Nutrition support and glycaemic variability in critically ill patients

Rationale: Glycaemic variability (GV) is associated with excess mortality. The aims of this study were to determine associations between GV and clinical outcomes (mortality, duration of ventilation and intensive care unit (ICU) length of stay) as well as between GV and nutrition provision, particularly carbohydrate provision, from artificial nutrition support.Methods: Data on 760 critically ill patients ventilated for>24 hours and exclusively fed artificial nutrition support in a heterogeneous ICU was extracted from a clinical database over 18 months. This data was analysed for associations as outlined in the aims. GV is defined as the coefficient of variation (standard deviation/mean of blood glucose levels) x100). Statistical analysis was performed using logistic, zero-truncated negative binomial and linear regression as appropriate to the distribution of the outcome variable using R software.Results: GV was associated with excess mortality (Odds Ratio (OR) (95% confidence interval (CI)) 1.04 (1.02e1.06); p<0.001). There was no association between GV and ventilation hours or length of ICU stay. Increasing insulin dose (Regression Coefficient 0.08 (0.06e0.09); p<0.001) was associated with GV. There were no significant associations between GV and nutrition provision; however, further exploratory analyses showed that of those patients requiring insulin, increased insulin dose was associated with increasing carbohydrate provision (Incidence Rate Ratio 1.003 (1.001e1.005); p¼0.001).Conclusions: GV is associated with excess mortality and therefore mitigation of high levels of GV is prudent. Increasing doses of insulin appears to be associated with greater GV. Strategies to reduce insulin use may be warranted with one potential consideration being a reduction in exogenous carbohydrate provision in light of its association with increased insulin doses.References: Krinsley J.S. et.al (2013) Critical CareDisclosure of interest: None declared.