A common side effect of stress and trauma in critically ill patients is a rapid increase in blood glucose levels. As with diabetes, the levels can be reduced and controlled by the infusion of insulin. But glucose levels peak and change much more quickly in the ICU environment and there is little room for trial and error. If the situation is not normalised, then complications and even deaths can and do occur.
Twice in the past, Europe-wide studies and trials were put in place to try and come up with a solution to the problem. But in both cases they were prematurely halted because researchers could not solve the problem of overcompensating and patients developing hypoglycaemia, or abnormally low blood sugar levels.
"What these studies did clearly indicate is that the establishment of normal glucose levels in critically ill patients is very difficult to achieve without some sort of automated system to help the nurses", stated Dr. Martin Ellmerer, scientific co-ordinator of the CLINICIP project which has developed just such a system.
CLINICIP started by surveying ICUs in a number of European hospitals and interviewing nursing staff. "We found out that ICU staff did not want to see additional catheters in patients, they did not want extra equipment taking up space, and costs had to be kept right down so as not to eat into funds for other vital equipment", stated Dr. Ellmerer. "So, right from the start the requirements were really tough."
Partners in this European Union-funded project, academic medical institutions plus one private-sector medical equipment manufacturer, decided they needed to develop a two-step approach. "We first developed a decision-support system which met all the criteria outlined by the ICU staff, and later developed a fully automated system", Dr. Ellmerer explained.
At the heart of both systems is a sophisticated bit of computer software - an algorithm - written especially for this project. With the decision-support system, nurses still have to draw blood from patients in the traditional way and test it for glucose levels. They enter the information via the user interface - a touch screen - the researchers have developed. The algorithm takes over at this stage, calculates how much insulin is needed and automatically administers it. It also alerts the nurse when a new blood sample needs to be taken and analysed - half an hour in the worst cases and up to four hours in less severe cases.
"We have fully functioning prototypes of the decision-support system which we successfully trialled in ICUs at different hospitals around Europe", Dr. Ellmerer stated. The project's industrial partner, B. Braun Melsungen AG, is ready to go into commercial production of the system working together with the clinical partners.
"We will first have to go through an approval process and the systems should be commercially available to hospitals in mid-2009", Dr. Ellmerer stated. B. Braun is one of the leading manufacturers of infusion systems used in hospitals, and the CLINICIP technology will be incorporated into these as it was during the trials.
At the same time the prototype was being developed and tested, CLINICIP researchers were working on sensors for a fully automated, closed-loop control system to both monitor glucose levels and administer insulin with no involvement from a nurse.
The drawback of this is that a dedicated needle is necessary. "Unfortunately, this is unavoidable for a fully automated system", Dr. Ellmerer pointed out. Using fibre-optic technology the needle draws blood, sends it for analysis and then returns it to the patient's vein as well as administering the necessary dose of insulin.
"We have performed a proof-of-concept study to show we are able to establish glucose control in a clinical setting", Dr. Ellmerer stated. To develop the sensor technology further and then commercialise it, a spin-off company will be set up with Dr. Ellmerer as CEO and one of the shareholders. The other shareholders are individuals from project partners in CLINICIP. The spin-off will work closely with B. Braun and the partners, although they are not stakeholders in it.
Dr. Ellmerer expects the fully automated two-step system to be commercially available in 2011. "Our research and the products which result from it should have a pretty fundamental impact on ICUs", he staed. "They should improve survival chances, reduce complications, such as sepsis and organ failure, and reduce the time patients need to spend in ICUs."