Modelling and control of hyperglycemia in critical care patients

Abstract

Critically ill patients are known to experience stress-induced hyperglycemia. Inhibiting the physiological response to increased glycemic levels in these patients are factors such as increased insulin resistance, increased dextrose input, absolute or relative insulin deficiency, and drug therapy. Although hyperglycemia can be a marker for severity of illness, it can also worsen outcomes, leading to an increased risk of further complications. Hyperglycemia has been quantified in critically ill patients showing the need for glucose control. The development of a relatively simple system model and the verification of both generic and patient specific parameters have been successful in control trials and simulations over a range of critically ill patients. Stepwise reduction of blood glucose values by adaptive control was shown to be accurate to within 20%, and average long-term fitting errors are within the measurement error of the glucose sensor. A control algorithm capable of tight regulation for a glucose intolerant ICU patient would thus reduce mortality, as well as the burden on medical resources and time with current experience-based control approaches used in most critical care units. Overall, the research presented is a significant step towards fully automated adaptive control of hyperglycaemia in critically ill patients.