INPATIENT DIABETES MANAGEMENT AT THE NEW MEXICO VA HEALTH CARE SYSTEM IN ALBUQUERQUE, NEW MEXICO.

Abstract

Purpose Several recent studies have demonstrated that better glycemic control is related to improved outcomes for hospitalized patients. Despite the evidence, poor glycemic control in the inpatient setting continues to be a problem. The first step toward solving this problem is to develop an understanding of current practices related to glycemic control in the inpatient setting. This study describes current practice at the Veterans Administration (VA) hospital in Albuquerque, NM. Methods We evaluated 325 patients (367 admissions) who were discharged from the VA hospital between December 1, 2004, and February 28, 2005. Patients on the medicine, psychiatry, and surgery services with diabetes as one of the admitting diagnoses were included in this study. ICU data were excluded. Blood glucose data were retrieved from the Computerized Patient Record System (CPRS) using a data extraction program. Where possible, missing data were acquired from the paper record. Results Of the 367 admissions, 79 did not have any capillary blood glucose (CBG) measurements during the admission. Of the remaining 288 admissions, 206 (71.5%) were admitted to medicine, 17 (5.9%) were admitted to psychiatry, and 65 (22.6%) were admitted to surgery. There were 5,297 CBG readings available of the 5,513 glucose test strips ordered (96%). CBGs ranged from 15 to 577 mg/dL and were grouped into five categories for data analysis: hypoglycemia ( 400). Of the 5,297 available CBGs, 107 (2.0%) were hypoglycemic, 2,154 (40.7%) were acceptable, 1,341 (25.3%) were high, 1,604 (30.3%) were hyperglycemic, and 91 (1.7%) were severely hyperglycemic. Patients on the psychiatry service had significantly better glycemic control compared with those on medicine and surgery services (p = Conclusions This descriptive study demonstrates that glycemic control at the Albuquerque VA hospital is not optimal. More than half of recorded CBGs were either unacceptably high or unacceptably low. CBGs for patients on the psychiatry service were more likely to be in the acceptable range compared with CBGs for patients on the surgery or medicine service. However, there were significantly fewer patients on the psychiatry service compared with the other services, which could introduce bias.