In 2015, the USPSTF recommended targeted screening for dysglycemia in those adults aged 40-70 years of age who were overweight or obese. However, dysglycemia was less likely to be detected among racial/ethnic minorities than for whites. This disparity occurred because more minority patients developed dysglycemia at normal weights and younger ages compared to whites.
Lead author Matthew J. O’Brien MD, from Northwestern Feinberg School of Medicine told EndocrineWeb, “When we looked at the population trends of rising rates in diabetes among younger people and those over 70, as well as those with a lower body weight, we speculated that the new USPSTF guidelines may have a low sensitivity.”
Their initial assumptions proved correct. “Our findings suggest that certain groups that are at a high risk for developing dysglycemia are more likely to be missed when using these criteria, particularly racial/ethnic minorities.”
The USPSTF is an independent group of specialists who make recommendations for the use of screening tests and services in medical practice. Health insurance plans must cover the recommended tests and services according to federal law.
The authors retrospectively evaluated EHR data from 50,515 adult primary care patients from six community health centers in the Midwest and Southwest serving vulnerable populations from 2008-2013. There was a 3-year follow-up period with the primary outcome being the development of dysglycemia. The exposure of interest was eligibility for screening based on the 2015 USPSTF recommendations.
The study cohort was predominantly less than 40 years old (62.7%), overweight/obese (66.4%), female (72.5%), and uninsured or publicly insured (74.3%). The racial/ethnic background was 34.7% Black, 33.9% Hispanic/Latino, 22% White and 8.7% other (Asian, American Indian, Hawaiian/Pacific Islander, biracial or multiracial).
Out of 50,515 patients available for screening, 59.3% (29,946) had a screening test within three years of the index visit. Screening was completed in 77.8% and 53.1% of patients eligible and not eligible according to the 2015 USPSTF criteria respectively. During the follow-up period, 5,960 patients developed prediabetes and 2,518 developed diabetes. The majority of dysglycemia cases were identified by A1C (77.9%) and fasted glucose (18.2%).
The USPSTF criteria had the following performance results based on all patient data over the 3-year follow-up period: sensitivity 45%, specificity 71.9%, positive predictive value 38.8%, and negative predictive value 76.8%. Of those cases missed by the USPSTF criteria, 77.7% were under 40 years of age, and 29.3% had a normal body weight.
Age ≥40, overweight/obesity, nonwhite race/ethnicity, hypertension, polycystic ovary syndrome, history of gestational diabetes and family history of diabetes were all significantly associated with developing dysglycemia during follow-up.
Higher odds of dysglycemia were found for Hispanic/Latino (OR: 1.46, 95% CI: 1.30-1.64), Asian, American Indian, Hawaiian/Pacific Islander, biracial or multiracial (OR: 1.33, 95% CI: 1.16-1.54) and Blacks (OR: 1.24, 95% CI: 1.09-1.40) compared to Whites.
Sensitivity based on racial/ethnic groups showed Hispanic/Latino the lowest at 37.7%, followed by Asian, American Indian, Hawaiian/Pacific Islander, biracial or multiracial at 42%, Blacks at 50.3% and Whites at 54.5%.
According to Dr. O’Brien, about half of the patients in the study received testing who currently would not be eligible according to the new USPSTF criteria. The lack of mandatory insurance coverage for those ineligible for screening may result in even lower screening rates than those found in their study, particularly for those patients unable to afford out of pocket expenses.
Additionally, because dysglycemia develops in racial/ethnic minorities at younger ages and normal bodyweights and the new guidelines miss a substantial proportion of this patient population, there is concern over potentially delayed interventions to prevent and treat diabetes.
Further research is needed to evaluate the performance of the new screening criteria in population-based studies in which all individuals are screened. “Our findings may not apply to other demographic groups that we didn’t study, so our team is currently investigating how the USPSTF screening criteria apply to a nationally-representative population,” Dr. O’Brien commented.
The authors recommend physicians should consider screening racial/ethnic minorities at earlier ages and lower weights than those recommended by the USPSTF criteria. “We recommend that providers screen patients based on a more global assessment of their diabetes risk, including other risk factors such as hypertension, and a family history of diabetes. Also, because racial/ethnic minorities are at a higher risk of developing diabetes than whites, providers should consider having a lower threshold for diabetes screening, even if patients are younger than 40 or have a normal weight,” Dr. O’Brien told EndocrineWeb.
Priyathama Vellanki, MD agreed and said, “The clinician should use their judgment to see which patients need screening for dysglycemia rather than follow the guidelines strictly for each patient. Strictly following the guidelines may lead to missed diagnoses for dysglycemia, missed opportunities for intensive lifestyle interventions and progression to overt diabetes.” Dr. Vellanki is Assistant Professor, Division of Endocrinology, Metabolism and Lipids at Emory University School of Medicine in Atlanta, and a member of the EndocrineWeb.com Editorial Board.
The authors have declared that no competing interests exist.
August 30, 2016
O’Brien MJ, Lee JY, Carnethon MR, et al. Detecting Dysglycemia Using the 2015 United States Preventive Services Task Force Screening Criteria: A Cohort Analysis of Community Health Center Patients. PLOS Med. http://dx.doi.org/10.1371/journal.pmed.1002074.