Annual rates of emergency department visits or hospitalizations for hypoglycemia (HYPO) were evaluated using a serial cross-sectional study design (2004-2015) of VHA/Medicare enrolled patients. Subjects were ≥65 years receiving secretagogues only (Secr), or insulin with or without Secr (Ins), and who had HbA1c value(s). Separately for Secr/Ins groups and by race/ethnicity, we assessed age/sex adjusted HYPO rates (per 1000 patients) and A1c values (from prior to an event or end of year for those without an event). In 2015, subjects were 99% men, 10% [B]lack, 6% [H]ispanic, and 75% [W]hite; 75% had major comorbidity categories; 24% had cardiovascular conditions only, and 9% advanced diabetes only. Ins use increased by 88%, similar among B/H/W. HYPO rates markedly declined for Ins and Secr in all races but remained higher for B. The percent of patients with A1c <7% decreased: 44.2%/32.9%[B], 51.0%/32.6%[W], and 41.3%/28.4%[H]; the percentage for A1c 7.0-7.9% increased: 28.6%/31.6%[B], 30.6%/35.8%[W], and 30.8%/32.4%[H]; the percentage for >9% also increased: 11.9%/15.8%% [B], 6.4/12.2%[W], and 12.0/17.6%[H]. Mean A1c values also increased in B/H/W, and were similar (difference ≤0.2%) between patients with and without HYPO. Guideline recommendations (post 2008) for individualized A1c goals may have mediated reductions in racial disparities for HYPO rates.


C. Tseng: None. O. Soroka: None. C.E. Myers: None. D.C. Aron: None. L.M. Pogach: None.


U.S. Department of Veterans Affairs (IIR14-082)

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