Background: The conflicting results from studies over tight glucose control in intensive care unit (ICU) patients ask for a continuous on-line real-time glucose monitoring in future. Here, intravenous microdialysis was tested in ICU patients and healthy volunteers. Primary aims were technical feasibility and accuracy.
Methods: A microdialysis catheter was inserted into a peripheral vein. ICU patients (n=10) were studied for up to 5 days. Healthy volunteers (n=6) were studied on one occasion. Recordings were monitored during 70 min each 24-h period to allow for an estimate of variability over time. Microdialysis glucose and lactate were compared with plasma glucose and whole blood lactate. Results are presented as medians (quartiles) of the differences between microdialysis and plasma concentrations over each of the 70-min recording periods.
Results: Out of the included ICU patients, no exclusions or early terminations were due to failure of the microdialysis catheter. The concordance was highly variable. The difference of medians over the recording periods differed by -34% (-40, -16) in patients and -22% (-31, -15) for the volunteers. In contrast, the overall variability within the individual measurement periods was low.
Conclusion: Technical feasibility was good, but the accuracy was not sufficient and the variability between the recording periods was high without calibrations. The non-availability of suitable peripheral veins was a problem in many patients screened but not included in the study. Intravenous microdialysis to obtain continuous on-line real-time glucose monitoring is technically feasible, but accuracy needs to be improved.