Tracking of cardiac output from arterial pulse wave.

In a previous issue of Clinical Science, Remmen et al. [1] addressed the question whether the Modelflow2 method can reliably assess cardiac output (CO) from an arterial pressure waveform without calibration. This question is not new, neither is the answer: it does not, as we have shown in a series of earlier studies [2–5]. If accurate absolute values are required, the methodology needs calibration against a ‘gold standard’, such as thermodilution [2] or Fick [6]. Otherwise, CO can be expressed as changes from control with the same precision in CO tracking [4]. Uncalibrated, the model cannot replace thermodilution, but after one calibration the Modelflow2 method can do without further thermodilutions for a long period of time [2]. Even under the adverse circumstances of low arterial pressure, e.g. in severe shock, calibrated Modelflow2 COs from radial or femoral arterial pressure track over a wide range of COs with a limited error (237%) for up to 2 days without recalibration [5]. The model equally accepts intra-arterial and non-invasively determined finger pressure as input, as we have shown during prolonged orthostatic stress [4]. Modelflow2 stroke volume estimates and those from thermodilution differed neither in direction nor in magnitude (338 ml for prolonged head-up tilt and 339 ml for standing). Rather than to take issue with the methodology of Remmen et al. [1] that may have inspired the overstated title of their paper, we want to stress the basic value of the Modelflow2 method for cardiovascular research and for clinical practice. We consider it highly important that changes in CO can be tracked from an arterial pressure waveform, which can be as peripheral as the radial artery [5,7] or finger [4]. We are aware that erroneously absolute values without calibration or comment have been reported in the literature [8,9]. Therefore we want to