Tight glycaemic control: intelligent technology or a nurse-wise strategy?

Despite disappointing findings with the computerized decision-supported tight glycaemic control (TGC) protocol, Shulman and colleagues [1] argue that one reason to proceed with computerized TGC protocols is that complex protocols remain mandatory for TGC. Indeed, most intensivists think of TGC as difficult and complex. In The Netherlands as many as 46 different protocols are in use, including protocols with flowcharts, sliding scales, calculators and conversion tables as well as computerized decision-support protocols (survey, de Graaff MJ, Royakkers AANM, Kieft H, Spronk PE, van der Sluijs HP, Schultz MJ, unpublished data); they all are exceptionally complex and frequently difficult to follow. We recently had the opportunity to visit the Leuven hospital and were surprised to see their protocol, which is remarkably concise, far from complex, and consequently very easy to follow. In fact, their protocol is no more than a small set of written guidelines that nurses hardly ever use after having gained experience with it for some months, because (to quote the nurses) 'TGC is something you do by heart, not from a sheet of paper'.

In addition, the large number of blood glucose level (BGL) measurements (almost one BGL measurement every 75 minutes) with the tested computerized TGC protocol [1] is not comparable with the original studies of TGC [2, 3] or with current practice in Leuven. Rather, such frequent measurements may explain the high incidence of protocol violations: it was common for BGL measurements to be performed too infrequently or too late.

There is a need to standardize TGC. But let's start the easy way first: a simple protocol that forms a solid and practical basis for nurses.