Prevalence of Hyper- and Hypoglycemia Among Inpatients With Diabetes

The recent demonstration (1–3) of the benefits of intensive glycemic control in hospitalized patients has renewed interest in inpatient management of diabetes. Poor glycemic control is a marker for poor quality of hospital care (4), as well as an important safety issue: insulin is one of five medications most associated with inpatient medication errors (5,6). Moreover, many hospitals continue to solely rely on insulin “sliding scales” despite the limitations of this approach (7,8). To gain a broader understanding of the current quality of inpatient diabetes management, we analyzed the prevalence and management of hyper- and hypoglycemia among 999 patients with known diabetes treated in 44 hospitals across the U.S. Data were derived from two sources: the University Health System Consortium (UHC) Diabetes Benchmarking Project and VHA, Inc. The UHC project collected inpatient and outpatient data in 2003 by standardized chart review of 274 patients aged ≥18 years with type 1 and type 2 diabetes (diagnosed by their outpatient physicians), who were admitted as inpatients to 1 of 29 academic medical centers located in 20 states. Chart reviewers identified the highest and lowest glucose values during hospital admissions and recorded the highest and lowest glucose results for the 2 days preceding and following the peak and nadir. In 2003–2004, 15 member hospitals of VHA, Inc, an alliance that serves ∼1,400 not-for-profit U.S. hospitals, performed baseline chart reviews on 725 general medical and surgical patients aged >18 …