What metrics should be considered beyond HbA1c to best measure diabetes outcomes?
Other aspects of glucose metabolism tell the full story.
HbA1c, by itself, is not a good enough measurement of diabetes outcomes. Specifically, it really does not tell the story of how that number was achieved. It could have been achieved by having glucose in the range of 250 mg/dL and 50 mg/dL or by having a range between 80 mg/dL and 120 mg/dL. I would like to see other measurements that will help provide that answer. However, putting any new measures in place will require robust studies, validation and education.
There are a lot of deficiencies with the HbA1c measurement. It is not the most accurate measure for patients with severe kidney disease, sickle cell anemia or various blood diseases, and studies have shown it is not an accurate measure for some ethnicities. But, on an average, it’s the best way to follow up on the patient’s glucose control. I don’t have an alternative.
Several experts and patient groups are suggesting continuous glucose monitoring (CGM) as a measure to assess diabetes outcomes in a different way, measuring how long a person was within, above or below goal. This indeed is helpful and may give me a better understanding of the diabetes control of the patient. But CGM is a technology most people do not or cannot use. CGM may be attainable by people with type 1 diabetes, but may not be available to most people with type 2 diabetes, especially those under CMS.
There are some other alternatives I would like to see. Measurements of fructosamine or glycated albumin or advanced glycation end-products (AGEs) are other aspects of glucose metabolism that could give us different milestones not necessarily based on red cells. AGE markers could help us identify which people are at higher risk for complications. It would also be helpful to look at how any diabetes therapy affects the eyes, kidneys, heart, etc.
However, translating all of that becomes more complicated. I’m very much in support of finding new measures beyond HbA1c that look at specific outcomes and specific types of measurements, such as glycated proteins, CGM and time within range and variability. I’d love to have all of these measures available for me and my patients. However, I don’t want to see any change before measures are validated. It is not enough to say, “We think it’s better.” I want to see validation for the outcome. And, once we make a change, there needs to be funding for the necessary education to get clinicians to be able to use the measures.
Yehuda Handelsman, MD, FACP, FNLA, FACE, is medical director and principal investigator for the Metabolic Institute of America in Tarzana, California. Disclosure: Handelsman reports receiving consultant and speaker fees and research grants from various pharmaceutical companies.
Focus on adherence to therapy.
HbA1c has evolved to become the gold standard outcome to measure the efficacy of glucose-lowering medications. It’s likely time to question this. HbA1c may not be an accurate measure for certain populations, such as African Americans, people with certain anemias or thalassemia or those taking certain medications or vitamin supplements. The question must be raised about whether sufficient screening is done in studies to detect individuals whose HbA1c results may not be accurate. As in clinical practice, one shouldn’t rely on an HbA1c result alone to assess the impact of glucose-lowering medications.
An important consideration for FDA approval of diabetes therapies should be adherence. While there is a big leap between a drug trial and the real world, there are ways FDA could objectively measure the ease of or challenges with adherence. Factors such as dosing frequency, need for refrigeration, ease of use of the delivery device and side effects all affect adherence. We can’t allow the excuse that ‘We can’t measure it, so it doesn’t matter.’
We know that adherence rates — whether to glucose-lowering medications, hypertension medications, lipid-lowering medications — are far from stellar. As we say in nutrition, ‘It’s not nutrition unless you eat it.’ In the case of medication, it’s not medication unless you use it.
At the FDA workshop, significant attention was paid to measuring hypoglycemia as an outcome. While this measure is certainly of importance with glucose-lowering medications that can cause hypoglycemia, such as insulin, today, there are fewer medications that can cause hypoglycemia. Clearly, hypoglycemia, both degree and frequency, is an important metric to track, but we don’t necessarily need to measure it when studying medications that don’t cause hypoglycemia.
The discussion surrounding patient-related outcomes is interesting. I support the inclusion of these as objective measures when possible.
A topic not discussed at this forum, but one that troubles me as a diabetes educator, is that there is typically not an arm of these studies that requires the inclusion of diabetes self-management education and support (DSMES) as a comparator or in addition to the use of a medication. For many years now, DSMES has been included in the standard of care and management of people with diabetes. Omitting this important therapy and standard of care in glucose-lowering medication studies should no longer be acceptable by the FDA.
Hope Warshaw, MMSc, RD, CDE, BC-ADM, owns Hope Warshaw Associates, LLC, a consulting practice in Northern Virginia. During 2016, Warshaw is serving as president of the American Association of Diabetes Educators. Disclosure: Warshaw reports no relevant financial disclosures.