Statistical and clinical analysis of alterations in glucose values after burns

The purpose of this study was to evaluate independent factors (mainly critical hyperglycaemic values on admission) with an impact on outcome in burn patients (sepsis and mortality) and analyse prevalence of critical hyperglycaemia on admission and during burn disease in adult and elderly patients with severe burns. This was an observational retrospective cohort study involving burn patients (≥ 20 years old) hospitalized in the ICU of the Burn Facility in Albania during 2010-2014. Patients were categorized as having euglycaemia, moderate or critical hyperglycaemia. Regression analysis, hyperglycaemia prediction test and risk measurement were performed for the population. Statistical significance for SIH was only found for glucose values on admission, presence of diabetes and BMI. Using 180 mg/dl as cut off for critical hyperglycaemia, we found that this test had a sensitivity of 66.67% (95% CI: 44.68% to 84.33%), specificity of 88.20% (95% CI: 84.16% to 91.51%), PPV of 29.63% (95% CI: 17.99% to 43.61%) and NPV of 97.26% (95% CI: 94.67% to 98.81%). Statistical significance was found for patients with critical hyperglycaemia on admission regarding relative and absolute risk measures for sepsis and mortality. Glucose values on admission, as one of the derangement features of burn shock, are prognostic factors in critical hyperglycaemia during disease, and have a close relationship with other outcomes (sepsis and mortality).