PS2-4: Diabetes Screening and Detection in an Ambulatory Clinical Population

Abstract

Background/AimsType 2 diabetes mellitus is associated with significant morbidity and mortality in the U.S. It is unknown whether racial/ethnic disparities in diabetes prevalence are being adequately addressed in clinical practice with respect to screening. Our aim was to determine diabetes screening and detection rates by race/ethnicity and whether they vary depending on screening criteria used. Racial/ethnic minority groups are generally found to have higher prevalence of diabetes, but little is known about whether screening and detection rates vary by race/ethnicity.MethodsData from electronic health records (EHR) in a large ambulatory group practice were used to construct a cohort of persons aged 35 and older without prevalent diabetes (N = 116,355). They were followed for diabetes screening (average follow up 17 months in 2007–2011). We computed standardized screening and detection rates in each racial/ethnic group and examined predictors of screening. We compared positive detection rates for those who met the U.S. Preventive Services Task Force (USPSTF) or American Diabetes Association (ADA) screening criteria and those who did not.ResultsMost patients (82%), including 75% of those meeting neither criteria, were screened during follow-up. Racial/ethnic minority groups had higher screening rates – Latino (82%), Asian subgroups (83% (Japanese) to 90% (Filipino)), non-Hispanic Black (92%) – than non-Hispanic White (80%) (P <0.001). Detection rates were also significantly higher for some minority groups (Latinos (2.1%), Asian Indians (2.4%), and Filipinos (3.8%)) than non-Hispanic White (1.3%) (P <0.05). Among screened, 1.6% had positive test results (number needed to screen (NNS) = 64). The NNS was 46 for those meeting ADA criteria, 36 for USPSTF criteria and 132 for those meeting neither.ConclusionsNearly all patients in our study population received screening, and racial/ethnic minority groups were not less likely to be screened. Current clinical practice screens a high proportion of individuals meeting neither ADA nor USPSTF guidelines. Given low screening costs that may be appropriate because many new cases were detected among those without risk factors. Given the substantial benefit of early detection of diabetes our study suggests that universal, preventive screening for diabetes for asymptomatic individuals may have utility.