Non-invasive fluid responsiveness monitoring: Patients selection considerations.

The story concerning non-invasive fluid responsiveness monitoring is not new. From many years, clinicians observed that plethysmographic waveform is generally different in patients responding to fluid therapy. Consequently, when the study of Michard et al. popularized the use of dynamic parameters for the prediction of fluid responsiveness, a great hope appeared to improve fluid management during the perioperative period.[1] Automated calculation of the plethysmographic respiratory variation with the pleth variability index (PVI), from Masimo Co, permitted to introduce the concept of continuous, non-invasive, optimization of dynamic parameters in high as in low risk patients.[2] Indeed, we confirmed that it possible and fruitful in randomized trials in major abdominal surgery.[34]

Surprisingly, despite all these evidences, in 2011, a survey showed that members of the American Society of Anesthesiology and the European Society of Anesthesiologists considered plethysmographic waveform variations as one of the worst predictor of cardiac output response following volume expansion in the context of high risk surgery.[5]

Some reasons can be proposed: First, certainly that an automated calculation with a validated device was not available for most of these physicians. Second, probably that the recent studies adding some evidence for the use of dynamic parameters, especially the PVI, may increase the interest for this kind of monitoring.[234] However third, the most important concerns were about the precision of the parameter, despite a demonstrated accuracy, rendering difficult the interpretation of the variables, even if we know that these are correct in average. Consequently, the works showing the limits of these kinds of monitoring are of prime importance. Studies showed that spontaneous ventilation,[6] vasoconstriction during the use of vasopressors[7] or during seconds after a surgical incision[8] as perioperative arythmia,[38] can limit the use of plethysmographic waveform analysis.

In this issue of the Journal, Mussa WF showed that hypercapnia, in range compatible to every day practice; however, important, lowers the PVI value compared to normocapnia.[9] These are very important data, as these may concerns many patients. However, what this study does not say is what we have to do with this lowered PVI value. At this time, we do not have the response if it is due to a decrease value but with preserved accuracy and precision. In this situation, we would have only to change the threshold in hypercapnic patients. On the contrary, accuracy could be modified, limiting the use of the PVI in these patients. Finally, the value could simply modify the preload dependency status of the patients.

In all cases, all these studies focusing on the limitations of the waveform analysis could be seen, not necessarily as an argument to not use this kind of monitoring, but rather as a chance to identify conditions were the parameter is not as precise. These works help then to trust in the PVI in the very greater number of situations that can be seen as optimal to assess it, in low as in high risk patients.