, coding new test strips and making sure they are not exposed to extreme temperature or humidity). Factors that can affect the accuracy of blood glucose meter readings include hematocrit abnormalities, inadequate blood sample sizes (a cause of falsely low readings), and severe dehydration or hypotension (Briggs and Cornell, 2004).
As one means of reducing non-compliance with SMBG, there is a move toward alternative ways of testing blood glucose that will decrease the inconvenience, time, and pain associated with traditional “finger-stick” testing. These include alternate site testing, minimally-invasive methods that do not penetrate into deep layers of tissues where nerve terminals are located – primarily glucose measurement in interstitial fluid (ISF) – and non-invasive methods using, for example, reverse iontophoresis (extraction of glucose through the skin and measuring the sample using an electrochemical-enzymatic sensor), crystalline colloidal arrays, infrared spectroscopy, radiowaves, photoacoustic spectroscopy, scatter changes and florescence, and optical rotation of polarized light (see Bansod and Shrivastava, 2004). Among innovations is a disposable contact lens that changes color according to blood glucose levels. Non-invasive methods are now, at best, in early clinical trials. The cost of most minimally- and non-invasive methods is generally high – see Bansod and Shrivastava, 2004.
The greatest promise for non-invasive glucose measurement lies in devices that measure blood glucose on a continuous basis (see e.g., Kovatchev et al., 2004; Tanenberg et al., 2004; Aussedat et al., 2000; Tamada et al., 1999). Continuous glucose sensor (CGS) technology has the potential to revolutionize diabetes management by providing patients with ongoing feedback about current blood glucose levels and the rate and direction of the change and signaling dangerous trends such as rapid glucose descents that may lead to hypoglycemia (Kovatchev et al., 2004). Although the DCCT (DCCT Research Group, 1993) and UKPDS (UKPDS Research Group, 1998a) clearly demonstrated a reduction in long-term complications when blood glucose was strictly controlled, hypoglycemic events also tripled – and as many as 7 blood glucose measurements per day were insufficient to detect a number of severe hypoglycemic and hyperglycemic events (DCCT Research Group, 1993). Most glucose sensors developed for continuous monitoring do not measure blood glucose directly, but rather rely on measurement of glucose levels in the interstitial fluid (ISF) of subcutaneous tissue (for a review, see Wilson and Gifford, 2005).
Most importantly, however, development of CGS technology is a step forward in the development of “closed-loop” insulin delivery; i.e., an “artificial pancreas” that eliminates the need for self-monitoring and calculation of insulin doses and approximates the natural physiologic insulin delivery of the pancreatic ß cells. Insulin pump technology (see below) has been in existence for at least 25 years; combining that technology with CGS technology – and an algorithm – will soon make closed-loop systems a reality (see e.g., Steil et al., 2004).
Several devices have been approved for alternate site testing (e.g., forearm, upper arm, thigh, calf, palm, and abdomen). One advantage of alternate site testing is that it requires smaller samples of blood; however, there is some controversy about whether it accurately represents the patient’s blood glucose state. Research indicates that there may be variations between finger-stick and alternative site readings, especially during times of rapidly changing glucose levels (e.g., after a meal or exercise) – see Bennion et al., 2002; Lock et al., 2002; Ellison et al., 2002; McGarraugh et al., 2001; Jungheim and Koschinsky, 2001; McGarrugh, 2001.
The cost-effectiveness of SMBG has not been studied apart from the costeffectiveness of intensive glycemic control. If patient education is sufficient to ensure adherence to a self-testing regimen, and if appropriate actions are taken in response to readings (e.g., modifications in diet, exercise, or insulin levels) so that clinical parameters are improved significantly, then the healthcare savings achieved by preventing complications should outweigh the short-term costs of increased use of self-monitoring (Davidson J, 2005; Karter et al., 2003). Econometric modeling is needed to establish that linkage.
Insulin Delivery Systems
The most common means of insulin delivery is multiple daily injections of insulin (ADA, 2005a; National Guideline Clearinghouse, 2005). These provide a significant reduction in quality of life for the diabetic due to the necessity of frequent monitoring of blood glucose and adhering to a regular meal schedule. In addition, conventional monitoring methods (i.e., frequent blood draws) and administration of shots can be painful. Our review focuses on two alternatives to multiple daily injections (MDI): continuous subcutaneous insulin infusion (CSII) and inhaled insulin (INH).
Continuous Subcutaneous Insulin Infusion (CSII)
Although insulin therapy has been in existence since the early 1920s, continuous subcutaneous insulin infusion (CSII), often called insulin-pump therapy, was introduced in the 1970s as a way of achieving and maintaining strict control of blood glucose concentrations in people with type 1 diabetes. More recently, it has found limited use with type 2 diabetics as well. CSII uses rapid-acting insulin to deliver a small amount of insulin on a continuous basis (i.e., to maintain a basal rate). At meal times, a bolus (increased) dose of insulin is delivered, and additional bolus doses can be given at other times to correct high blood glucose levels. Use of the insulin pump requires selfmonitoring of blood glucose by the user and a decision about how much insulin to deliver. Compared to traditional intensive insulin therapy consisting of multiple daily injections (MDI), CSII is thought to reduce the frequency and/or duration of hypoglycemic events, provide better glycemic control, improve quality of life by providing for more flexibility, and possibly reduce the risk of ketoacidosis34 (Schuffham and Carr, 2003; Pickup et al., 2002; Lenhard and Reeves, 2001). Prevention of hypoglycemia is a key advantage. MDI may lead to blood glucose levels falling below normal, robbing the brain of its essential glucose supply. The consequences of these hypoglycemic events vary. In some cases they cause only a feeling of hunger and sweating, but when they occur at night they may disturb sleep. Severe hypoglycemia can lead to behavioral disturbances, unconsciousness, convulsions, and even death. Children experiencing frequent hypoglycemic events may also show impairment of intellectual function (Colquitt et al., 2004). Most importantly, fear of hypoglycemia may cause those undergoing intensive insulin therapy via MDI to keep their blood glucose levels higher than recommended in order to avoid hypoglycemic episodes. This provides for less than optimal glycemic control and increases the risk of diabetic complications.
It is estimated that about 200,000 insulin pumps are in use worldwide, about 140,000 of those in the U.S. (Colquitt et al., 2004). As a percentage of the total
3 Diabetic ketoacidosis is a life-threatening metabolic disturbance related to shortage of insulin. It is often brought on by other illnesses such as infections that increase the body’s insulin needs.
population, insulin-pump use is relatively small (e.g., about 8% of type 1 diabetics in North America). Figure 5 gives comparative percentages for several countries. Modern pumps are battery operated and hold enough insulin for several days. Most have alarms for empty cartridges, low batteries, occlusion of tubing, or faulty electronics.
CSII use by Country Source:
Colquitt et al., 2004.
Clinical Effectiveness of Continuous Subcutaneous Insulin Infusion (CSII)
Since the insulin pump was introduced in the late 1970s, it has been the subject of a large number of studies seeking to compare glycemic control with CSII to multiple daily injections of insulin (MDI). Although the majority have shown at least slightly better glycemic control for patients using pumps, the quality of available evidence makes firm conclusions difficult to draw. There are several reasons for this. Many studies have methodological flaws, poor research designs, no comparison groups or controls, and small sample sizes. Others combine subjects from different age groups (e.g. children and adults) or mix persons with type 1 and type 2 diabetes. Many also were undertaken in the 1980s with first-generation pumps whose performance was suboptimal compared to those in use today (see e.g., Colquitt et al., 2004; Retnakaran et al., 2004). Early studies also used regular (human) insulin instead of the rapid-acting insulin analogues such as insulin lispro and insulin aspart or used human insulin in one group, analogues in the other (see e.g., Wood, 2005; Retnakaran et al., 2004). Equally importantly, outcome measures have consisted largely of narrow measurements of glycemic control (e.g., A1C concentrations or rates of hypoglycemia) and have only recently begun to address issues of quality of life, including increased demands on self-care (the need to learn new skills, do more frequent monitoring of blood glucose and urine ketones, and increase awareness of insulin-to-carbohydrate ratios) and potential psychosocial demands – e.g., CSII may affect an individual’s body image and desire to engage in sexual intimacy (Weissberg- Benchell et al., 2003).
We could find only four systematic reviews comparing CSII and MDI in type 1 adult diabetics: Colquitt et al., 2004 (a review of 20 studies, with meta-analysis performed where appropriate); Retnakaran et al., 2004 (a pooled analysis of three randomized controlled trials); Weisberg-Benchell et al., 2003 (a meta-analysis of 52 studies); and Pickup et al., 2002 (a meta-analysis of 12 RCTs). A comprehensive review is now being undertaken by Misso and colleagues (2005), but results are not yet available.
The conclusions of the four reviews must be viewed with the above limitations in mind; however, results are generally consistent. The studies conclude that CSII provides better glycemic control, as measured by reduction in baseline AlC (glycated hemoglobin), in patients with higher baseline levels (i.e., those with poorer glycemic control at the outset), but there may be no advantage of CSII over MDII for those with lower baseline levels. This suggests that CSII may be more appropriate for higher-risk patients. Because of data limitations, the studies were unable to differentiate between CSII and MDII in terms of complications; e.g., frequency of hypoglycemic events or weight gain. Results from the French registry representing a collaboration of all implanting centers in France (the EVADIAC group) are more positive. For example, one EVADIAC study (Jeandidier et al., 1996) found that severe hypoglycemia occurred onesixth as frequently with the implanted pump (0.69 per patient-year with MDI, but only 0.11 per patient-year with CSII). Other EVADIAC studies (e.g., Hanaire-Broutin et al., 1995; Broussole et al., 1994) demonstrated improved glycemic control even for patients who were previously receiving intensive insulin therapy via MDI (i.e., those whose glucose was already reasonably well controlled) and reported a low (ca. 3 %) rate of explantation (removal of the device) because of complications (e.g., a slowing of the pump rate due to insulin deposits on the pumping mechanism, catheter obstruction, or infection at the implantation site). Nevertheless, use of the insulin pump requires a degree of patient management that is not attractive to everyone eligible for CSII (see e.g., Johansson et al., 2005). In one EVADIAC center, 16% of all treated patients chose to give up the implanted pump. This group of patients did not show fewer benefits or more adverse events than other patients; rather they found the demands of the pump too onerous (Schaepelynck-Bélicar et al., 2005). This perception contrasts with general findings from reviews (e.g., Colquitt et al., 2004; Radermecker and Scheen, 2004; Lenhard and Reeves, 2001) that cite lifestyle advantages, among them increased flexibility, and increased levels of patient satisfaction with CSII.
There are scarce data on the use of CSII in children and adolescents, and these studies have significant limitations (Lenhard and Reeves, 2001). For example, none is randomized, and most have a small number of subjects and are of a limited duration. Results are also contradictory. Some studies (e.g., Boland et al., 1998; Oesterle et al., 1998; Tamboriane et al., 1989; De Beaufort and Bruning, 1987; Schiffrin et al., 1984; Rudolf et al., 1982) report that insulin pump therapy provides as good or better metabolic and glycemic control than MDI and that it is well tolerated. Others report that it is associated with lower or comparable rates of complications such as hypoglycemia, ketoacidosis, and weight gain (e.g., Boland et al., 1999; Steindel et al., 1995; Bougneres et al., 1984). Other studies, however, found that CSII did not provide better glycemic control and was not well tolerated in children and adolescents and that it was associated with more complications (e.g., Knight et al., 1986; Becker et al., 1984). Some contradictory results may be explained by pump improvements since the 1980s; however, the quality of evidence does not permit firm conclusions. The state of current evidence may be best expressed by one recent study (Weintrob et al., 2003) that concluded that “intensive insulin therapy by either CSII or MDI is safe in children and young adolescents with similar diabetes control and a very low rate of adverse events – and both modes should be available.”
Results of studies examining the efficacy of CSII in pregnant women with type 1 diabetes, particularly its potential role in reducing the rate of fetal malformation, have been equally contradictory or inconclusive (for a review, see Colquitt et al., 2004; Lenhard and Reeves, 2001). The role of CSII for pregnant women with type 2 or gestational diabetes has not been well studied (Lenhard and Reeves, 2001). A review of CSII in pregnant women with all three types of diabetes is ongoing (Tuffnell et al., 2005); however, no results are available at this time.
We found only two randomized studies comparing MDI with CSII in adult type 2 diabetic patients (Herman et al., 2005; Wainstein et al., 2005), but sample sizes were small (98 and 40 subjects, respectively) and results inconclusive. The Hermann study found no differences in efficacy of glucose control by baseline A1C, sex, BMI, or study site, or in the number of hypoglycemic events, weight gain, or treatment satisfaction. The Wainstein study, which enrolled only obese patients with severe insulin resistance and poor glycemic control, found CSII to be superior to MDI in reducing A1C without significant change in weight or insulin dose. This seems to correspond to similar, albeit tentative, findings for type 1 diabetics that those with poor initial control may benefit most from CSII. Two earlier studies on short-term CSII in adult type 2 diabetic patients (Dupuy et al., 2000; Valensi et al., 1997) were inconclusive.
Cost-Effectiveness of CSII
Most cost-effectiveness studies of insulin therapy in diabetes have focused on comparison of intensive insulin therapy (which includes both MDI and CSII) to conventional therapy – see e.g., Gray et al., 2000; DCCT Research Group, 1996. The cost of intensive therapy has been estimated at $16,000-$30,000 per quality-adjusted life year gained for type 2 diabetics (ADA, 2003; DeWitt and Hirsch, 2003), well below the benchmark value of $50,000 that is generally accepted as “cost-effective” in the U.S. The CDC Cost-Effectiveness Group (2002) estimates a cost-effectiveness ratio somewhat higher: $41,384 per QALY. However, cost-effectiveness ratios are tied to age at diagnosis. The cost per QALY is only $9,614 for those diagnosed between the ages of 25 and 34, but reaches $2.1 million for patients aged 85 to 94 (CDC Cost Effectiveness Group, 2002). The incremental yearly cost of intensive therapy appears to be small – $1,866 in the UKPDS trial (Gray et al., 2000). In one Japanese trial, injection therapy in type 2 patients reduced costs from $31,525 for conventional therapy to $30,310 by decreasing complications (Wake et al., 2000). Significant short-term cost savings resulting from improved glycemic control in type 2 diabetics have been demonstrated in U.S. studies as well (e.g., Wagner et al., 2001; Testa and Simonson, 1998). The DCCT trial of type 1 diabetics concluded that, “from a healthcare perspective, intensive therapy was well within the range of cost-effectiveness considered to represent good value,” although the marginal benefits would vary depending on the level of control achieved with conventional treatment (DCCT Research Group, 1996). Although the DCCT included patients treated with both MDI and CSII, they were not assigned randomly, and the study design as a whole did not allow direct comparison of the two methods of intensive therapy.
To our knowledge, only two studies have attempted to systematically assess the cost-effectiveness of CSII compared to MDI (Colquitt et al., 2004 and Schuffham and Carr, 2003). In both studies, data came from the available literature on costs and effectiveness of MDI and CSII taken separately. Schuffman and Carr constructed a Markov model to estimate the costs and outcomes for patients with type 1 diabetes. The primary outcome measure used was quality-adjusted life years. Colquitt and colleagues reviewed cost and outcome data for both type 1 and type 2 diabetes. Given the limitations of the data (primary among them that there are no data linking pumps to longer-term complication or mortality data), no firm conclusions could be drawn. CSII is more costly than MDI. The cost of the pump itself is around $5,000-$6,000 (ADA, 2004; Lenhard and Reeves, 2001); infusion sets and catheters must be purchased regularly; and additional costs are incurred for patient management and education. These incremental costs must be offset by larger benefits in order for CSII to be more cost-effective than MDI. It appears from limited evidence available that CSII offers cost-benefit advantages over MDI only if targeted to those who may benefit most from it (Schuffham and Carr, 2003).
Although methods of subcutaneous insulin delivery have advanced considerably in the last 80 years, they still represent a degree of invasiveness that is inconvenient, in one way or another, to the patient. Several non-invasive methods of insulin delivery are being investigated, some still in animal studies (see e.g., DeWitt and Dugdale, 2003; McAuley, 2001). The most promising at present is inhaled insulin. Two pulmonary insulin systems, one using insulin in powder form and one aerosol, are now in advanced stages of development and testing and at least one (the powder insulin delivery system) is currently under evaluation by the U.S. Food and Drug Administration (FDA) and is expected to be approved (Trubo, 2005).45 Inhaled insulin has the potential to reduce the number of injections to as few as one long-acting insulin per day and provide a closer match to the natural production of insulin because of more rapid absorption from the lung (Royle et al., 2005). Key potential advantages of inhaled insulin are convenience (see e.g., Gerber et al., 2001) and a more natural physiologic process that provides levels of insulin closer to those produced naturally by the pancreas in non-diabetic persons. The principle disadvantage of inhaled insulin is a range of bioavailability estimated at 10-30% (see Royle et al., 2005; Skyler, 2001); i.e., the percentage of insulin that reaches the
4 The FDA announced recently that it was postponing approval of Exubera until further data could be examined: http://www.exuberareport.com/.
bloodstream is somewhat lower than with injected insulin and larger quantities of insulin are required.
The most recent systematic review of inhaled insulin use in diabetes mellitus is that of Royle and colleagues (2005); another (Burt et al., 2005) is ongoing. Royce and colleagues examined six randomized controlled trials with a total number of 1,191 participants and whose results were published in 2001 or 2002. The trials compared the efficacy of inhaled insulin (INH) to subcutaneous insulin (SC). Outcomes reported were A1C concentrations, patient satisfaction, quality of life, hypoglycemic episodes, changes in antibody levels, weight change, and other adverse effects. Three trials included subjects with type 1 diabetes (n = 735) and three included subjects with type 2 (n = 456). Overall results showed equivalence between INH and SC (that is, glucose control was comparable); patient satisfaction was higher with INH; quality-of-life improvement was seen for INH; there was little or no difference in the frequency of hypoglycemic episodes; INH-treated patients generally developed increased levels of serum antibody binding, but the higher levels did not appear to have clinical significance;56 and weight gain was similar or slightly less for INH (see e.g., Cefalu, 2001). There was no evidence of pulmonary side effects other than an early increased incidence of coughing that decreased over the course of the study. No differences were found between type 1 and type 2 patients. These findings are confirmed in more recent studies for both type 1 and type 2 patients (see e.g., Skyler et al., 2005; Quattrin et al., 2005; DeFronzo et al., 2005;
5 The lack of clinical significance is supported by two recent studies (Fineberg et al., 2005; Heise et al., 2005). The increased development of antibodies did not impair glucose tolerance, alter the time-action profile of the insulin, or give rise to allergic or other hypersensitivity reactions.
Freemantle et al., 2005; Hermansen et al., 2004; Hollander et al., 2004; Weiss et al., 2003).
Inhaled insulin holds particular promise for type 2 diabetics who may benefit from short-term insulin therapy or for whom lifestyle changes and oral agents are ineffective. There is some preliminary evidence that elderly patients may need to inhale more insulin than young patients to achieve similar glycemic control (Henry et al., 2003), but the ramifications of this, if indeed the finding holds in future studies, are unknown at this time. The most likely effect would be a reduction in cost-effectiveness, but no costeffectiveness analyses are possible since the devices are not yet on the market and prices are unknown.
Patient satisfaction seems to be the only conclusive distinguishing factor between the two regimens. Cappelleri et al. (2000) reported a level of patient satisfaction almost three times higher for INH than for SC. More recently, Rosenstock et al. (2005) reported that 85% of patients chose to remain on INH after the study period, and 75% of those on SC chose to switch to INH. Sample sizes in both studies are small; hence results can be used only to support the general consensus that patient satisfaction is higher with INH. A much longer experience with INH will be needed to draw firm conclusions about its efficacy – and particularly to investigate potential long-term pulmonary effects. Equally importantly, the economic viability of inhaled insulin will need to be assessed.
Conclusions and Implications
Compared to most chronic diseases, diabetes is unusually burdensome, but also holds unusual promise for effective management and control. Diabetes is unusually burdensome for several reasons – it is a major cause of mortality; unless vigilantly controlled, diabetes is a strong risk factor for the development of other chronic diseases and conditions; and monitoring and achieving glycemic control requires rigorous selfdiscipline and typically includes painful testing and insulin delivery.
For all of its risks and complications, diabetes is ultimately a (largely) controllable disease, especially type 2 diabetes, which accounts for 90-95% of the illness burden. Medical technology is the critical element in both improved disease management observed to date and promising avenues for further improvements in control and management. Evolution of technology for the control and management of diabetes occurred rapidly. Until the late 1970s, the only practical regimen for diabetics was monitoring blood glucose levels via “finger-sticking” and controlling them via multiple daily insulin injections. Since then there have been substantial improvements: less invasive blood glucose monitors and continuous subcutaneous insulin infusion pumps. Moreover, recent devices are more effective and user-friendly than those that first became available in the late 1970s.
Obstacles to Utilization of New Technologies
As detailed in this report, evidence about the effectiveness and efficacy of new means for blood glucose monitoring and insulin delivery is lower in both quantity and quality than desirable. Nonetheless, empirical evidence is sufficient to conclude that the new technologies, when applied correctly and consistently, benefit glycemic control and may improve quality of life as well. Despite good reasons for widespread use of these technologies, utilization rates remain low – and even diabetics who use these technologies often do not derive the full benefits from them.
Obstacles to utilization include both impediments to the timely diffusion of the technologies and suboptimal utilization patterns. We offer comments on both of these issues.
Slow Diffusion of Devices.
Little is known about the actual rate of diffusion of cutting-edge monitoring and insulin delivery systems. It is abundantly clear, however, that diffusion has been sluggish. Evidence suggests that very few type 2 diabetics use insulin pumps, and relatively few use glucose monitors. Factors accounting for the slow diffusion of these technologies have not been systematically investigated (i.e., the relative importance of specific factors is unknown). Nonetheless, some of the reasons for less than optimal utilization of these devices are known.
Given the very low rates of utilization, it seems apparent that physicians are not prescribing them widely. Reasons for this are unclear, but probably include insufficient evidence that the new devices (e.g., CSII) are superior to traditional methods of delivering insulin (MDI) or that self-monitoring of blood glucose is worth the effort and cost, lack of experience with using the devices and teaching patients to use them, and insufficient information to choose the specific devices that will best meet each patient’s needs and lifestyle. For whatever reasons, it is clear that physicians have not been eager advocates of self-monitoring devices and insulin pumps.
The greater cost of managing diabetes with self-monitoring devices and insulin pumps is undoubtedly another factor that decreases utilization of them. Higher costs are especially important because of the demographics of diabetes – i.e., its disproportionate prevalence among racial and ethnic minorities, persons of low socioeconomic status, and the elderly. It is worth noting that these devices are reimbursable under Medicare, but utilization rates have not increased dramatically since their approval – again suggesting that the major impediment to more widespread use is resistance by physicians rather than by patients.
In addition, self-monitoring devices and insulin pumps require a different skill set than finger sticks and self-administered injections. It is not clear that the skills needed for the appropriate use of these devices are necessarily more inherently difficult than those required for more conventional techniques, but they are clearly different. Patient education is a critical component of the use of these devices, and most observers believe that it is frequently short-changed. The demands that adequate patient education place on health providers may make them more resistant to prescribing these devices, and fears that the devices may be too complicated may make some patients resistant to them.
A final consideration is the fact that type 2 diabetes accounts for the overwhelming disease burden but that only about 27% of type 2 diabetics use insulin (Koro et al, 2004). Clearly then, the potential market for insulin pumps is much smaller than the overall prevalence of type 2 diabetes. Non-insulin treatment may decrease use of devices for self-monitoring of glucose levels as well. Both physicians and patients may believe that rigorous self-monitoring is less important for diabetes that can be effectively treated without insulin.
Suboptimal Use of Devices.
Both self-monitoring devices and insulin pumps have the potential to contribute to more effective management and control of diabetes than conventional approaches. But translation of potential benefits to actual benefits is dependent on the commitment and self-discipline of the patient. Glucose-monitoring devices will not improve glycemic control if they are not used properly and with sufficient frequency. Nor will they improve illness course if the detection of abnormal values is not accompanied by adjustments in insulin delivery and/or relevant behaviors. Insulin pumps require a high degree of patient management to provide the desired result.
The degree of non-compliance with glucose-monitoring regimens in clinical studies to date is sobering. Substantial proportions of patients participating in both clinical trials and observational studies fail to use the devices at the prescribed frequency. These patterns of non-compliance are surprising and disappointing for two reasons. First, although patient education may be unpredictable in the broader medical community, investigators in clinical studies are typically rigorous about patient education. Thus, lack of knowledge is undoubtedly less of an issue in clinical studies than in routine medical practice. Second, patients who volunteer for clinical studies are typically more motivated to ensure that they receive state-of-the-art and cutting-edge treatments than patients who choose not to participate in clinical studies. In short, the results from controlled clinical studies are typically superior to those found in routine medical practice.
It is quite clear that self-monitoring devices and insulin pumps will not achieve their potential in the absence of high-quality patient education. But there may be a larger issue here. Results from clinical studies to date suggest that new forms of patient education may be needed. Thus, in addition to more rigorous investigations of the benefits of these devices, research also may be needed to identify more successful methods of patient education.
Looking to the Future
Devices and technological advances for the control and management of diabetes are emerging rapidly. As noted in the body of this review, inhaled insulin is very close to FDA approval and widespread availability. An unusually large number of clinical trials have been published on the effectiveness of inhaled insulin – unusual in the sense that such a large number of trials have been completed and published prior to FDA approval. Results of these trials are consistently quite positive with regard to the provision of insulin at appropriate doses. Even more impressive are the very high levels of patient satisfaction observed. Patient satisfaction is one dimension of quality of life, and it appears that the pain and inconvenience of other modes of insulin delivery are substantially reduced when insulin is inhaled. In short, although it was believed that insulin pumps would be more convenient for patients than multiple daily injections, it appears that inhaled insulin will greatly outdistance all previous methods of insulin delivery in terms of patient satisfaction.
The most promising direction in technological advances, however, will be devices that “close the loop” between glucose monitoring and insulin delivery. At this point it is unclear exactly what form future devices will take. But given the availability of reliable technologies for monitoring glucose levels and delivering insulin, development of technologies for combining these devices into a single, comprehensive monitoring and delivery system is a feasible goal – and one that will probably be achieved in the near future.
The potential of “closed loop” systems for monitoring glucose and delivering insulin will be a major advance in managing the purely medical dimensions of diabetes (e.g., glycemic control, avoiding hypoglycemic events). But these systems also will greatly improve the quality of life of diabetes patients. It also is hoped that they will solve much of the non-compliance problem with current devices. It is likely that the field is currently at the point where significant technological advances have been achieved, but they are less simple and user-friendly than conventional management strategies. Closed loop systems have the potential to achieve even better medical management than is currently possible, while simultaneously reducing burden on patients and improving their quality of life.
Adams AS, Mah C, Soumerai SB, Zhang F, Barton MB, Ross-Degnan D (2003). Barriers to self-monitoring of blood glucose among adults with diabetes in an HMO: a cross-sectional study. BMC Health Services Research 3:6: http://www.biomedcentral.com/1472-6963/3/6.
ADA – American Diabetes Association (2005a). http://www.diabetes.org
ADA – American Diabetes Association (2005b). Standards of medical care in diabetes. Diabetes Care 28 (Suppl. 1): S4-S36.
ADA – American Diabetes Association (2004). Five things you should know before pumping. Diabetes Forecast May 2004: http://www.diabetes.org/diabetes-forecast/may2004/pump.jsp.
ADA – American Diabetes Association (2003). Economic costs of diabetes in the US in 2002. Diabetes Care 26: 917-932.
Aussedat B, Dupire-Angel M, Gifford R, Klein JC, Wilson GS, Reach G (2000). Interstitial glucose concentration and glycemia: implications for continuous subcutaneous glucose monitoring. Am J Physiol Endocrinol Metab 278: #716- E728.
Bansod P, Shrivastava MC (2004). Comparative study of current approaches for minimally-invasive and non-invasive blood glucose monitoring. IETE Technical Review 21: 45-54.
Becker D, Kernesky K, Transue D, et al. (1984). Continuous subcutaneous insulin infusion (CSII) in adolescents with IDDM [abstract]. Int Study Group Diabetes Child Adolesc Bull 11: 10.
Bennion N, Christensen NK, McGarraugh G (2002). Alternative site glucose testing: a crossover design. Diabetes Technol Ther 4: 25-33.
Bergenstal RM, Gavin JR (2005). The role of self-monitoring of blood glucose in the care of people with diabetes: report of a global consensus conference. American Journal of Medicine 118: 1S-6S.
Berger B, Stenstrom G, Sundkvist G (1999). Incidence, prevalence, and mortality of diabetes in a large population: a report from the Skaraborg Diabetes Registry. Diabetes Care: 22: 773-778.
Blonde L, Karter AJ (2005). Current evidence regarding the value of self-monitored blood glucose testing. American Journal of Medicine 118: 20S-26S Suppl. 9A.
Boland E, Grey M, Oesterle A, et al. (1999). Continuous subcutaneous insulin infusion: a new way to lower risk of severe hypoglycemia, improve metabolic control, and enhance coping in adolescents with type 1 diabetes. Diabetes Care 22: 1779- 1784.
Boland E, Ahern J, Grey M (1998). A primer on the use of insulin pumps in adolescents. Diabetes Education 24: 78-86.
Brandle M, Zhou H, Smith BR et al. (2003). The direct medical cost of type 2 diabetes. Diabetes Care 26: 2300-2304.
Briggs AL, Cornell S (2004). Self-monitoring blood glucose (SMBG): now and the future. Journal of Pharmacy Practice 17: 29-38.
Broussole C, Jeandidier N, Hanaire-Broutin H (1994). The EVADIAC Study Group: French multicentre experience of implantable insulin pumps. The Lancet 343: 514-515.
Brown SA, Garcia AA, Kouzekanani K, Hanis CL (2002). Culturally competent diabetes self-management education for Mexican Americans: the Starr County border health initiative. Diabetes Care 25: 259-268.
Burt AL, Green S, Kwan I, Mugglestone M, Thomas J (2005). Intranasal insulin for type 1 diabetes mellitus. [Protocol] Cochrane Metabolic and Endocrine Disorders Group: Cochrane Database of Systematic Reviews 4, 2005. Accession Number: 00075320-100000000-04251.
Cappelleri JC, Cefalu WT, Rosenstock J, Kourides IA, Gerber RA (2002). Treatment satisfaction in type 2 diabetes: A comparison between an inhaled insulin regimen and a subcutaneous insulin regimen. Clinical Therapeutics 24: 552-564.
Caro JJ, Ward AJ, O’Brien JA (2002). Lifetime costs of complications resulting from type 2 diabetes in the U.S. Diabetes Care 25: 476-481.
Cefalu WT, Skyler JS, Kourides IA, Landschulz WH, Balagtas CC, Cheng SL, Gelfand RA (2001). Inhaled human insulin treatment in patients with type 2 diabetes mellitus. Annals of Internal Medicine 134: 203-207.
CDC – Centers for Disease Control (2005). National Diabetes Fact Sheet 2003: http://www.cdc.gov/diabetes/pubs/factsheet.htm
CDC Diabetes Cost-Effectivness Group (2002). Cost-effectiveness of intensive glycemic control, intensified hypertension control, and serum cholesterol level reduction in type 2 diabetes. JAMA 287: 2542-2551.
Chan WB, Chan JC, Chow CC, et al. (2000). Glycaemic control in type w diabetes: the impact of body weight, beta-cell function and patient education. QJM 93: 183- 190.
Colquitt JL, Green C, Sidhu MK, Hartwell D, Waugh N (2004). Clinical and costeffectiveness of continuous subcutaneous insulin infusion for diabetes. Health Technology Assessment Vol. 8: No. 43.
Cooper DS, DeAngelis CD (2001). Diabetes mellitus – a call for papers. JAMA 286: 968- 969.
Coster S, Gulliford MC, Seed PT, Powrie JK, Swaminathan R (2000). Self-monitoring in Type 2 diabetes mellitus: a meta-analysis. Diabetic Medicine 17: 755-761.
Davidson J (2005). Strategies for improving glycemic control: effective use of glucose monitoring. American Journal of Medicine 118: 27S-32S.
Davidson MB, Castellanos M, Kain D, Duran P (2005). The effect of self-monitoring of blood glucose concentrations on glycated hemoglobin levels in diabetic patients not taking insulin: A blinded, randomized trial. American Journal of Medicine 118: 422-425.
Davidson MB (2005). Counterpoint: Self-monitoring of blood glucose in type 2 diabetic patients not receiving insulin: A waste of money. Diabetes Care 28: 1531-1533.
Davidson MB, Castellanos M, Kain D (2004). Self-monitoring of blood glucose (SMBG) linked to nutritional counseling in minority type 2 diabetic patients on diet and oral antidiabetes medication does not improve glycemic outcomes. Diabetes 53 (Suppl.): A73.
De Beaufort CD, Bruning GJ (1987). Continuous subcutaneous insulin infusion in children. Diabetic Medicine 4: 103-108.
DeFronzo RA, Bergenstal RM, Cefalu WT, Pullman J, Lerman S, Bode BW, Phillips LS (2005). Efficacy of inhaled insulin in patients with type 2 diabetes not controlled with diet and exercise - A 12-week, randomized, comparative trial. Diabetes Care 28: 1922-1928.
DeWitt DE, Dugdale DC (2003). Using new insulin strategies in the outpatient treatment of diabetes: clinical applications. JAMA 289: 2265-2269.
DeWitt DE, Hirsch IB (2003). Outpatient insulin therapy in type 1 and type 2 diabetes mellitus. JAMA 289: 2254-2263.
DCCT (Diabetes Control and Complications Trial) Research Group (1996). Lifetime benefits and costs of intensive therapy as practiced in the Diabetes Control and Complications Trial. JAMA 276: 1409-1415.
DCCT (Diabetes Control and Complications Trial) Research Group (1993). The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. New England Journal of Medicine: 329: 977-986.
DCCT-EDIC (Diabetes Control and Complications Trial-Epidemiology of Diabetes Interventions and Complications) Research Group (2000). Retinopathy and nephropathy in patients with type 1 diabetes four years after a trial of intensive therapy. New England Journal of Medicine 342: 381-389.
Diabetes Research Working Group (1999). Conquering Diabetes: A Strategic Plan for the 21st Century. Bethesda, MD: National Institutes of Health. Publication No. 22- 4398.
Dixon D, Luce BR, Mauskopf JA, Paramore LC, Prasad M, Sloan F (2004). MEDTAP International Inc.: Value of Investment in Health Care: Better Care, Better Lives. White paper prepared for The Value Group, Washington, DC. Available at http://www.medtap.com/Products/policy.cfm.
Dupuy O, Mayaudon H, Palou M, Sarret D, Bordier L, Banduceau B (2000). Optimized transient insulin infusion in uncontrolled type 2 diabetes: evaluation of a pragmatic attitude. Diabetes and Metabolism 26: 371-375.
Eastman RC, Javitt JC, Herman WH et al. (1997a). Model of complications of NIDDM, I: model construction and assumptions. Diabetes Care 20: 725-734.
Eastman RC, Javitt JC, Herman WH, et al. (1997b). Model of complications of NIDDM, II: analysis of the health benefits and cost-effectiveness of treating NIDDM with the goal of normo-glycemia. Diabetes Care 20: 735-744.
Ellison JM, Stegmann JM, Colner SL, et al. (2002). Rapid changes in postprandial blood glucose produce concentration differences at finger, forearm, and thigh sampling sites. Diabetes Care 25: 961-964.
Farmer A, Gibson OJ, Tarassenko L, Neil A (2005). A systematic review of telemedicine interventions to support blood glucose self-monitoring in diabetes. Diabetic Medicine 22: 1372-1378.
Fineberg SE, Kawabata T, Finco-Kent D, Liu C, Krasner A (2005).Antibody response to inhaled insulin in patients with type 1 or type 2 diabetes. An analysis of initial phase II and III inhaled insulin (Exubera) trials and a two-year extension trial. Journal of Clinical Endocrinology and Metabolism 90: 3287-3294.
FDA – Food and Drug Administration (2005a). Glucose meters and diabetes management: http://www.fda.gov/diabetes/glucose.html.
FDA – Food and Drug Administration (2005b). Insulin: http://www.fda.gov/diabetes/insulin.html.
Franciosi M, Pellegrini F, De Berardis G, Belfiglio M, Di Nardo B, Greenfield S, Kaplan SH, Rossi MCE, Sacco M, Tognoni G, Valentini M, Nicolucci A (2005). Selfmonitoring of blood glucose in non-insulin-treated diabetic patients: a longitudinal evaluation of its impact on metabolic control. Diabetic Medicine 22: 900-906.
Freemantle N, Blonde L, Duhot D, Hompesch M, Eggertsen R, Hobbs FDR, Martinez L, Ross S, Bolinder B, Stridde E (2005). Availability of inhaled insulin promotes greater perceived acceptance of insulin therapy in patients with type 2 diabetes. Diabetes Care 28: 427-428.
Gallichan M (1994). Self-monitoring by patients receiving oral hypoglycaemic agents: a survey and a comparative trial. Practical Diabetes 11: 28-30.
Gatling W, Budd S, Walters D, Mullee MA, Goddard JR, Hill RD (1998). Evidence of an increasing prevalence of diagnosed diabetes mellitus in the Poole area from 1983 to 1996. Diabetic Medicine 15: 1015-1021.
Gerber RA, Cappelleri JC, Kourides IA, Gelfand RA (2001). Treatment satisfaction with inhaled insulin in patients with type 1 diabetes - A randomized controlled trial. Diabetes Care 24: 1556-1559.
Gilmer TP, O’Connor PJ, Manning WG, Rush WA (1997). The cost to health plans of poor glycemic control. Diabetes Care 20: 1847-1853.
Gray A, Raikou M, McGuire A, Fenn P, Stevens R, Cull C, et al. (2000). Cost effectiveness of an intensive blood glucose control policy in patients with type 2 diabetes: economic analysis alongside randomized controlled trial (UKPDS 41). United Kingdom Prospective Diabetes Study Group. British Medical Journal 320: 1373-1378.
Guerci B, Droulin P, Grangé V, Bougnères P, Fontaine P, Kerlan V, Passa P, Thivolet CH, Vialettes B, Charbonnel B (2003). Self-monitoring of blood glucose significantly improves metabolic control in patients with type 2 diabetes mellitus: the Auto-Surveillance Intervention Active (ASIA) study. Diabetes and Metabolism 29: 587-594.
Hanaire-Broutin H, Broussole C, Jeandidier N. et al. (1995). The EVADIAC Study Group: Feasibility of intraperitoneal insulin therapy with programmable implantable pumps in IDDM. Diabetes Care 18: 388-392.
Harris MI (2001). Frequency of blood glucose monitoring in relation to glycemic control in patients with type 2 diabetes. Diabetes Care 24: 979-982.
Harris MI, Eastman RC (2000). Early detection of undiagnosed diabetes mellitus: a U.S. perspective. Diabetes and Metabolism Research and Reviews 16: 230-236.
Heise T, Bott S, Tusek C, Stephan JA, Kawabata T, Finco-Kent D, Liu C, Krasner A (2005).The effect of insulin antibodies on the metabolic action of inhaled and subcutaneous insulin - A prospective randomized pharmacodynamic study. Diabetes Care 28: 2161-2169.
Henry RR, Mudaliar S, Chu N, Kim D, Armstrong D, Davis TT, An B, Reinhardt RR (2003). Young and elderly type 2 diabetic patients inhaling insulin with the AERx (R) Insulin Diabetes Management System: A pharmacokinetic and pharmacodynamic comparison. Journal of Clinical Pharmacology 43: 1228-1234.
Herman WH, Ilag LL, Johnson SL, Martin CL, Sinding J, Harthi AA, Plunkett CD, LaPorte FB, Burke R, Brown MB, Halter JB, Raskin P (2005). A clinical trial of continuous subcutaneous insulin infusion versus multiple daily injections in older adults with type 2 diabetes. Diabetes Care 28: 1568-1573).
Hermansen K, Ronnemaa T, Petersen AH, Bellaire S, Adamson U (2004). Intensive therapy with inhaled insulin via the AERx insulin diabetes management system. Diabetes Care 27: 162-167.
Hoffman RM, Shah JH, Wendel CS, Duckworth WC, Adam KD, Bokhari SU, Dalton C, Murata GH (2002). Evaluating once- and twice-daily self-monitored blood glucose testing strategies for stable insulin-treated patients with type 2 diabetes: the Diabetes Outcomes in Veterans Study. Diabetes Care 25: 1744-1748.
Hogan P, Dall T, Nikolov P (2003). American Diabetes Association: economic costs of diabetes mellitus in the U.S. in 2002. Diabetes Care 26: 917-932.
Hollander PA, Blonde L, Rowe R, Mehta AE, Milburn JL, Hershon KS, Chiasson JL, Levin SR (2004). Efficacy and safety of inhaled insulin (Exubera) compared with subcutaneous insulin therapy in patients with type 2 diabetes - Results of a 6- month, randomized, comparative trial. Diabetes Care 27: 2356-2362.
IDF – International Diabetes Federation. Facts and Figures: http://www.idf.org/home/index.cfm?node=37.
Ipp E, Aquino RL, Christenson P (2005). Point: Self-monitoring of blood glucose in type 2 diabetic patients not receiving insulin - the sanguine approach. Diabetes Care 28: 1528-1530.
Jeandidier N, Selam JL, Renard E, et al. (1996). The EVADIAC Study Group: Decreased severe hypoglycemia frequency during intraperitoneal insulin infusion using programmable implantatble pumps. Diabetes Care 19: 780.
Johansson U-B, Adamson U, Lins P-E, Wredling R (2004). Patient management of longterm continuous subcutaneous insulin infusion. Journal of Advanced Nursing 51: 112-118.
Jungheim K, Koschinsky T (2001). Risky delay of hypoglycemia detection by glucose testing at the arm. Diabetes Care 24: 1303-1304.
Karter AJ, Stevens MR, Herman WH, et al. (2003). Out-of-pocket costs and diabetes preventive services: the Translating Research into Action for Diabetes (TRIAD) study. Diabetes Care 26: 2294-2299.
Karter AJ, Ackerson LM, Darbinian JA, et al. (2001). Self-monitoring of blood glucose levels and glycemic control: the Northern California Kaiser Permanente Diabetes Registry. American Journal of Medicine 111: 1-9.
Knight G, Boulton AJM, Ward JD (1986). Experience of subcutaneous insulin infusion in the outpatient management of diabetic teenagers. Diabetic Medicine 3: 82-84.
Koro CE, Bowlin SJ, Bourgeois N, Fedder DO (2004). Glycemic control from 1988 to 2000 among U.S. adults diagnosed with type 2 diabetes: a preliminary report. Diabetes Care 27: 17-20.
Kovatchev BP, Gonder-Frederick LA, Cox DJ, Clarke WL (2004). Evaluating the accuracy of continuous glucose-monitoring sensors. Diabetes Care 27: 1922- 1928.
Kwon HS, Cho JH, Kim HS, et al. (2004). Establishment of blood glucose monitoring system using the Internet. Diabetes Care 27: 478-483.
Lawson ML, Gerstein HC, Tsui E, Zinman B (1999). Effect of intensive therapy on early macrovascular disease in young individuals with type 1 diabetes: a systematic review and meta-analysis. Diabetes Care 22 (Suppl. 2): B35-B39.
Lawton J, Peel E, Douglas M, Perry O (2004). 'Urine testing is a waste of time': newly diagnosed Type 2 diabetes patients' perceptions of self-monitoring. Diabetic Medicine 21: 1045-1048.
Lenhard MJ, Reeves GD (2001). Continuous subcutaneous insulin infusion: a comprehensive review of insulin pump therapy. Archives of Internal Medicine 161: 2293-2300.
Lock JP, Szuts EZ, Malomo KJ, Anagnostopoulos A (2002). Whole-blood glucose testing at alternative sites. Diabetes Care 25: 337-341.
Mäkimattila S, Nikkilä K, Yki-Järvinen H (1999). Causes of weight gain during insulin therapy with and without metformin in pateients with Type II diabetes mellitus. Diabetologia 42: 406-412.
McAuley L (2001). Inhaled insulin for the treatment of diabetes mellitus. The Canadian Coordinating Office for Health Technology Assessment (CCOHTA): Issues in Emerging Health Technologies 18: June 2001. McGarraugh G, Price D, Schwartz S, Weinstein R (2001). Physiological influences on off-finger glucose testing. Diabetes Technol Ther 3: 367-373.
McGarraugh G (2001). Response to Jungheim and Koschinsky. Diabetes Care 24: 1304- 1305.
Meier JL, Swislocki AL, Lopez JR, Noth RH, Bartlebaugh P, Siegel D (2002). Reduction in self-monitoring of blood glucose in persons with type 2 diabetes results in cost saving and no change in glycemic control. American Journal of Managed Care 8: 557-565.
Miles P, Everett J, Murphy J, Kerr D (1997). Comparison of blood or urine testing by patients with newly diagnosed non-insulin dependent diabetes: patient survey after randomized cross-over trial. British Medical Journal 315: 348-349.
Misso ML, O'Connor DA, Egberts KJ, Shaw J (2005). Continuous subcutaneous insulin infusion (CSII) versus multiple insulin injections for type 1 diabetes mellitus. [Protocol] Cochrane Metabolic and Endocrine Disorders Group: Cochrane Database of Systematic Reviews 4, 2005. Accession Number: 00075320- 100000000-04078.
Murata GH, Duckworth WC, Shah JH, Wendel CS, Mohler MJ, Hoffman RM (2005). Hypoglycemia in stable, insulin-treated veterans with type 2 diabetes: a prospective study of 1662 episodes. Journal of Diabetes and Its Complications 19: 10-17.
Murata GH, Duckworth WC, Hoffman RM, Wendel CS, Mohler MJ, Shah JH (2004a). Hypoglycemia in type 2 diabetes: a critical review. Biomedicine and Pharmacotherapy 58: 551-559.
Murata GH, Shah JH, Duckworth WC, Wendel CS, Mohler MJ, Hoffman RM (2004b). Food frequency questionnaire results correlate with metabolic control in insulintreated veterans with type 2 diabetes: the Diabetes Outcomes in Veterans Study. Journal of The American Dietetic Association 104: 1816-1826.
Murata GH, Hoffman RM, Shah JH, Wendel CS, Duckworth C (2004c). A probabilistic model for predicting hypoglycemia in type 2 diabetes mellitus. The Diabetes Outcomes in Veterans Study (DOVES). Archives of Internal Medicine 164: 1445- 1450.
Murata GH, Duckworth WC, Shah JH, Wendel CS, Hoffman RM (2004d). Factors affecting hypoglycemia awareness in insulin-treated type 2 diabetes: The Diabetes Outcomes in Veterans Study (DOVES). Diabetes Research and Clinical Practice 65: 61-67.
Murata GH, Hoffman RM, Duckworth WC, Wendel CS, Shah JH (2004e). Contributions of weekly mean blood glucose values to hemoglobin A1C in insulin-treated type 2 diabetes: the Diabetes Outcomes in Veterans Study (DOVES). American Journal of The Medical Sciences 327: 319-323.
Murata GH, Duckworth WC, Shah JH, Wendel CS, Hoffman RM (2004f). Sources of glucose variability in insulin-treated type 2 diabetes: the Diabetes Outcomes in Veterans Study (DOVES). Clinical Endocrinology 60: 451-456.
Murata GH, Shah JH, Adam KD, Wendel CS, Bokhari SU, Solvas PA, Hoffman RM, Duckworth WC (2003a). Factors affecting diabetes knowledge in Type 2 diabetic veterans. Diabetologia 46(8): 1170-1178.
Murata GH, Shah JH, Hoffman RM, Wendel CS, Adam KD, Solvas PA, Bokhari SU, Duckworth WC (2003b). Intensified blood glucose monitoring improves glycemic control in stable,insulin-treated veterans with type 2 diabetes: the Diabetes Outcomes in Veterans Study (DOVES). Diabetes Care 26: 1759-1763.
Murata GH, Shah JH, Wendel CS, Hoffman RM, Adam KD, Bokhari SU, Solvas PA, Duckworth WC (2003c). Risk factor management in stable, insulin-treated patients with type 2 diabeles: the Diabetes Outcomes in Veterans Study. Journal of Diabetes and its Complications 17: 186-191.
Nathan DM (2002). Initial management of glycemia in type 2 diabetes mellitus. N Engl J Med 347: 1342-1349.
NDIC – National Diabetes Clearinghouse (2005): Diabetes in America, 2nd Edition. http://diabetes.niddk.nih.gov/ dm/pubs/America/contents.htm NDEP – National Diabetes Education Program (2005). http://www.ndep.nih.gov National Guideline Clearinghouse (2005). Best practice guideline for subcutaneous administration of insulin in adults with type 2 diabetes: http://www.guideline.gov/summary/summary.aspx?ss=15&doc_id=5736&nbr=& string= Nettles A (1993). User error in blood glucose monitoring. Diabetes Care 16: 946-948.
NIDDKD – National Institute of Diabetes and Digestive and Kidney Diseases (2005). http://www.niddk.nih.gov/ Olefsky JM (2001). Prospects for research in diabetes mellitus. JAMA 285: 628-632.
Oesterle A, Boland E, Yu C., et al. (1998). CSII: a new way to achieve strict metabolic control and lower the risk of sever hypoglycemia in adolescents [abstract]. Diabetes 21 (Suppl 1): 1320A.
Owens D, Barnett AH, Pickup J, et al. (2004). Blood glucose self-monitoring in type 1 and type 2 diabetics: reaching a multidisciplinary consensus. Diabetes Primary Care 6: 8-16). Peel E, Parry O, Douglas M, Lawton J (2004). Blood glucose self-monitoring in noninsulin- treated type 2 diabetes: a qualitative study of patients’ perspectives. British Journal of General Practice 54: 183-188.
Pickup J, Mattock M, Kerry S (2002). Glycaemic control with continuous subcutaneous insulin infusion compared with intensive insulin injections in patients with type 1 diabetes: meta-analysis of randomised controlled trials. British Medical Journal 324: 705-708.
Quattrin T, Belanger A, Bohannon NJV, Schwartz SL (2004). Efficacy and safety of inhaled insulin (exubera) compared with subcutaneous insulin therapy in patients with type 1 diabetes - Results of a 6-month, randomized, comparative trial. Diabetes Care 27: 2622-2627.
Radermecker RP, Scheen AJ (2004). Continuous subcutaneous insulin infusion with short-acting insulin analogues or human regular insulin: efficacy, safety, quality of life, and cost-effectiveness. Diabetes Metabolism Research and Reviews 20: 178-188.
Retnakaran R, Hochman J, DeVries JH, Hanaire-Broutin H, Heine RJ, Melki V, Zinman B (2004). Continuous subcutaneous insulin infusion versus multiple daily injections. Diabetes Care 27: 2590-2596.
Rosenstock J, Cappelleri JC, Bolinder B, Gerber RA (2004). Patient satisfaction and glycemic control after 1 year with inhaled insulin (exubera) patients with type 1 or type 2 diabetes. Diabetes Care 27: 1318-1323.
Royle P, Waugh N, McAuley L, McIntyre L, Thomas, S (2005). Inhaled insulin in diabetes mellitus. [Systematic Review] Cochrane Metabolic and Endocrine Disorders Group: Cochrane Database of Systematic Reviews 4, 2005: Accession Number 00075320-100000000-02896.
Rudolf MC, Ahern J, Genel M, et al. (1982). Optimal insulin delivery in adolescents with diabetes: impact of intensive treatment on psychosocial adjustment. Diabetes Care 5 (Suppl 1): 53-57.
Sarol JN Jr., Nicodemus NA Jr., Tan KM, Grava MB (2005). Self-monitoring of blood glucose as part of a multi-component therapy among non-insulin requiring type 2 diabetes patients: a meta-analysis (1966-2004). Current Medical Research and Opinions 21: 173-183.
Saudek CD (2002). Progress and promise of diabetes research. JAMA 287: 2582-2584.
Schaepelynck-Bélicar P, Dufaitre-Patouraux L, Lassmann-Vague V (2005). What could be the reasons for giving up the implanted pump treatment? Diabetes Metabolism 31: 87-89.
Schiffrin AD, Desrosiers M, Aleyossine H, et al. (1984). Intensified insulin therapy in the type 1 diabetic adolescent: a controlled trial. Diabetes Care 7: 107-113.
Schwedes U, Siebolds M, Mertes G – the SMBG Study Group (2002). Meal-related structured self-monitoring of blood glucose effect on diabetes control in noninsulin- treated type 2 diabetic patients. Diabetes Care 25: 1928-1932.
Scuffham P, Carr L (2003). The cost-effectiveness of continuous subcutaneous insulin infusion compared with multiple daily injections for the management of diabetes. Diabetic Medicine 20: 586-593.
Skyler JS, Weinstock RS, Raskin P, Yale JF, Barrett E, Gerich JE, Gerstein HC (2005). Use of inhaled insulin in a basal/bolus insulin regimen in type 1 diabetic subjects - A 16-month, randomized, comparative trial. Diabetes Care 28: 1630-1635.
Skyler JS, Cefalu WT, Kourides IA, Landschulz WH, Balagtas CC, Cheng SL, Gelfand RA (2001). Efficacy of inhaled human insulin in type 1 diabetes mellitus: a randomised proof-of-concept study. The Lancet 357: 331-335.
Steil GM, Panteleon AE, Rebrin K (2004). Closed-loop insulin delivery – the path to physiological glucose control. Advanced Drug Delivery Reviews 56: 125-144.
Steindel BS, Roe TR, Costin G, et al. (1995). Continuous subcutaneous insulin infusion (CSII) in children and adolescents with poorly controlled type 1 diabetes. Diabetes Research and Clinical Practice 27: 199-204.
Stratton IM, Adler AI, Neil HA, Matthews DR, Manley SE, Cull CA, Hadden D, Turner RC, Holmman RR (2000). Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS 35): prospective observational study. British Medical Journal 321: 405-412.
Tamada JA, Garg S, Jovanovic L, Pitzer KR, Fermi S, Potts RO, Cygnus Research Team (1999). Noninvasive glucose monitoring: comprehensive clinical results. JAMA 282: 1839-1844.
Tamboriane WV, Batas SE, Rudolf MC, et al. (1989). Comparison of continuous subcutaneous insulin infusion versus multiple daily injections in adolescents with insulin-dependent diabetes. Adv Diabetologia 2 (Suppl 1): 24-27.
Tanenberg R, Bode B, Lane W, Levetan C, Mestman J, Harmel AP, Tobian J, Gross T, Mastrototaro J (2004). Use of the continuous glucose monitoring system to guide therapy in patients with insulin-treated diabetes: A randomized controlled trial. Mayo Clinic Proceedings 79: 1521-1526.
Testa MA, Simonson DC (1998). Health economic benefits and quality of life during improved glycemic control in patients with type 2 diabetes mellitus: a randomized, controlled, double-blind trial. JAMA 280: 1490-1496.
Trubo R (2005). Interest in inhaled insulin grows. JAMA 294: 1195-1196.
Tuffnell DJ, West J, Farrar D (2005). Continuous subcutaneous insulin infusion versus multiple daily injections of insulin for pregnant women with diabetes. [Protocol] Cochrane Pregnancy and Childbirth Group: Cochrane Database of Systematic Reviews 4, 2005. Accession Number: 00075320-100000000-04534.
Ubink-Veltmaat LJ, Bilo HJ, Groenier KH, Houweling ST, Rischen RO, Meyboom-de Jong B (2003). Prevalence, incidence and mortality of type 2 diabetes revisited: a prospective population-based study in the Netherlands (ZODIAC-1). European Journal of Epidemiology 18: 793-800.
UKPDS (UK Prospective Diabetes Study) Group (1998a). Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). The Lancet 352: 837-853.
UKPDS (UK Prospective Diabetes Study) Group (1998b). Effect of intensive bloodglucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). The Lancet 352: 854-865.
Valensi P, Moura I, Le Magoarou M, Paries J, Perret G, Attali JR (1997). Optimized transient insulin infusion in uncontrolled type 2 diabetes: evaluation of a pragmatic attitude. Diabetes and Metabolism 23: 51-57.
VHA – Veterans Health Administration (2005). Veterans Health Administration Diabetes Program: http://www1.va.gov/health/diabetes/ Wagner EH, Sandhu N, Newton KM, McCulloch DK, Ramsey SD, Grothaus LC (2001). Effect of improved glycemic control on health care costs and utilization. JAMA 285: 182-189.
Wainstein J, Metzger M, Boaz M, Minuchin O, Cohen Y, Yaffe A, Yerushalmy Y, Raz I, Harman-Boehm I (2005). Insulin pump therapy vs. multiple daily injections in obese type 2 diabetic patients. Diabetic Medicine 22: 1037-1046.
Wake N, Hisashige A, Katayama T, et al. (2000). Cost-effectiveness of intensive insulin therapy for type 2 diabetes. Diabetes Research and Clinical Practice 48: 201-210.
Weintrob N, Benzaquen H, Galatzer A, Shalitin S, Lazar L, Fayman G, Lilos P, Dickerman Z, Phillip M (2003). Comparison of continuous subcutaneous insulin infusion and multiple daily injection regimens in children with type 1 diabetes: a randomized crossover trial. Pediatrics 112: 559-564.
Weiss SR, Cheng SL, Kourides IA, Gelfand RA, Landschulz WH (2003). Inhaled insulin provides improved glycemic control in patients with type 2 diabetes mellitus inadequately controlled with oral agents - A randomized controlled trial. Archives of Internal Medicine 163: 2277-2282.
Weissberg-Benchell J, Antisdel-Lomaglio J, Seshadri R (2003). Insulin pump therapy. Diabetes Care 26: 1079-1087.
Welschen LMC, Bloemendal E, Nijpels G, Dekker JM, Heine RJ, Stalman WAB, Bouter LM (2005). Self-monitoring of blood glucose in patients with type 2 diabetes who are not using insulin. The Cochrane Database of Systematic Reviews 2005, Issue 2. Art. No. CD005060.pub2. DOI: 10.1002/14651858.CD005060.pub2: http://www.cochrane.org/cochrane/revabstr/AB005060.htm.
Wen L, Parchman ML, Linn WD, Lee S (2004). Association between self-monitoring of blood glucose and glycemic control in patients with type 2 diabetes mellitus. American Journal of Health-System Pharmacy 61: 2401-2405.
Wilson GS, Gifford Raeann (2005). Biosensors for real-time in vivo measurements. Biosensors and Bioelectronics 20: 2388-2403.
Yki-Järvinen H, Ryysy L, Nikkilä K, Tulokas T, Vanamo R, Heikkilä M (1999). Comparison of bedtime insulin regimens in patients with type 2 diabetes mellitus – a randomized controlled trial. Annals of Internal Medicine 130: 389-396.