Veronica Hackethal, MD
April 20, 2018
Current diabetes screening criteria that use age and weight alone, as recommended by the US Preventive Services Task Force (USPSTF), may miss up to half of prediabetes and diabetes cases, according to a study published online April 12 in the Journal of General Internal Medicine.
Using expanded criteria on the basis of other high-risk factors (gestational diabetes, polycystic ovarian syndrome, racial/ethnic minority, and/or family history of diabetes) may improve detection of abnormal blood glucose levels.
The study is the first nationally representative evaluation of how using expanded screening criteria could improve diabetes detection.
"This seems like a no-brainer to screen patients who have any of these additional risk factors," first author Matthew O'Brien, MD, assistant professor of medicine at Northwestern University Feinberg School of Medicine, Chicago, Illinois, said in a press release. "By demonstrating how well these expanded criteria work in identifying patients with prediabetes and diabetes, we're proposing a better path for the USPSTF to strengthen its screening guidelines."
The USPSTF 2015 recommendations call for prediabetes and diabetes screening in adults aged 40 to 70 years who are overweight or obese (referred to as limited criteria). The USPSTF also suggests, but does not formerly recommend, earlier screening in people with certain diabetes risk factors, including a history of gestational diabetes, polycystic ovarian syndrome, membership of an ethnic/racial minority, or a family history of diabetes (expanded criteria).
Early screening is important because it can enable earlier pharmacotherapy and lifestyle modification, potentially warding off more serious complications of diabetes.
"The earlier patients are diagnosed with these conditions, the sooner they can begin to combat them," O'Brien added.
To compare the limited USPSTF screening criteria to expanded criteria, O'Brien and colleagues conducted a cross-sectional study using data from a nationally representative sample of participants in the 2011 to 2014 National Health and Nutrition Examination Surveys study. The study included 3643 adults who had never been diagnosed with diabetes. The study defined abnormal blood glucose as an A1c ≥5.7%, fasting blood glucose ≥100 mg/dL, and/or 2-hour blood glucose ≥140 mg/dL.
Overall, the researchers found that 49.7% of the study population had undiagnosed abnormal blood glucose. By ethnicity/race, the prevalence was 48.6% among non-Hispanic whites, 54.0% among blacks, 50.9% among Hispanic/Latinos, and 51.2% among Asians.
Extrapolating from Census data, the researchers estimate that 105.1 million Americans have undiagnosed dysglycemia.
However, if they restricted their analysis to only those who met the limited criteria, they would only identify 47.3% of those who truly have abnormal blood glucose (sensitivity). The limited criteria would correctly weed out 71.4% of people who truly do not have abnormal blood glucose (specificity).
The expanded criteria performed better in identifying true cases of abnormal blood glucose and identified 76.8% of people who were truly positive. In contrast, the expanded criteria would only weed out 33.8% of people who truly did not have abnormal blood glucose.
The results mean that using the expanded criteria will miss fewer cases of abnormal blood glucose, but would result in further testing for more individuals, many of whom will have normal results.
Because the Affordable Care Act requires health plans to fully cover services recommended by the USPSTF, the results raise the question of whether insurance should cover blood glucose screening in people who meet the expanded criteria.
"This could be a particular problem for people of low socioeconomic status who are at high risk of developing diabetes and may be unable to pay for a screening test," O'Brien commented in the press release.
The results also showed that using the limited criteria would miss other high-risk groups. These include women with polycystic ovarian syndrome and/or a history of gestational diabetes, and younger age groups who are increasingly at risk for diabetes.
Importantly, racial/ethnic minorities have higher rates of abnormal blood glucose and are at higher risk for diabetes complications than whites. Blacks develop abnormal blood glucose levels at younger ages, and Asians do so at lower body weights.
In general, the expanded criteria showed better performance for racial/ethnic minorities. In contrast, the limited criteria showed a trend for lower detection of truly abnormal blood glucose levels in all minority groups, especially Asians. Using the limited criteria, 70% of Asians with prediabetes or diabetes would go undiagnosed.
The study was limited by the use of National Health and Nutrition Examination Surveys data, in which the definition of Asian race does not include Pacific Islanders, for whom diabetes risk is particularly high.
The study was sponsored by the National Institute of Diabetes and Digestive and Kidney Diseases and the Centers for Disease Control and Prevention. The authors have disclosed no relevant financial relationships.