1.5 Million Fingersticks Later: Benchmarking Inpatient Glycemic Control from a National Sample of Hospitals
Abstract Number: 397-P
Authors: PREM G. THOMAS, KASIA J. LIPSKA, JANIS E. BOZZO, SILVIO E. INZUCCHI
Institutions: New Haven, CT
Results: With growing interest in improving the quality of hospital glycemic control, several guidelines now recommend specific target blood glucose (BG) ranges in both critically and non-critically ill patients. There are, however, few published data on the average BG of hospitalized patients in the US. Glucometrics is a free, open-access, university-based internet site that allows hospitals to upload and analyze their de-identified capillary BG data. A variety of metrics are calculated and reported back to the end-user. These employ the patient sample, each 'patient-day', and each 'patient-stay' as analytical models, allowing for a standardized methodology.
During Glucometrics' first 2 years, 1,593,766 individual samples have been submitted from 31 geographically diverse US institutions; of these, 1,559,676 were suitable for analysis. From intensive care units (ICUs), 20 hospitals contributed 515,134 BGs, representing 20,893 patient-stays, and 113,805 patient-days. The average patient-day mean BG was 151.8±53.9 mg/dl (median, 139.5). A patient-day mean BG of 70-149 was achieved in 57.9% of patient-days (range from individual hospitals, 32.8%-75.5%). 8.2% (range, 0%-17.4%) of patient-days had at least one BG<70 (hypoglycemia) and 8.4% (range, 2.9%-18.1%) at least one BG≥300 (severe hyperglycemia).
From non-ICU medical-surgical wards, 28 hospitals contributed 1,044,542 BGs (38,882 patient-stays, 276,358 patient-days.) The average patient-day mean BG was 152.7±55.3 mg/dl (median, 141.0). 55.8% of patient-days achieved a mean BG of 70-149 (range, 27.2%-89.2%). 9.5% (range, 3.9-15.7%) of patient-days had at least one hypoglycemic BG and 8.5% (range. 1.8-39.3%) at least one severely hyperglycemic BG.
Among hospitals using the website, the average BG is ∼150 mg/dl in glucose monitored inpatients, whether in the ICU or on general wards. Based on current guidelines, there is room for improvement in achieving glycemic targets, especially in the critically ill. We also noted marked variability in glycemic control between hospitals. The Glucometrics website may serve as a benchmarking tool in the assessment of national trends in hospital glucose management.