Under pressure from insurance providers to cut corners whenever possible, doctors struggle with the question of which tests or procedures can be safely eliminated and which cannot. One test that has come under scrutiny is laboratory testing for gestational diabetes (GD). This screening test, which usually costs $75 to $85, measures blood sugar response to glucose consumed in a drink and is performed between the 24th and 28th week of pregnancy or at about six months. Many obstetricians do this test routinely while others weigh risk factors before ordering it. These risk factors include: family history of diabetes, prior large baby or stillbirth, being overweight, age (over age 25) and ethnicity (Hispanic, African-American, Native American, Asian American or Pacific Islander). Experts presenting at the American Diabetes Association’s 60th Annual Scientific Sessions in San Antonio took on this issue and presented the case for both selective and universal screening for GD.A proponent of selective testing for GD, Anne Kenshole, MD, of Women’s College Hospital in Toronto, Canada, recommended using a scorecard of five criteria to eliminate unnecessary testing: under age 25, low-risk ethnicity, negative family history, no previous GD and no adverse obstetrical problems. She argued that having had a large baby (over nine pounds) is not necessarily a risk factor because 80 percent of large babies are born to women with normal glucose tolerance. (Currently, standard screening criteria includes having given birth to a large baby.)
Dr. Kenshole contended that at least 10 percent of the 4.3 million U.S. women who become pregnant every year should be eliminated from GD testing using standard screening criteria. She noted that studies indicate that only three percent of women eliminated using risk factor criteria develop GD. “Selective screening could eliminate 430,000 women from testing, resulting in significant cost savings,” she said. “If three percent is missed, that equates to 14,000. The question is if this is a medical-legal nightmare or a justifiable allocation of resources.”
Furthermore, Kenshole suggested that a false-positive result creates a sense of “psychological sequela” (aftereffect of disease) in women who begin to worry that their baby isn’t healthy. A false-positive diagnosis may cause the doctor to overreact, which “may lead to inappropriate treatments that are meddlesome and not cost effective,” she added. In a review of 3,013 women tested for GD in Toronto, 18 percent had a false-positive result. “Targeting high-risk and lowering the threshold (reducing screening criteria), reduced the false-positive ratio.”
An advocate of universal screening for GD, Michael Greene, MD, of Harvard University School of Medicine, contended that all pregnant women should receive both risk factor screening and the biological test. He argued that the cost of testing the 10 percent of patients Kenshole would eliminate is trivial by comparison with the consequences of missing even three percent of patients with GD. “It’s pennywise and dollar foolish,” Dr. Greene said. “If it costs $13 million to test everyone, this is peanuts when you consider total dollars spent on healthcare every year.”
Additionally, Greene contended that risk-based screening is not that simple. “As the schemes (for risk factor diagnosis) get more complex, it’s less practical for doctors in a busy practice to use this (standard risk factor screening). There’s greater probability that patients will fall through the cracks,” he stressed.
Editorial Commentary by Dr. Anne Peters: When determining the value of a screening test, it is important to remember the cost of missing a diagnosis versus the cost of the test. With regards to gestational diabetes, it is important to identify women who have the disease in order to treat them and their developing fetus appropriately. Although selective testing eliminates the need for testing women who are at extremely low risk of having the condition, it also means that more cases of gestational diabetes will be missed. Additionally, physician behavior is an important component of implementing any testing strategy. To make testing complex, requiring thought about each patient and their need for testing, may make physicians less likely to perform the test. Most pregnant women that I have treated are very sensitive about the need to do everything possible to ensure a healthy outcome. If I were pregnant and was told that taking a simple test half way through my pregnancy could help determine if I had a disease as significant as gestational diabetes, I would gladly take the test. If I were given a false positive result (told that I had gestational diabetes when, in fact, I didn’t), I would rather deal with a further evaluation than miss something in the first place. Pregnant women are by and large one of the most motivated groups of patients I know of. I believe that to continue with the recommendations for universal screening is the wiser path, even if screening costs money. A poor outcome from a missed case of gestational diabetes could mean harm to mother and baby and high legal and medical costs. For a host of reasons, it seems better to screen than not, at least until we are more certain of the validity of risk assessment for the development of the disease.