ABSTRACT
PURPOSE: This study investigated: (1) the type of corticosteroid associated with the greatest degree of hyperglycemia, assessed using bedside capillary blood glucose monitoring, in hospitalized patients; and (2) the pattern of hyperglycemia throughout the day with the use of each type of corticosteroid. METHODS: This single-center, retrospective study used data from 964 adult inpatients receiving oral or IV corticosteroids. Data on capillary blood glucose concentrations and time taken over 7 days were collected. A mixed model for repeated measures was applied to investigate changes in glucose concentration over time with the use of four different corticosteroids. An autoregressive covariance structure was used to model correlations between repeated measurements. FINDINGS: Across all 7 days, the mean blood glucose concentration was greater with dexamethasone compared to that with hydrocortisone (mean difference, 16.6 mg/dL [95% CI, 8.1-24.8] [0.92 mmol/L (95% CI, 0.45-1.38)]) or prednisolone (mean difference, 20.0 mg/dL [95% CI, 14.2-25.7] [1.11 mmol/L (95% CI, 0.79-1.43)]). The mean blood glucose concentration was greater with methylprednisolone compared to that with hydrocortisone (mean difference, 23.9 mg/dL [95% CI, 11.3-36.4] [1.33 mmol/L (95% CI, 0.63-2.02)]), and with methylprednisolone versus prednisolone (mean difference, 27.4 mg/dL [95% CI, 16.4-38.3] [1.52 mmol/L (95% CI, 0.91-2.13)]). There were no significant differences in the patterns of hyperglycemia at six time points of the day with each type of corticosteroid. IMPLICATIONS: Treatment with oral or IV dexamethasone or methylprednisolone was associated with greater hyperglycemia in comparison to prednisolone and hydrocortisone. More vigorous monitoring and intervention, when necessary, are suggested in adult inpatients receiving corticosteroids, in particular dexamethasone and methylprednisolone.