PURPOSE: To evaluate hyperglycemia and correlation with length of stay (LOS) during corticosteroid treatment of obstructive airway disease.
METHODS: This retrospective analysis of 100 admissions (67 patients) for exacerbations of COPD or asthma admitted between January 2001 and December 2003 assessed demographic data, corticosteroid dose and duration, and glucose levels and management. Associations between mean glucose (MG), highest glucose recorded (HG) and LOS (the primary endpoint) were analyzed with ANOVA. Data were expressed as mean ± standard deviation.
RESULTS: Sixty-six percent were male, 16% asthmatic, 98% were smokers and 2% known diabetics. Mean age was 65.8 ± 13.5 years. Mean length of intravenous (IV) therapy was 4 ± 2.2 days, mean length of oral corticosteroid therapy was 2.4 ± 1.2 days, and mean total daily corticosteroid dose was 135.6 ± 50.1 mg. Mean LOS was 5.1 ± 2.6 days. MG and HG were, respectively, 152 ± 41 and 200 ± 84 mg/dl. Insulin sliding scale coverage was prescribed in only 44% admissions with insulin administration occurring in 37% and a baseline insulin regimen in 1%. Both MG and HG were divided into 3 groups: I (≤125 mg/dl), II (125-200 mg/dl) and III (≥200 mg/dl). A significant difference in LOS between the groups was found for MG (p=0.007) and HG (p=0.013). Although MG (p=0.807) and HG (p=0.232 ) were not affected by daily corticosteroid dose, HG (p = 0.009) correlated with total dose per stay. Glucose levels were recorded in 62% on discharge (mean 150.5 ± 47 mg/dl). Although 90% were prescribed prednisone, only 9% received glucose monitoring instructions upon discharge.
CONCLUSION: In patients hospitalized with COPD and asthma exacerbations, poor glucose control is prevalent and associated with increased LOS. However, glucose monitoring with insulin coverage is not routinely prescribed.
CLINICAL IMPLICATIONS: Glycemic control is known to affect outcomes in the critical care setting, but has not been similarly scrutinized in corticosteroid-treated obstructive airway disease. A prospective study assessing the effect of glycemic control on outcome in this population is needed.
DISCLOSURE: Omar Nehme, None.