Based on our results, the present study may suggest that CI measurements obtained by means of calibrated radial arterial pulse variation catheter (ProAQT®) may be similar from those obtained with the PAC after complicated major cardiac surgery. To our knowledge, this is the first study that has evaluated the accuracy of ProAQT® sensor measurements in comparison with PAC during the postoperative course of patients who underwent cardiac surgery with cardiopulmonary bypass. A similar study was performed but in patients who underwent off-pump bypass surgery [6].
We agree that the use of PAC should be ideally restricted to the most complex hemodynamic scenarios whereas less-invasive monitoring, such as ProAQT® sensor, should be used in more stable patients [2, 13]. Hemodynamic monitoring is necessary in unstable patients after cardiac surgery for guiding fluid resuscitation, and the use of less-invasive devices is appropriate when there is a contraindication for the use of PAC or more invasive devices [1, 2]. Despite the lack of accuracy compared with the gold standard, some clinical conditions (e.g., inability to monitor the femoral artery for PiCCO2 placement) make impossible the use of invasive monitoring and they leave less invasive devices as the only choice for any type of hemodynamic monitoring. In addition, the ProAQT® sensor has the advantage of not needing an additional line placement or procedure since a radial artery is used to be placed in each patient who underwent cardiac surgery.
Similar to previous studies, our measurements obtained by an arterial pressure waveform sensor have a limited accuracy, which may be inherent to the technology of these less-invasive devices [4–6]. In order to improve accuracy of measurement as much as possible, we have used only auto-calibrated mode in the ProAQT® sensor since it seems to improve measurements, especially the trending ability of CI [5]. Indeed, the limited precision of uncalibrated measurements of CI obtained by less invasive devices has been widely reported in surgical patients [14].
It is important to point out that the technology is based on algorithms incorporating data on normal vascular anatomy and function, which is not the case of almost every patient [14]. The absolute values of CI measured by ProAQT® sensor after cardiac surgery has been shown to be reliable whereas a high percentage error has been reported in shock patients admitted to the ICU [6, 15]. Thus, our results of CI are moderately accurate, especially if we consider we have performed an evaluation in a short sample of complicated cardiac surgery patients.
The ProAQT® sensor measurements are based on arterial pressure waveform analysis and a severe vasoplegia, which is reflected by low SVRI, can influence the accuracy of these measurements [16]. It has been shown that the inaccuracy of CI measurements in ProAQT® sensors may be related with large variations of SVRI during major surgery (i.e., liver transplantations) [4]. However, vasopressor use may help to correct these variations and the influence of SVRI over CI could be minimal [16]. In addition, these variations have been reported during surgery and not during postoperative period, which could be largely influenced by hypovolemia caused by intraoperative bleeding and insensitive losses, especially in major abdominal surgeries. Despite our patients representing a complicated cohort of postoperative cardiac surgery patients needing vasopressor and inotropic support; we think that vasoplegia might have slightly influenced our results regarding SVRI since patients showed appropriate MAP and urine output during the study period.
Our study presents certain limitations. The most important are the single-centre observational nature of our study and the lower size of our sample. Despite our results should be taken cautiously, the methodology we have used to evaluate measurements of ProAQT® sensors with PAC seems appropriate [10–12, 14]. Another point of criticism could be the measurement of CI by means of continuous thermodilution instead of intermittent thermodilution, which has been considered clinical gold standard. However, continuous thermodilution monitoring of CI with PAC has proven to be accurate and precise in the critically ill patients when compared with the "standard" intermittent bolus thermodilution technique, even when hemodynamics are highly variable (e.g., during cardiac surgery interventions) [17, 18]. Indeed, bolus thermodilution CO measurements may be affected by variations in injectate volume, rate, and temperature. These variations are eliminated when CI is measured by a continuous automated thermal technique, which has been performed in our study [19].
In our opinion, a task force guided by scientific societies involving all the healthcare professionals involved in hemodynamic monitoring should establish the standard conditions for the design and development of validation studies for these types of devices.
In conclusion, our study may suggest that the ProAQT® sensor may be useful to monitor CI in patients undergoing cardiac surgery and it may provide a reliable estimate of its absolute value compared with gold standard. More studies are needed in order to validate the ProAQT® sensor and elucidate its proper use within the different clinical scenarios, as well as, provide larger evidence on its use.