Skip to main content

Automated blood glucose control in the intensive care unit

Fig 1: Conventional insulin therapy (left) versus intensified, automated insulin therapy (right).

Fig 2: Closed loop with patient and assistance system for automated insulin therapy

Contact person

Project description

Glucose is the central energy carrier of energy metabolism in humans. During glucose transport in the blood, the body's natural glucose regulation using insulin and glucagon regulates blood glucose levels in the normoglycemic range of 100-120 mg/dL. In ICU patients, this glucose metabolism can be affected by stress due to severe infectious disease or after surgery. In this setting, these patients often suffer from elevated blood glucose levels (hyperglycemia) and severe blood glucose fluctuations (glucose variability).

To control blood levels, as well as blood glucose levels, patients have a blood gas analysis taken every 4-8h. If blood glucose levels are high, therapy can be given by external injection of insulin (see Fig. 1). In this regard, the common conventional therapy provides for intervention only when blood glucose levels are severely elevated (>180 mg/dL).

In intensive insulin therapy, blood glucose levels are measured regularly and the insulin infusion is adjusted accordingly. Closer measurement of blood glucose levels allows for better regulation of blood glucose levels and allows regulation closer to the normoglycemic range. This can reduce hyperglycemia and high glucose variability.

Diverse studies show that intensive insulin therapy can reduce mortality and length of stay in the intensive care unit. However, this type of therapy requires regular blood glucose monitoring and adjustment of insulin infusion to avoid dangerous hypoglycemia. Thus, the effort required by clinical staff is higher compared with conventional therapy.

An assistance system based on a glucose sensor, an insulin perfusor, and the control device enables safe intensive insulin therapy with reduced effort for the nursing staff (see Fig. 2). The sensor monitors the patient's blood glucose level at regular intervals, and the data are transmitted to a control unit. A control algorithm can then calculate the optimum amount of insulin from the measured blood glucose level and deliver it to the patient via a perfusor. Robust control algorithms provide a high degree of stability and safety of the controller.

Project goals

  • Modeling of blood glucose levels in patients in intensive care
  • Construct a hardware in the loop system for testing blood glucose sensors

    Project partner

    • Trace Analytics, Braunschweig
    • University Hospital RWTH Aachen, Aachen