Screening Patients for Diabetes in a Large, National Clinical Database

Abstract

Objective: To characterize diabetes screening patterns among eligible patients per ADA Standards of Medical Care in Diabetes using a large, geographically diverse clinical dataset. Study Design: Retrospective descriptive analysis conducted in a clinical database containing over 25 million patient records. Population Studied: A total of 5.1 million patients aged 18-75 in 23 health care organizations (HCOs) with at least one outpatient visit in the study period (7/2016-6/2017), no prior diagnosis of diabetes mellitus, and no prior diabetes medication prescriptions (except metformin). Principal Findings: Among 5.1 million patients, 74% were eligible for screening and 55% were screened according to guidelines. Adequate screening ranged from 47%-76% across 23 HCOs. Adults 65 and older were most likely to be adequately screened (61%; P < .01). Least likely to be screened were patients with low-income insurance types (P < .01); non-white race (P < .01) or lower education by zip code (P < .01). Of 1.7 million patients screened, 2.8% and 36.5% had evidence of diabetes and prediabetes, respectively. Diabetes/prediabetes was identified in 30% of patients aged 35-44 and 48% of patients 65 and older. Patients with lower education were less likely to be adequately screened, but yielded greater rates of diabetes/prediabetes, 47% to 39% from least to most educated (P < .01). More African Americans (46%) screened positive for diabetes/prediabetes compared to other race/ethnicities (42%) (P < .01). Conclusions: Screening practices vary greatly, mostly within and across HCOs. Patients least likely to be adequately screened were often those with highest yields of diabetes/prediabetes. The identification of inadequately screened, low-income, lower education, or race/ethnic minority populations may facilitate targeted efforts to decrease disparities in screening practices. Identification of organizations with the highest net yield may also provide insight into strategies for efficient screening practices. Disclosure N. Stempniewicz: None. J.K. Cuddeback: Research Support; Self; Novo Nordisk Inc.. Other Relationship; Self; Johnson & Johnson Diabetes Institute, LLC., Merck & Co., Inc., Sanofi US, Boehringer Ingelheim Pharmaceuticals, Inc., Eli Lilly and Company. C.R. Rattelman: None. E.L. Ciemins: None.