I'm going to examine some of the evidence for the relationship between self-monitoring of blood glucose (SMBG) and glycemic control.

We've heard from Dr. Gavin about the epidemic of diabetes; this was an 11-year-old girl with type 2 diabetes who appeared on the cover of Time magazine. About the time that this appeared, there was a paper in the Journal of the American Medical Association (JAMA) that estimated that for people born in the year 2000, 33% of men, 39% of women, and 53% of Latino-Hispanic women would develop diabetes in their lifetime if we don't change something.

Even today, diabetes costs a fortune; $132 billion, mostly from the complications Dr. Gavin talked about.

We're all familiar with the goals for glycemia, blood pressure, and lipid control. Dr. Gavin reviewed these.

He showed the relationship between lowering hemoglobin A1C (A1C) and lowering complications, both micro- and macrovascular.

Dr. Gavin also showed that, in fact, we're not doing very well in terms of getting people to goal; most recently, only 37% of diabetic patients attained goal glycemia.

Recently, the American Association of Clinical Endocrinologists (AACE) had their annual meeting. Dr. Davidson chaired a task force that released a press release, and featured at this press conference was Surgeon General Carmona and Della Reese. What we announced was the state of diabetes health. This was a survey of 157,000 people looking at laboratory results across the country, and it was found that, using the American College of Endocrinology (ACE) goal, two thirds of people were above the recommended goal of less than 6.5%. And in every state tested, more than half the people were above the goal. Dr. Gavin showed you data from the year 2000; these are data from the year 2005, and we're not seeing much progress in the right direction.

Why don't people get to goal? We have already heard, and will continue to hear, a lot about new therapies. That's terrific, and they will be very beneficial, but we should be able to do much better with current therapies than what we are achieving. We know there's a failure of many physicians to adopt a treat-to-target approach; that there's suboptimal adherence of patients to lifestyle and pharmacologic treatments, and there is a lack of optimal systems of care that play a role.

This is from the Kaiser database. They looked at the length of time between the first monotherapy A1C value of greater than 8%, and when somebody did something about it, and it was 15 to 21 months. So there's this incredible lag and this failure of adopting a treat-to-target approach.

When you talk about adherence of patients -- and Dr. Gavin mentioned the increased adiposity -- certainly we're becoming less fit. Although there's been a lot of attention to the fast food industry about this, and they have responded (in fact, virtually all fast food chains are trying to offer healthier food), it isn't just fast food; it isn't just what we eat. I've often said that I think Americans have raised the avoidance of exercise to an art form in our society.

So for those of us who want to improve the organization of diabetes care, the National Diabetes Education Program (NDEP) has a Web site called betterdiabetescare.nih.gov that is designed to help people structure better, more organized diabetes care.

For all of us healthcare professionals, going to sites such as the NDEP and the American Diabetes Association (ADA), we can download the guidelines, and we also can have access to a lot of tools for our patients.

At the ADA site is a Visit Planning Tool on which people can write down what their goals are, track their results, and bring them to the healthcare professional when they have their visits.

The NDEP also has a number of resources that can help people better adhere to lifestyle measures; there's a GAMEPLAN tool kit and risk assessment tool, and a food and activity tracker.

There are a number of patient information resources and culturally specific messages to help people better adhere to therapy.

One of my favorites is from the Hispanic/Latino work group of the NDEP, Prevengamos La Diabetes Tipo 2 Paso a Paso (Prevent Type 2 Diabetes Step by Step). There are a lot of resources that we can use to really help our patients.

Another resource that we can use is to encourage our patients to use self-monitoring of blood glucose (SMBG), and for us to help them respond appropriately to the values they get. SMBG can help patients and their healthcare professionals better adjust therapy and assess the responses to therapy. It can help healthcare professionals implement a treat-to-target approach. It can help people better adhere to treatment by showing them the responses they are having to their treatment.

We know that in terms of monitoring glycemic control, A1C provides an assessment of time-averaged glycemic control over the previous 2 to 3 months; it's a standard method of guiding management decisions, and has become a surrogate for long-term complications. But SMBG provides the real-time feedback on glycemic response to diet, activity, and medications, and identifies hypoglycemia and postprandial hyperglycemic excursions.

Obviously, the advantages of A1C is this correlation with mean plasma glucose (MPG) over the previous 2 to 3 months, and one sees here this very strong correlation between mean A1C on the x-axis, and MPG on the y-axis. This has been considered a gold standard for long-time glycemic control, and a surrogate for the risk of complications.

The limitations are that this is a time-averaged value, so you don't actually see the individual glycemic excursions. The excursions are very common, even with A1C values that are not markedly elevated. In this small study of 25 patients with type 2 diabetes, all of whom had A1C values of less than 7.5%, you can see these rather dramatic excursions in glycemia; so this is made up of a lot of highs and lows. And A1C doesn't provide immediate feedback about glycemia and its response to various perturbations such as diet, exercise, or medications. Indeed, one can have the same A1C -- in this case, 7.5% -- with values in which there's minimum fluctuation, or one can have the same A1C with a lot of real highs and a lot of real lows.

There's increasing evidence that it isn't just the level of A1C, but the fluctuation itself that may have an impact on both complications, and how people feel and function. In this study that Testa presented at ADA 2 years ago, she showed that the day-to-day variation in glucose predicted quality of life in addition to the effect of high glucose values. And the only way to really look at day-to-day variation is with SMBG.

In the Verona Diabetes Complications Study, if one looks at tertiles of the mean fasting plasma glucose level, survival was less with the highest level of A1C. But look at the tremendous impact of the greatest coefficient of variation; variation actually was a greater predictor than just glucose value itself, at least in this trial.

What about SMBG? The real advantages are that real-time feedback can detect glycemic excursions, and it can be used as an educational and motivational tool. The limitations are the inconvenience, discomfort of a fingerstick, cost of supplies, and the requirement for training and education of patients and healthcare professionals about appropriate analysis and use of data.

SMBG is an integral part of insulin therapy in type 1 and type 2 patients; Dr. Gavin showed you data from the Diabetes Control and Complications Trial (DCCT), and the Kumamoto trial.

There is little debate about the role of SMBG in people who are taking insulin, particularly multiple daily injections of insulin. In this study from Dr. Raskin's group a number of years ago, they showed a clear correlation between increasing frequency of SMBG in type 1 diabetic patients and the A1C levels; as frequency went up, the A1C level went down.

Where there has been a more variable opinion is related to the role of SMBG in patients who have type 2 diabetes and are not treated with insulin. There are a number of studies, and I'm going to review briefly a few that have attempted to look at this. One of the issues is that there have been a large number of different kinds of trials that have been used, and there have been some methodologic issues with some of the trials as well.

In this longitudinal study from the Kaiser group, they looked at over 24,000 adult patients with diabetes in a large, group model managed-care organization. They showed that there was a relationship between SMBG and A1C in type 1 patients (if they did it 3 or more times per day), and pharmacologically-treated type 2 patients, no matter what the pharmacologic treatment was. If they performed SMBG at least once a day, this was associated with a lower A1C than with less frequent monitoring. So type 1 patients who monitored 3 or more times per day had a 1% lower A1C than those who monitored less frequently or didn't monitor. Type 2 patients who monitored 1 time a day or more had a 0.6% lower A1C than those monitoring less. Even in this study, nonpharmacologically treated type 2 patients who practiced SMBG at any frequency had a 0.4% lower A1C level than those not practicing at all. Dr. Karter, who's responsible for these data, presented some data at the European Association for the Study of Diabetes (EASD) last year from a longitudinal study that starts to show the same sort of evidence.

Other studies that show benefit -- in this study, almost 1000 diabetic patients not treated with insulin or previously monitored were randomized to standard care or standard care plus self-monitoring. SMBG was asked to be performed 6 times per week on 3 different days, and this was associated with a significantly lower A1C in the SMBG group compared with the conventional group at 6 months. In the SMBG group, there was a reduction of 0.9% vs a reduction in the control group of 0.5% that was statistically significant.

Recently, Davidson in the American Journal of Medicine reported a study that was a single-blind, randomized study of type 2 patients not taking insulin. Notice this was a much smaller group of patients, and monitoring was asked to be performed 6 times per week on 6 different days in the intervention group compared with the conventional group. But the people who were -- as I understand the paper -- responsible for adjusting therapy were blinded to the SMBG results. Both groups, interestingly, had a rather significant reduction in A1C. The monitoring group actually did better; it had a greater reduction, but the difference between the two was not statistically significant. I guess that's not surprising to me in that I would anticipate it would have taken a lot more patients to be able to show a statistically significant reduction in this study. Also, adherence to monitoring appeared to have been a problem in this study as well.

In contrast, an interesting study by Kwon and colleagues involved 110 type 2 patients in which the treatment arm and the control group had usual care, monthly visits; both groups were asked to perform SMBG, but only one group was followed. Their results were put on the Internet, and they interacted with their healthcare professionals on the Internet and got feedback about what to do about their SMBG values based on this Internet reporting. Both groups were told to do similar frequency of monitoring, but the actual frequency was significantly greater in the group that was getting regular feedback (which I think is one of the messages from this study), and there was a greater reduction in A1C in the intervention group. So I think one of the things that's clear is that one has to respond to the SMBG values to get a benefit.

A meta-analysis from Sarol of 8 randomized, controlled clinical trials comparing A1C with and without SMBG in non-insulin-treated patients has been published. They found 14 randomized, controlled trials that potentially could be included; they excluded a number of these and they gave specific reasons, such as absence of extractable data for non-insulin-treated patients only, lack of usable outcome data, and noncomparability between the 2 groups. They ended up with 8 studies that they then put into this meta-analysis.

In the 8 studies in the meta-analysis, they found that the addition of SMBG was associated with a 0.42% greater reduction in A1C compared with the group that did not perform SMBG.

As you're probably aware, in a recent issue of Diabetes Care was another meta-analysis by Welschen and colleagues that looked at 5 studies, including the Davidson study that I just showed you.

And they found, very similar to the Sarol data, a 0.39% greater A1C reduction in the SMBG group compared with the group that was not randomized to SMBG. Again, these are non-insulin-using subjects. They did indicate that there were some methodologic issues, even with some of these studies, and of course that's been an issue.

In the same issue, there was a point/counterpoint. Ipp and colleagues addressed the point of SMBG in type 2 diabetic patients not receiving insulin; their recommendation for doing it.

And Dr. Davidson wrote a counterpoint saying that he felt that this was often not worth doing.

There was actually quite a bit of concordance among all 3 papers in some of the things that they said. What I took away from it was that if one looks at the studies in which we have not been able to see a reported benefit, they are often underpowered, and where you wouldn't expect, necessarily, to be able to see a benefit. To impact control, patients and healthcare professionals have to assess SMBG values and act on the data. Patients need education about SMBG and self-management training to respond appropriately to values, and healthcare professionals must endorse the importance by actually reviewing the data with patients.

Here are data from Schwedes and colleagues. Patients were randomized into either monitoring or not monitoring, but they were given a specific counseling algorithm addressing what to do with the values. Again, these were non-insulin-requiring patients, and at the end of 24 weeks, there was about a 0.4% greater reduction in A1C with monitoring.

Obviously, people who have hypoglycemia, particularly people who have hypoglycemic unawareness, which can occur particularly in people with type 1 diabetes, must monitor more frequently. One of the key roles for monitoring is the ability to identify, treat, and subsequently appropriately avoid hypoglycemia.

In summary, achieving good glycemic control requires accurate assessment of blood glucose values. A1C and SMBG are clearly -- and I think we all know this -- complementary methods for evaluating glycemic control. A1C reflects mean glucose levels over the long term, and is the standard for adjusting therapy. SMBG provides patients and healthcare professionals with real-time feedback on daily glycemic variation.

Apart from clinical trials, I believe that as we treat patients, SMBG can have an incredible impact. For instance, in starting someone on a pharmacologic therapy, whether one starts with metformin or a sulfonylurea or with a thiazolidinedione (TZD), one doesn't need to wait 3 months to see the reflection of a therapy or therapy dose on the A1C; one can make changes based on SMBG, and most of the studies haven't really commented or looked at this.

In addition, patients can get tremendous feedback. I have patients all the time who say that they find that certain foods make their blood sugars go up more than other foods; certain combinations of food. The only way they can learn that is actually seeing the impact of various meals on pre- and postprandial glycemic control. Dr. Gerich is going to talk about postprandial hyperglycemia, a subject that I'm very interested in because I live in New Orleans, Louisiana, where people are continuously in the postprandial phase of glycemic control!

In the last part of this presentation, I thought I would review some recent recommendations for frequency of SMBG in type 2 patients.

The ADA at the present time, as I look at their present recommendations and the standards of care, say that for type 1 patients, pregnant women taking insulin, and all patients using multiple insulin injections, SMBG is recommended 3 or more times a day. For patients using less frequent insulin injections, oral agents, or medical nutrition therapy alone, SMBG is useful in achieving glycemic goals. Patients with type 2 diabetes on insulin typically will need to perform SMBG more frequently than those not using insulin; for type 2 diabetes patients, SMBG should be sufficient to facilitate reaching glycemic goals. And when adding to or modifying treatment, type 1 and type 2 diabetic patients should test more often than usual. To achieve postprandial glucose targets, postprandial SMBG may well be appropriate, although there haven't been a lot of studies that have specifically targeted this intervention. I personally find it very hard to object to anything in these recommendations, although I'd like to see a little more direct guidance about specific frequency in specific patient types.

The American Academy of Family Physicians published a monograph that was reported a couple of years ago, and their recommendations on frequency were that type 2 diabetes patients using multiple daily injections of insulin should perform SMBG as often as those who have type 1 diabetes, at least 3 times a day. Many type 2 patients taking oral antidiabetic agents who have not achieved their A1C goal may require SMBG multiple times per day -- 2 to 4. So the first of these is just like the ADA; the second doesn't really disagree but gives a more concrete frequency recommendation.

All patients who have diabetes should own a glucose meter and know how to use it. Patients whose diabetes is not well controlled should test multiple times per day for several days to produce sufficient data for clinical decision making.

Dr. Davidson chaired an implementation conference for the outpatient management of diabetes, conducted by the American Association of Clinical Endocrinologists and the American College of Endocrinology (ACE).

As part of their recommendations, they said that, when performed with sufficient frequency, SMBG readings allow patients and their healthcare professionals to make informed decisions about lifestyle choices and adjustments in pharmacologic therapy. SMBG can also provide ongoing feedback to patients about their nutrition and physical activity. It's a very important educational tool, and we should use SMBG to support and advise therapeutic decisions and enhance patient education. There are going to be expanded recommendations, I believe, and they may well be associated with even more specific recommendations about SMBG frequency.

There also was a global consensus conference conducted by the International Diabetes Center and the World Health Organization, "Improving Tight Control in the Patient with Diabetes: A Reappraisal of SMBG." This was held in the fall of 2004.

Dr. Gavin was one of the co-chairs of this conference, and this really was an international conference. Dr. Gerich and I were also participants, and Dr. Davidson wrote one of the position papers.

There were 4 position papers drafted and distributed to all participants prior to the meeting. There were 2 unpublished studies presented during the meeting, and then recommendations were made after reviewing the above.

At present, the recommendations are that SMBG frequency for those on multiple daily injections, or on an insulin pump, should be a minimum of at least 3 times daily. But the recommended frequency should be varied for individual patients, especially for those not at glycemic targets, or in the setting of other special clinical circumstances. Certainly more frequent monitoring is often required to include preprandial and postprandial and occasionally 2 AM to 3 AM values, whenever people are above target, or whenever they have frequent episodes of hypoglycemia.

They then made recommendations for SMBG frequency for those on oral agents, with or without once-daily insulin, and for those above target on oral agents plus insulin, on insulin alone, or oral agents alone; the recommendation was at least twice daily. The consensus conference also indicated that once-daily insulin is generally less effective in achieving targets than either multiple daily injections, or therapy combining insulin and/or oral agents. Again, the recommended frequencies, the group said, should be varied for individual patients, particularly those not at target.

For the group on oral agents alone or insulin alone at target, the recommendation was at least once daily, including a weekly profile. They didn't specify the frequency of profile, but they did indicate that profile meant values at different times of the day that would include preprandial and postprandial values; for those at target on oral agents plus once-daily insulin, at least once a day with more frequent profiles.

And finally, SMBG frequency recommendations for those on nonpharmacologic therapy who are at or above target; at least 1 weekly profile and recommended frequencies should be varied for individual patients, especially those not at glycemic targets, or in the setting of other special clinical circumstances. This should be used to guide nutrition and physical activities, and the values should be used as a trigger for when pharmacologic therapy, for patients who are consistently above target, needs to be implemented.

Obviously, additional self-monitoring should always be performed in situations such as acute illness; changes in medication; with symptoms of hypoglycemia or hypoglycemic unawareness; during pregnancy; worsening or out-of-range A1C; or changes in nutrition, physical activity, or stress.

SMBG clearly is an important component in the management of all people who have diabetes. Data from the DCCT, Kumamoto, and other studies have demonstrated the benefit of SMBG testing in insulin-treated diabetic patients. Several recent studies including 2 meta-analyses of randomized, controlled trials have shown the benefit of SMBG in non-insulin-treated patients. A large longitudinal cohort study demonstrating the association between higher SMBG frequency and improved glycemic control was also presented.