To the Editor:—
The paper by Wietasch et al. 1describes a new technique, transcerebral thermodilution, to evaluate cerebral blood flow (CBF) at the bedside based on a double-indicator method (dye and iced water). The agreement of this new technique with the Kety-Schmidt reference method, with use of argon as a tracer gas, in patients undergoing coronary bypass surgery is reported as reasonable. In fact, the agreement of transcerebral thermodilution technique with the Kety-Schmidt method is poor, with a bias of 7 ml · min−1· 100 g−1, which is 14% of the normal value for CBF (50 ml · min−1· 100 g−1), with 95% limits of agreement of ±26 ml · min−1· 100 g−1, which are 50% of the normal values for CBF. Moreover, the authors do not report the in vivo variability for repeated measurements with the transcerebral thermodilution technique. In the intensive care setting, continuous jugular thermodilution has a better agreement with the Kety-Schmidt reference method (bias −0.9 ml · min−1· 100 g−1, with 95% limits of agreement of ±7.2 ml · min−1· 100 g−1). 2More important is the inaccuracy of the measurement at low CBF. If we look at the Bland and Altman diagram, figure 5 in the article by Wietasch et al. , 1it is obvious that the transcerebral thermodilution technique as compared with the Kety-Schmidt method underestimates CBF below 30 ml · min−1· 100 g−1. This point is of crucial importance for a technique proposed for use at the bedside in a critical care unit to monitor patients with low cerebral blood flow, which occurs in most brain-injured comatose patients. 3