An audit of the impact of a change in the blood glucose targets protocol on intensive care mortality for diabetic patients admitted to the intensive care unit

None listed

Many people with type-1 and type-2 diabetes develop critical illness, which inevitably leads to deterioration in glycaemic control. However, most critically ill patients with hyperglycaemia are not diabetic, but develop disordered glucose metabolism that normalises once the acute illness resolves – so-called critical-illness induced hyperglycaemia (CIIH). While a number of studies have evaluated the effects of modulating glycaemia, using insulin, on outcomes in the critically ill, a major limitation of these studies, given recent information, is that critically ill patients with hyperglycaemia have been considered a homogenous cohort, rather than classifying hyperglycaemia in the critically ill according to pre-morbid glycaemia or presence of diabetes. Recent data indicate that a paradigm shift is required, and that patients with CIIH should be considered separately to those critically ill patients known to have diabetes, particularly those patients with ‘chronic’ hyperglycaemia, or ‘poorly controlled’ diabetes. Given these data (and that it intuitively makes sense) the medical staff at Austin Hospital Intensive Care Unit believe we should be personalising blood glucose targets and that, in diabetic patients, the target glucose level in ICU should simulate the sort of glycemic levels these patients are likely to commonly experience in their daily lives, especially during situations of physiological stress (between 10-14 mmol/L). Such target glucose level would also be expected to prevent any episodes of even mild hypoglycaemia, which have been associated with increased risk of death. However, allowing slightly increased blood glucose concentrations in critically ill patients known to have diabetes is a slight change from current practice, and evidence to support this change, although logical and strong, is from observational studies only. Accordingly, we wish to collect data after the implementation of the new protocol to audit the effects of glycaemia on outcomes and ensure that this protocol proves safe and does indeed prevent hypoglycemic episodes as expected. This audit represents a formal assessment of a quality improvement initiative dedicated at increasing the quality of glycaemic care in critically ill diabetic patients.