Noninvasive continuous versus intermittent arterial pressure monitoring: evaluation of the vascular unloading technique (CNAP device) in the emergency department

Adequate monitoring of cardiovascular function in acutely ill patients is of outstanding importance, as belatedly detected hemodynamic instability might worsen the outcome of these patients [1]. In high dependency unit and intensive care unit (ICU) patients, arterial pressure (AP) is therefore monitored continuously using an arterial catheter placed in the radial or femoral artery. As this invasive method of AP monitoring is associated with risks [2] and requires specially trained medical staff, its application is limited and not suitable for broad use in the emergency department (ED). Patients requiring AP monitoring in the ED are, therefore, routinely connected to automated oscillometric devices which measure AP intermittently at defined time intervals (depending on the patient’s clinical condition). However, oscillometry does not meet the need for continuous AP monitoring. During the last years, methods for continuous noninvasive AP monitoring have become available for evaluation in clinical studies [3‐17]. One technology for noninvasive continuous AP assessment is the vascular unloading technique (VUT) [3‐11]. The CNAP technology (CNSystems Medizintechnik AG, Graz, Austria) is one of two commercially available AP monitoring systems that use VUT. Previous studies have described its application during general anesthesia and showed clinically acceptable agreement between VUT using the CNAP technology and invasive continuous AP monitoring [5, 7, 8]. Furthermore, the CNAP technology was evaluated during medical procedures, e.g., interventional endoscopy [18] or spinal anesthesia for caesarean section [19]. The aim of the present study was to compare AP values recorded with VUT (CNAP technology) over a 2-hour period with intermittently measured AP values obtained using oscillometry in acutely ill ED patients. In addition, we aimed to investigate whether continuous noninvasive AP monitoring allows detection of relevant hypotensive episodes that might be missed with intermittent AP monitoring.

For the evaluation of agreement between the noninvasive continuous AP measurements using VUT and the intermittently obtained oscillometric AP measurements, 1,155 paired AP measurements in 130 patients were statistically analyzed. The mean value (±standard deviation) of VUT-derived AP was 125 mmHg (±27 mmHg) for SAP, 71 mmHg (±15 mmHg) for DAP, and 92 mmHg (±19 mmHg) for MAP, respectively (Table 2). The mean value (±standard deviation) of AP obtained by the oscillometric method was 130 (±24 mmHg), 74 (±17 mmHg), and 97 mmHg (±20 mmHg) for SAP, DAP, and MAP, respectively. The mean difference (±standard deviation; 95% limits of agreement) between AP measurements obtained by VUT and oscillometric AP measurements was -5 mmHg (±22; -47 to +37 mmHg) for SAP, -2 mmHg (±15; -32 to +27 mmHg) for DAP, and -6 mmHg (±16; -37 to +26 mmHg) for MAP, respectively. Figure 1 shows the corresponding Bland-Altman plots for SAP, DAP, and MAP.

In the interval between two oscillometric measurements, VUT detected hypotensive episodes (≥4 minutes) defined as either SAP <90 mmHg or MAP <65 mmHg at least once in 30 patients (with a total of 67 episodes) and 16 patients (with a total of 27 episodes), respectively. The median duration of these episodes was 6.7 (4.5 – 12.1) and 5.7 (4.8 – 8.7) minutes, respectively. In 3 (SAP <90 mmHg) and 2 (MAP <65 mmHg) measurements in 2 and 1 patients, respectively, the recorded hypotensive episodes lasted at least 45 minutes. In 11 (SAP <90 mmHg) and 6 (MAP <65 mmHg) of the patients (i.e., in 25 and 7 of the episodes, respectively), hypotension was also detected by the subsequent intermittent oscillometric AP measurement. As the time interval between two oscillometric measurements was 15 minutes, the detection of a hypotensive episode by VUT took place with an average time advantage of 8.1 minutes for MAP <65 mmHg and 9.5 minutes for SAP <90 mmHg compared to the subsequent oscillometric measurement.

This exploratory study compares continuous noninvasive AP monitoring obtained by VUT (CNAP technology) with the method of intermittent oscillometric AP measurements which is widely used and generally accepted by health care professionals. Thus, the oscillometric method is also routinely used and considered as criterion standard for AP monitoring in acutely ill patients presenting to the ED. One major limitation of this diagnostic device is, however, that it does not provide uninterrupted monitoring of the patient’s AP behavior. Particularly in hemodynamically unstable patients, only continuous AP recordings including real-time visualization of the AP waveform allow the treating ED staff to rapidly identify hypotensive phases and treat the patient accordingly in order to avoid permanent organ damage due to hypoperfusion. In clinical routine, the criterion standard for continuous AP monitoring are arterial catheter-derived AP measurements usually reserved for ICU patients as invasive AP measurements are often not practically applicable in the ED. However, in specific medical conditions, the first hours that a patient often spends in the ED when admitted to the hospital might be crucial when it comes to physical long-term damage and survival. Kumar and colleagues demonstrated the importance of early detection of hypotension in septic shock patients, as the administration of antimicrobial therapy within the first hour of recorded hypotension was associated with a significantly higher survival rate [1]. Given the possible damage caused by unregistered episodes of hypotension, but also the limitations in establishing invasive AP monitoring in the ED, a continuous noninvasive AP monitoring technology like the CNAP system might help to increase patient safety during the first hours of medical treatment. There already exist validation studies comparing CNAP-derived AP with invasively assessed AP measurements that reported clinically acceptable agreement for SAP, DAP, and MAP and, therefore, suggested CNAP as an alternative for invasive AP monitoring [5, 8]. Furthermore, recent studies were able to demonstrate the potential value of continuous noninvasive AP monitoring by VUT using the CNAP device compared with intermittent oscillometric AP monitoring: Ilies and colleagues detected more hypotensive episodes after spinal anesthesia in women requiring caesarean section using the CNAP device than with the intermittent AP measurements [19]. In a study by Siebig and colleagues, rapid AP changes during sedation in interventional endoscopy patients were identified using VUT for continuous AP monitoring which were not detected by the standard oscillometric method [18].

Nevertheless, the comparability of the noninvasive device with the current criterion standard is important for its use in clinical practice. Our method comparison study revealed a satisfactory agreement between noninvasive AP monitoring using VUT and intermittent oscillometric AP measurements resulting in a bias (95% limits of agreement) of -5 mmHg (-47 to +37 mmHg) for SAP, -2 mmHg (-32 to +27 mmHg) for DAP, and -6 mmHg (-37 to +26 mmHg) for MAP, respectively.

In a similar study, Nowak and colleagues compared intermittent oscillometric AP measurements with noninvasive continuous AP measurements using the Nexfin device (BMEYE, Amsterdam, The Netherlands) that also uses VUT in 40 patients treated in the ED [22]. In this trial, a mean difference and 95% limits of agreement of +0.87 mmHg and -45.8 to +47.54 mmHg for SAP, of -1.24 mmHg and -31.74 to +29.27 mmHg for DAP, and of -2.05 mmHg and -33.83 to +29.74 mmHg for MAP were observed. The respective bias for SAP, DAP, and MAP in the study of Nowak and colleagues is lower indicating a higher accuracy of the noninvasive device compared with our results, whereas the limits of agreement and therefore the precision are reasonably comparable. Among other things, these differences may be explainable by the fact that the CNAP device uses automated calibration to upper arm oscillometric AP whereas the Nexfin device uses a different technique [10]. To what extent the calibration method has an impact on the accuracy and precision of VUT is the subject of further studies. Regarding the types of diagnoses, the patients investigated in the study of Nowak et al. were comparable with those included in our trial with a smaller total number of cases enrolled in this previous study as compared with our work. In contrast to our study, the exclusion criteria of Nowak and colleagues did not take into account AP differences between the right and left upper arm which might have had an impact on study results as they also used one patient’s upper extremity for oscillometric measurements and the contralateral one for continuous noninvasive monitoring.

In addition to the sole comparison of methods, we analyzed AP recordings of each patient for hypotensive phases in the interval between two oscillometric measurements. For this purpose we deliberately chose 4 minutes as cut-off value in order to ensure that only clinically relevant hypotensive phases are identified.

Several limitations of our study have to be mentioned to ensure the correct interpretation of our results. Firstly, we did not define the limits of agreement a priori. Although relevant guidelines for the comparison of two AP measurement methods from the Association for the Advancement of Medical Instrumentation (AAMI) exist, these guidelines explicitly do not cover finger AP measurement devices and no other guidelines addressing this type of measurement comparison studies are available [23]. Considering this fact, it would have been arbitrary to define limits of agreement a priori. Secondly, critically ill patients who presented hemodynamically unstable were immediately admitted to the ICU, provided that an ICU bed was available. Therefore, patients included in our study were relatively hemodynamically stable.

According to our study, noninvasive continuous AP assessment using VUT might be a promising monitoring option for acutely ill ED patients with the potential for increasing patient safety. An application of the noninvasive continuous AP monitoring technique in the ED is conceivable because devices using VUT are easy to apply and safer in contrast to invasive AP measurements. Therefore it seems to be particularly suitable in the ED as in many patients there is no clear indication for invasive AP monitoring or time until invasive AP measurements become possible has to be bridged.

In conclusion, VUT using the CNAP system for noninvasive continuous AP measurement shows reasonable agreement with intermittent oscillometric measurements in acutely ill ED patients. Continuous AP monitoring allows immediate recognition of clinically relevant hypotensive episodes, which are missed or only belatedly recognized with intermittent AP measurement.