Dysglycaemia, or abnormal blood‑glucose levels, has been estimated to affect up to 90% of patients in intensive care (De Block et al. 2008). Seriously ill patients often develop high blood‑glucose levels because the body reacts to trauma or surgery by producing counter‑regulatory stress hormones that cause insulin resistance. Around 215,000 adults in English NHS intensive care units had dysglycaemia in the year to March 2011. Of these, dysglycaemia followed surgical or anaesthetic procedures in 32.0% of cases and accidents in 10.1%. It is estimated that 7% of those patients with dysglycaemia in intensive care died before they were able to be discharged (Health and Social Care Information Centre 2012). Because dysglycaemia is associated with increased mortality (Dellinger et al. 2013; Jacobi et al. 2012), lowering blood‑glucose levels is the focus of care for these patients. Insulin infusions are used to lower blood‑glucose levels but they can cause glucose to drop below the healthy range, which also increases the risk of death. For this reason, glucose monitoring is needed to avoid hypoglycaemia as well as dysglycaemia (Jacobi et al. 2012).
Intensive care patients therefore benefit from regular blood‑glucose monitoring; usually this is every 1 or 2 hours until glucose values and insulin infusion rates are stable, and every 4 hours thereafter to maintain control within specified limits (Dellinger et al. 2013). A number of care protocols have been developed to achieve this, and more recent computer‑based predictive algorithms have shown better performance with fewer adverse effects than standard paper‑based protocols (Horvorka et al. 2007; Hoekstra et al. 2009).
Blood‑glucose control protocols need frequent blood sampling, and this increases the workload on nursing staff. Intermittent sampling may not always detect significant dysglycaemic events, so continuous glucose monitoring – and real‑time updating of the protocols – may provide a better and safer means to manage blood‑glucose levels in intensive care patients.