American Association for Clinical Chemistry
Better health through laboratory medicine
August 10, 2006
 

In This Issue . . .

Diagnosing Diabetes: In Defense of the Oral Glucose Tolerance Test

Julie L McDowell


Identifying patients who are likely to develop diabetes before clinical onset, such as those with impaired glucose tolerance (IGT), is critical since medications and other preventive efforts can be implemented to thwart the disease. This is the position of many American clinicians, including William H. Herman, MD, MPH, a professor of internal medicine and epidemiology at the University of Michigan and Director of the Michigan Diabetes Research and Training Center. Herman was recently recognized by the American Diabetes Association (ADA) for his epidemiological work in studying diabetes in large populations, although his views on diagnosing diabetes through oral glucose tolerance testing (OGTT) differ somewhat from the organization’s current guidelines. This issue of Strategies examines Herman’s thoughts on diabetes diagnosis, and its potential implications for the clinical lab.

The ADA 2006 Position Statement on Diagnosis and Classification of Diabetes Mellitus outlines three ways that the disease should be diagnosed, as depicted on pg. S47 of the “Criteria for the Diagnosis of Diabetes Mellitus,”  which can be found online at http://care.diabetesjournals.org/cgi/reprint/29/suppl_1/s43 (Diabetes Care 29: S. 43-48, 2006).  But the best method for diagnosing diabetes is performing the OGTT, said William H. Herman, MD, MPH, in a speech at the ADA’s Annual Scientific Sessions in June, where he received the Kelly West Award for Outstanding Achievement in Epidemiology. This thought runs counter to the ADA’s recommendations, which state that the OGTT is not recommended for routine clinical use. But since diabetes is associated with insulin deficiency and hyperglycemia, a stimulated or post-glucose load glucose level will distinguish normal from abnormal better than an unstimulated or fasting glucose level, says Herman.

“Diabetes diagnosis should be based on fasting glucose and post-glucose load glucose levels, to give a true picture of glucose tolerance,” Herman noted. “The problem with fasting glucose is that it’s not as sensitive as a post-glucose load glucose level, but it’s easer to do when compared to a two-hour glucose load glucose level.”

While he acknowledges that these tests are more complex and expensive to perform, they can also identify which patients have IGT. Based on his background conducting large population-level studies, Herman believes that screening millions of at-risk people using these tests can be done economically and will detect patients who process glucose in an abnormal manner at the earliest stages of IGT and diabetes. With proper intervention, including nutrition, exercise, and medication, full blown diabetes can be thwarted. Herman pointed to his own research and that of others showing that tight blood sugar control can delay or reduce the chances of problems associated with long term diabetes-related complications.

Herman also takes issue with the ADA’s current definition of diabetes as outlined in the diagnostic criteria, suggesting instead that diabetes should be defined as hyperglycemia sufficient to cause adverse health outcomes. Since IGT is clearly associated with risk of retinopathy, neuropathy, cardiovascular disease, and cardiovascular mortality, Herman says that the distinction between IGT and diabetes is unnecessary and diabetes should be defined to include either IGT or diabetes (fasting glucose >= 126 mg/dl or 2-hr post 75 gm glucose load glucose >=140 mg/dl). “The latter would really simplify diagnosis and facilitate identification of individuals who could benefit from intervention,” he explains.

Labs Should Adhere to Standards and Guidelines

When comparing fasting plasma glucose and OGTT, it’s important to note that OGTT is indeed more sensitive, explained William Winter, MD, Professor, Departments of Pathology and Laboratory Medicine and Pediatrics, University of Florida College of Medicine in Gainesville.

“In part, this reflects the fact that physiologically, it’s easier to control a preprandial than a postprandial glucose test,” he said. “If you fast long enough, and if you have sufficient insulin secretion and sensitivity, you may have a normal fasting glucose. However, when you are put under the stress of an oral glucose load you may have impaired glucose tolerance or even frank hyperglycemia. We know that people with diabetes have a defective first phase insulin response, regardless of whether they have type 1 or type 2 diabetes.”

In addition to these testing questions, one of the primary issues for the clinical laboratory in regard to diagnosing diabetes is standardization, according to Winter. The best way to screen for pre-diabetes is the standard oral glucose tolerance test because there are validated reference intervals for both fasting and the two-hour glucose levels. “I don’t think it’s good medical practice to simply measure glucose following a meal because there are no standards to interpret these data,” he explained. “I think that labs should be able to offer the standard oral glucose tolerance tests and perform them correctly.”

It’s important to follow the ADA guidelines in order to minimize variables that may give aberrant results. The guidelines state that patients should not be acutely ill, should not have been recently hospitalized, and should be on their regular diet and regular exercise regiment. Patients need to drink glucose over the specified length of time. Of course, the patient should not be exercising beforehand. Between the first and second draw two hours later the patient should be sedentary—and should avoid smoking and caffeine, which can affect glucose, said Winter.

Update on Medically Unbelievable Edits

Proposed “medically unbelievable edits” (MUE) codes for most of the pathology and laboratory services have now been removed, the College of American Pathologists (CAP) reported in late July, after receiving notice of a letter from officials with the Centers for Medicare and Medicaid Services (CMS) to the American Medical Association (AMA). In addition, CMS has renamed the proposed code changes as “medically unlikely edits.”

MUEs are part of a plan to curb coding errors and reign in Medicare spending. Earlier this year, CMS’s National Correct Coding Initiative (NCCI) announced a plan to establish payment limits on medical services, including diagnostic tests, billed for a particular code by a health care provider for the same Medicare patient performed on the same date. If a provider reports a claim that exceeds these limits, the claim would automatically be denied (www.aacc.org/AACC/publications/strategies/030906.htm). CMS released the proposed MUE codes for public comment in March with plans to implement in July, although the agency soon postponed that date. Many medical associations, including AACC, CAP, and the AMA, criticized the program.

CAP reports that the MUEs will be implemented in quarterly phases beginning in January 2007. The current list of MUEs now open for public comment is available online at www.cap.org/apps/docs/statline/pdf/mue_public_comment.pdf .