There are limited data which examine the current trends in emergency department (ED) use among adults with all-type diabetes by demographic characteristics. We analyzed data from the nationally representative Medical Expenditure Panel Survey (MEPS), which uses a rotating panel design—a new cohort is initiated annually, and health care use is recorded over the course of 2 years. We pooled panels, each with 2-year follow-up into years 2009-2012 and 2014-2017 to improve precision. We identified adults with diabetes by self-report, presence of diabetes-specific ICD-codes, or prescriptions of antidiabetic agents and used all-cause event-level ED data to estimate weighted counts and age-adjusted rates of average annual ED visits per 1000 persons with diabetes. We then compared rates of ED use between the two time periods by calculating absolute and percent change. Increases in use were greatest among the groups: 65-74 yrs (38.1%), 75+ yrs (28.6%), West (27.3%), Midwest (16.5 %), Female (19.4%), and White (15.9%). The share of visits by the uninsured dropped (10.1% to 5.3%), while share of visits increased (29% to 36.3%) for those covered by Medicaid (data not shown). Further analysis will be conducted to analyze the trend over time from earlier time periods (2004-2007), understand the type of event (nondiabetes vs. diabetes-related) and stratify by potentially preventable visit and insurance type.


T.S. Uppal: None. G. Fernandes: Employee; Spouse/Partner; Janssen Pharmaceuticals, Inc. Employee; Self; Merck Sharp & Dohme Corp. J. Haw: None. M.K. Shah: None. S. Turbow: Research Support; Self; Gilead Sciences, Inc. S. Rajpathak: Employee; Self; Menarini Group. K. Narayan: None. M.K. Ali: Research Support; Self; Merck & Co., Inc.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at