Poor agreement between cardiac output measurements from the pressure recording analytical method and pulmonary artery thermodilution in patients with low cardiac output: preliminary results from a prospective observational method comparison pilot study

Background: Continuous hemodynamic monitoring is essential for guiding goal-directed therapy in high-risk cardiac surgery patients, particularly those prone to low-output syndrome. The pressure recording analytical method (PRAM) is a unique, uncalibrated pulse contour analysis technology that estimates cardiac output (CO) by directly analyzing the arterial pressure waveform.

Objectives: The primary objective of this prospective observational pilot study was to investigate the feasibility of the PRAM method (test method) in a challenging, high-risk cardiac surgery population defined by a severely impaired left ventricular ejection fraction (LVEF < 35%). The exploratory objective was to perform a preliminary analysis of the agreement between PRAM-derived and pulmonary artery catheter (PAC)-derived continuous cardiac output (FastCCO) measurements.

Design: This was a prospective, observational method comparison pilot study.

Methods: Seven patients (LVEF < 35%) undergoing coronary artery bypass surgery were included. Concurrent measurements (n = 826) were collected via radial artery (PRAM) and PAC. To account for repeated measures, agreement was assessed using linear mixed-effects (LME) modeling, and 95% confidence intervals (CI) were derived using 100,000 bootstrap statistics. Trending ability was assessed via four-quadrant and polar plot analysis.

Results: Feasibility was high (98.8% data acquisition). However, the LME constant bias model revealed a significant population-level underestimation by PRAM of −2.02 L/min (limits of agreement: −5.69 to 1.64 L/min). The bias-corrected and accelerated-bootstrapped percentage error was 134.3% (95% CI: 122.7–148.3%), and the least significant change was 4.48 L/min (95% CI: 4.31–4.74 L/min). Trending ability was severely compromised, with a four-quadrant concordance rate of 36.6% and a polar concordance rate of 10.1%.

Conclusions: While feasible in terms of data acquisition, the PRAM method, when applied via a peripheral arterial site, showed poor agreement and unreliable trending ability compared with PAC FastCCO in cardiac surgery patients with severely reduced left ventricular function. These findings suggest that the complex pathophysiology, particularly the associated severe low ejection fraction, may compromise the accuracy of PRAM in this high-risk population. Further investigation is warranted to understand the influence of these specific conditions on the PRAM algorithm.

Clinical Trial Registration: https://drks.de/search/en/trial/DRKS00017260/details.

1 Introduction

Coronary artery bypass graft (CABG) surgery after myocardial infarction (MI) can be a potentially life-saving procedure, but it can be complicated by low-output syndrome (LOS), which is associated with high morbidity and mortality (1–3). Goal-directed vasopressor and inotropic therapy may improve outcome, and through continuous hemodynamic monitoring techniques, cardiac output (CO) provides the anesthetist with a target for intervention (4, 5). Traditionally, pulmonary artery catheters (PACs) have been the gold standard for hemodynamic monitoring in critically ill patients (6). While PAC provides comprehensive insights into cardiac preload and afterload, its use is limited by its invasive nature and potential for complications (7).

This has driven the development of less-invasive alternatives for CO estimation. A class of less-invasive techniques relying on analysis of the arterial pressure waveform to estimate various hemodynamic parameters is called pulse contour analysis (8). This approach estimates hemodynamic parameters by deriving stroke volume (SV) from the area under the systolic portion of the arterial pressure wave, typically by using a mathematical model of the systemic vascular system and adjusting for a measure of aortic impedance (8).

Established pulse contour methods often require an initial calibration against a reference method (like thermodilution) or rely on pre-estimated demographic data (e.g., age and sex) to estimate vascular impedance (8). Consequently, the accuracy of these systems can be susceptible to factors like severe arrhythmia and changes in vascular tone (9, 10). A unique and newer approach within this class is the pressure recording analytical method (PRAM). Unlike traditional pulse contour methods, PRAM is an uncalibrated technology that continuously analyzes the arterial pressure waveform (11) to derive the patient's unique instantaneous systemic impedance, Z(t), directly from the morphology of the pressure wave in a beat-to-beat fashion, focusing on specific points of instability within the waveform (12). The PRAM algorithm integrates the effects of upstroke and vascular compliance across the entire cardiac cycle to calculate SV based on the instantaneous pressure gradient (12). This unique, a priori modeling allows PRAM to theoretically provide accurate CO measurements without external calibration or reliance on demographic data, even during dynamic conditions where vascular properties are rapidly changing (11, 12). Several papers have been published on the accuracy of PRAM, showing a general level of agreement using both transesophageal echocardiography and PAC as reference methods (11, 13–15). Notably, volume status does seem to play a role in PRAM, as shown in animal studies (16). A recent study in veno-venous extracorporeal membrane oxygenation (ECMO) patients by Greiwe et al. (15) showed agreement between TEE and PRAM within the criteria proposed by Critchley and Critchley (17). These commonly used guidelines for comparing cardiac output measurements state that new methods should be judged against an accuracy of up to 30% in the percentage error (PE) (17). A newer take on this is the recently published COMPARE Statement (18), which additionally suggests identification of the trending ability and systematic offsets rather than a fixed percentage error.

With these promising results, the performance of PRAM can be further investigated in challenging populations with hemodynamic instability, such as patients undergoing CABG with a severely impaired left ventricular ejection fraction (LVEF < 35%).

It was the primary objective of this prospective observational pilot study to investigate the feasibility of the PRAM method (test method) in this high-risk cardiac surgery population by assessing the ability to obtain high-quality, comparable, paired PRAM and PAC measurements. We hypothesized that comparable measurements could be obtained in >90% of measurement cycles.

The exploratory objective was to perform a preliminary analysis of the agreement between PRAM-derived (test method) and PAC-derived (reference) CO measurements.

2 Materials and methods

2.1 Participants and recruitment

This study adheres to the previously published COMPARE Statement (18). This feasibility study was a sub-study of a larger observational study investigating the use of near-infrared spectroscopy for the early detection of heart failure after CABG in the sense of low cardiac output syndrome. Since the study was based on PAC measurements as the reference method for CO in patients with low left ventricular ejection fraction, we included PRAM measurement in the existing study. The study was approved by the local ethics committee (No. 119/19) and was registered with the German Clinical Trial Registry (DRKS00017260); final inclusion comprised 44 patients (19). However, the PRAM method was introduced late in the study period, and patient inclusion was only initiated in October 2022. The inclusion criteria included elective cardiac surgery, a preoperative left ventricular ejection fraction of ≤35%, and patient consent. The exclusion criteria included inability to place a PAC, non-elective emergent cardiac surgery, or a preoperative left ventricular ejection fraction of >35%.

2.2 Setup

After written informed consent at least 1 day prior to surgery, patients were included. Before anesthesia induction, an invasive arterial line was placed in the awake patient using ultrasound in the radial artery. After induction, patients received standard American Society of Anesthesiologists (ASA) monitoring. All patients received a standard induction with sufentanil, propofol, and rocuronium, followed by maintenance with sevoflurane and sufentanil. Inotropes, vasopressors, and fluids were administered in accordance with the mean arterial pressure (MAP) target of >65 mmHg until heart–lung bypass. The changes were recorded manually into the electronic anesthesia protocol by the anesthetist. Central venous lines were placed under ultrasound guidance in the right internal jugular vein. A PAC was inserted with the catheter being placed in a wedge position. Transesophageal echocardiography was performed as standard of care prior to surgery to estimate ejection fraction and valve insufficiency. Transfusions of blood, platelets, and fresh frozen plasma were administered in accordance with standard care. Where possible, patients were extubated in the operating room (OR) prior to being discharged to the cardiothoracic intensive care unit.

CO was measured via PAC with an Edwards Lifesciences HemoSphere monitor (Edwards Lifesciences, Irvine, USA) using the FastCCO algorithm via a special PAC catheter (Continuous Cardiac Output Thermodilution Catheter, 139HF75; Baxter/Edwards Critical Care, Irvine, USA). The MostCare Up (Vygon GmbH & Co. KG, Aachen) was connected to the arterial line via a three-way switch using the MostCare Up transducers. The PRAM method requires accurate detection of the dicrotic notch such that integration of systolic and diastolic portions of the curve is available to the algorithm (12). For this, the MostCare Up detects the dicrotic notch position and displays a marker but allows the user to introduce a shift to optimize the results visually.

Prior to measurement, dampening of the measurement setup and detection of the dicrotic notch were optimized to ensure good visual detection results. Measurements for the study began once the patient was positioned in the OR. Correct positioning of the PAC was confirmed by obtaining wedge pressure measurements before incision and during suturing. The data were then exported using the export functions to removable storage devices from the respective devices for further processing. The clinical data were obtained from the electronic health record systems (Dräger Integrated Care Manager and Dedalus ORBIS).

2.3 Sample size estimation

As this was a feasibility study, we did not perform a sample size calculation prior. We expected to achieve feasibility by having >90% concurrent measurements.

2.4 Statistical analysis

Statistical analysis was performed using MATLAB R2022b (MathWorks, Natick, MA, USA). Data from the PAC and the PRAM were synchronized and averaged into 1 min intervals to ensure temporal alignment. Given the hierarchical nature of the data, characterized by multiple repeated measurements nested within individual subjects, a linear mixed-effects (LME) modeling approach was employed to account for within-subject correlation. Agreement was evaluated using two distinct LME models. First, a constant bias model was defined as Difference ∼ 1 + (1|Patient). This model allowed for the partitioning of total variance into between-subject and within-subject components to derive adjusted limits of agreement (LoA). Second, a proportional bias model was utilized to determine if the measurement error was dependent on the magnitude of the cardiac output Difference ∼ MeanCO + (1|Patient). The significance of the slope was assessed to identify systematic divergence using the likelihood ratio test as employed by MATLAB R2022b.

Precision and diagnostic error were further quantified using a bootstrap resampling technique and the first constant bias model. We performed 100,000 iterations of random resampling of all measurement points to generate robust 95% confidence intervals (CI) for the PE and the least significant change (LSC). To account for potential skewness, the bias-corrected and accelerated (BCa) method from MATLAB R2022b was utilized for all bootstrap interval calculations.

Trending ability was evaluated through a combination of four-quadrant and polar plot analyses. For the four-quadrant plot, an exclusion zone of ±0.3 L/min was applied to filter out clinically insignificant fluctuations. Concordance was defined as the percentage of data points where the direction of change in PRAM-derived CO matched the reference PAC-derived change. For the polar plot analysis, an angular exclusion zone (AEZ) of ±30° around the line of identity (45° and 225°) was established. The polar concordance rate was calculated as the percentage of points falling within these boundaries, reflecting the method's ability to accurately track both the direction and magnitude of hemodynamic shifts.

The clinical data for the cohort were summarized by age, height, weight, and pre-existing conditions. Finally, volume balance and preload were provided via the central venous pressures prior to thoracotomy and the fluid balance during surgery and after the first day of postoperative ICU treatment. To determine volume balance and preload, we provide the central venous pressures prior to thoracotomy, the fluid balance during surgery, and the fluid balance after the first day of postoperative ICU treatment.

3 Results

3.1 Patient characteristics

Patients were recruited at the University Hospital Bonn between October 2022 and January 2023. During this period, a total of nine patients with an EF < 35% were enrolled. All received CABG surgery for severe triple vessel disease. Two patients had to be excluded from the analysis because positioning of the PAC correctly in the wedge position for the duration of the operation was not possible. Table 1 shows the characteristics of the patients included in the study with pressures reported.

Table 1

Table 1. Patient characteristics reported as mean ± standard deviation or as count of true (false).

The patient collective included individuals with chronic obstructive pulmonary disease, which may affect the lungs’ vascular system due to remodeling effects. Additionally, valvular disease affecting all valves was present with mitral valve insufficiency being the most common. Furthermore, most patients were arrhythmic with one patient in atrial fibrillation (AF) and three patients with significant ventricular premature beats.

3.2 Method comparison and trending ability

When looking at the measurements, except for one off-pump case, most patients required on-pump surgery. Thus, measurements were primarily taken prior to bypass in the OR. Overall, a median of 121 valid measurement points per patient was included with values ranging from 60 to 188 valid measurements. This shows an overall rate of 98.8% of valid measurements outside of bypass, thus fulfilling the criterion for the feasibility of comparing PRAM and PAC. We then investigated the bias and trending ability in cardiac output estimation between PRAM (test method) and PAC (reference method). The LME-based Bland–Altman analysis revealed a variation in cardiac output estimates between PRAM and PAC. The constant bias model identified a fixed underestimation by the PRAM method, with a mean bias of −2.02 L/min. The associated limits of agreement were wide, spanning from −5.69 L/min to 1.64 L/min. The proportional bias model confirmed that the differences observed were not uniform across the measurement range with a highly significant proportional bias detected (p < 0.0001) with a slope of −1.04, demonstrating that the degree of underestimation by PRAM exacerbated as the cardiac output increased. The results can be seen in Figure 1.

Figure 1

Figure 1. Bland–Altman analysis of all patients with mean difference marked as black lines and the upper and lower limits of agreement given by 1.96 × standard deviation as red lines. On the right-hand side, the constant bias-based LME model is shown. On the left-hand side, the proportional bias model of the.

The bootstrap statistic for the error metrics showed a mean percentage error of 134.28%, with BCa-adjusted 95% confidence intervals ranging from 122.70% to 148.25%. Similarly, the mean least significant change was 4.48 L/min with BCa-adjusted 95% confidence intervals ranging from 4.31 to 4.74 L/min. These values are summarized in Table 2.

Table 2

Table 2. Summary of measurement metric from LME models.

The trending performance of PRAM was assessed using four-quadrant analysis after applying the 0.3 L/min exclusion zone (17, 20); the concordance rate was 36.58%. The polar plot analysis (18) yielded a polar concordance rate of 10.07% within the ±30 ° AEZ, suggesting that PRAM failed to accurately replicate the vector of change in the majority of hemodynamic transitions. Figure 2 shows the polar plot.

Figure 2

Figure 2. Polar plot of all measurement results between PAC CO and PRAM CO. Trending ability can be deduced from this.

Additionally, we report the initial central venous pressure prior to thoracotomy and the fluid balances at the end of surgery and by the first postoperative morning in the ICU. All patients had received significant amounts of fluids, resulting in a net positive balance. We report Ea, the arterial elastance, as reported by the MostCare Up monitor. The patients show values of vasoconstricted and dilated states with Patients 2, 3 and 4 closest to the normal range. These are summarized in Table 3. The supplement includes the dynamic Ea (SSV/PPV) additionally, which was in the normal range.

Table 3

Table 3. The table metrics of fluid state.

4 Discussion

In our small pilot study, we report a good ability to obtain corresponding measurements between the reference PAC CCO method and the PRAM method of 98.8% of valid PAC CCO measurements. However, our preliminary data show a bias between PAC and PRAM in patients undergoing CABG with an initial low ejection fraction. These results contrast with previously published work on PRAM by Zangrillo et al. (21), who used thermodilution with a cold bolus. In contrast, we used the Fast CCO functionality provided by the monitor in our patients. However, Zangrillo et al.'s cohort was restricted to patients in sinus rhythm, and preload was optimized prior to measurement without further specification. This was not done in our collective as routine care was performed. In this pilot study, only patients with low ejection fraction were included which limits the interpretation of the findings to these patients. Additionally, there was no restriction concerning arrhythmia at inclusion. Thus, there may be several reasons for the discrepancies compared with other studies.

Considering the measurement site, patients received one central CO estimation via PAC in the small circulatory system and one peripheral CO estimation via PRAM. Existing research shows that measurements located in the small circulatory system may be affected by issues due to regurgitation in patients with valvular disease (17, 22). Similarly, arrhythmia and ventricular extrasystolic beats can affect CO measurements in the patient population with an already severely reduced ejection fraction due to suboptimal preload via the Frank–Starling mechanism and may exacerbate the mixing issues (23).

Peripheral CO measurement using PRAM is highly dependent on the shape of the radial arterial waveform (12), which can be altered by several factors. Animal models with the PRAM method have shown a large dependence on volume status (16). Additional factors to be considered include disease of the peripheral arterial system which may alter both compliance and elasticity (24–26), activation of the renin–angiotensin system which results in increased vascular resistance in patients with heart failure (27, 28), the site of measurement due to varying distance from the central waveform (29), and measurement effects such as dampening of the arterial line (30–32). Previous studies have found a high prevalence of dampening in the ICU (30) and have demonstrated its effect on CO measurement (32, 33). Flushing test for optimization and identification of damping exist (34), and Saugel et al. (35) has recently proposed a five-step process for accurately measuring invasive blood pressure. With regard to the measurement site, an observational study showed disagreement between radial and femoral cardiac output estimation (36). This was also seen in liver transplant patients using pulse contour analysis vs. PAC (37). With respect to the results presented in this study, the waveform shape was continuously monitored not only by the research team but also by experienced cardiothoracic anesthetists who were present. Additionally, the dicrotic notch detection was monitored during measurements to ensure the best possible inputs were maintained throughout. However, neither the HemoSphere nor the MostCare Up monitoring devices provides the ability to export the waveform used by the algorithm via the built-in export features to allow for post hoc investigation of the dependence on the underlying input signal. The HemoSphere provides an SQI indicator ranging from 1 to 4. Retrospectively, investigation of the included results without prior selection showed values ≥3. In our measurements, the MostCare Up PRAM algorithm seemed to underestimate CO on a net basis compared with the FastCCO.

Additionally, the PRAM does not seem to be able to follow the same trends in CO estimation compared with FastCCO. The patient collective in which the measurements were performed had several factors that may be attributed to this. Specifically, intrinsic factors such as changes in the elasticity and compliance of the arterial system due to atherosclerosis or the effects of left ventricular insufficiency can lead to alterations of the propagation of the pressure wave. This is in part indicated by the measure Ea values by the MostCare Up, which are largely outside of the range described by the manufacturer as the normal range from 1.10 to 1.40 (24, 38–40). The LME analysis using the proportional bias showed a systematic error with increasing CO. This effect may be related to the alterations of the pressure wave in this critically ill population and varying vascular resistance compared with healthy subjects.

Considering that our patient collective was in the operating room for CABG, some degree of atherosclerosis most likely will have been present, and two patients already had peripheral arterial disease described. Furthermore, given the low ejection fraction, it should be assumed that all patients included have heart failure with left ventricular insufficiency.

In summary, under the best possible circumstances regarding the inputs of the measurement devices, this study suggests that the PRAM vs. PAC CO measurements may not align in patients with low cardiac output. The PRAM method has been shown to provide accurate results within the proposed limits of <30% in critically ill patients such as those on veno-venous ECMO support (15), after cardiac surgery (14), and in critically ill patients receiving vasopressor and inotropic therapy (16). Yet, another study (41) found similar disagreement between PRAM and thermodilution bolus measurements, attributing the results to suboptimal arterial input curves (33, 34). Thus, maybe the differences can be attributed to the underlying physiology itself. In various types of shock, the shift in perfusion commonly seen allows the body to preserve the function of central organs such as the heart and brain. For this, sympathetic tone is often increased in response to activation of central baroreceptors to increase peripheral vascular resistance and reduce flow (42). While the cardiac output, as seen in the small circulatory system of the heart and lung, stays constant, the resulting increase in systemic resistance reduces peripheral end-organ perfusion. During anesthesia, hypnotic agents lead to a reduction in sympathetic tone and vasodilation with vasopressors and inotropes balancing this. The MostCare Up allows for monitoring of this using Ea, the arterial elastance, which was outside of the normal range described by the device for all but two patients. Thus, this may explain why the central PAC measurement provides an increased CO estimate compared with the peripheral PRAM estimation.

The main limitation of this case series is the small number of patients included. This is due to the highly selective cohort based on multiple factors such as admission to a tertiary care center with severe vessel disease and ischemic cardiomyopathy, and thus, the results may not be generalizable. Future validation of the results is required, but they are indicative of the need to further our understanding of the pathophysiology and its potential effects on measurement. Especially, the effect of the measurement site and the underlying vascular conditions on the cardiac output estimation is poorly understood.

Additionally, these critically ill patients required constant adjustment of vasopressors, inotropes, and fluid therapy during CABG. Thus, steady state measurements in which the consistency of the cardiac output estimation method with respect to itself could not be assessed. This is an important missing step and would require longer measurements in these patients ideally prior to thoracotomy and surgical manipulation after induction. This constant change might be the reason for the poor trending ability.

Overall, with these results, we would like to propose the need for notification of failure of measurement. One possibility would be to notify the user of a poor input signal in such a way that allows them to detect insufficient inputs using a range of quality metrics. Another possibility would be to augment the signal such as proposed by Michard (43). Given that clinicians use advanced hemodynamic monitoring to guide their therapy on a daily basis, we need to understand the scenarios in which our measurement tools may introduce bias due to the nature of the input data, the intrinsic system, or the deranged physiology in which compensation mechanisms may play a major role. Since the description of the first measurements of pulse wave velocity in humans was recorded in 1922 (44), technology has enabled us to further our knowledge and treatment of diseases. Previously, approaches of mathematical mechanistic modeling of the arterial system have been hampered by computational power (45). Applying such models to the diseased states may allow us to gain a better understanding of the underlying mechanisms, not only in terms of physiology but also with regard to measurement error (46, 47). First efforts in this direction have been made using a database to explore the effects of fundamental arterial system parameters, such as compliance, on the arterial waveform (48). These efforts should be supported further to improve CO algorithms by allowing a better quantization of the observed system. Animal experiments or even in vitro experiments involving different vascular systems may give us a better understanding of the compound effects on the pulse pressure waveform. This may lead to more accurate measurements in critically ill patients.

5 Conclusion

To our knowledge, this is the first report to compare cardiac output measurements in patients undergoing coronary artery bypass graft surgery with a severely reduced ejection fraction (<35%). While the feasibility of comparison could be confirmed, results show that cardiac output estimation remains challenging in this group of patients. The results revealed differences between the measurements with mean percentage errors >30% in all patients with severely reduced left ventricular ejection fraction. This may be due to numerous factors ranging from the underlying physiology to the quality of the input signal at the site of measurement. Modern signal processing technologies such as signal augmentation could correct potential effects in this regard. Additionally, clinicians need to be made aware when estimation fails or when there is high uncertainty in the measurement. Further study of this subject is warranted to better understand the differences between central and peripheral measurements and the effects of pathophysiological considerations on measurements.

Data availability statement

The data supporting the conclusions of this article will be made available by the authors in anonymized form.

Ethics statement

The studies involving humans were approved by Ethikkommission an der Medizinischen Fakultät Bonn. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

Author contributions

MO: Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing – original draft, Writing – review & editing. JK: Investigation, Methodology, Writing – original draft, Writing – review & editing, Project administration, Conceptualization, Resources. MS: Conceptualization, Investigation, Methodology, Project administration, Resources, Writing – original draft, Writing – review & editing, Data curation, Funding acquisition, Supervision. CN: Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Writing – original draft, Writing – review & editing. SK: Conceptualization, Funding acquisition, Methodology, Project administration, Resources, Supervision, Validation, Writing – original draft, Writing – review & editing. MT: Investigation, Methodology, Project administration, Supervision, Writing – original draft, Writing – review & editing, Formal analysis, Visualization.

Funding

The author(s) declared that financial support was not received for this work and/or its publication.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence, and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fanes.2025.1718771/full#supplementary-material

References

1. Lomivorotov VV, Efremov SM, Kirov MY, Fominskiy EV, Karaskov AM. Low-cardiac-output syndrome after cardiac surgery. J Cardiothorac Vasc Anesth. (2017) 31(1):291–308. doi: 10.1053/j.jvca.2016.05.029

PubMed Abstract | Crossref Full Text | Google Scholar

2. Pérez Vela JL, Jiménez Rivera JJ, Alcalá Llorente MÁ, González de Marcos B, Torrado H, García Laborda C, et al. Low cardiac output syndrome in the postoperative period of cardiac surgery. Profile, differences in clinical course and prognosis. The ESBAGA study. Med Intensiva. (2018) 42(3):159–67. doi: 10.1016/j.medin.2017.05.009

PubMed Abstract | Crossref Full Text | Google Scholar

3. Schoonen A, van Klei WA, van Wolfswinkel L, van Loon K. Definitions of low cardiac output syndrome after cardiac surgery and their effect on the incidence of intraoperative LCOS: a literature review and cohort study. Front Cardiovasc Med. (2022) 9:926957. doi: 10.3389/fcvm.2022.926957

PubMed Abstract | Crossref Full Text | Google Scholar

4. Kapoor PM, Magoon R, Rawat RS, Mehta Y, Taneja S, Ravi R, et al. Goal-directed therapy improves the outcome of high-risk cardiac patients undergoing off-pump coronary artery bypass. Ann Card Anaesth. (2017) 20(1):83–9. doi: 10.4103/0971-9784.197842

PubMed Abstract | Crossref Full Text | Google Scholar

5. Patel H, Shah R, Parikh N, Patel R, Thosani R, Shah P, et al. Effect of goal-directed hemodynamic therapy in postcardiac surgery patients. Indian J Crit Care Med. (2020) 24(5):321–6. doi: 10.5005/jp-journals-10071-23427

PubMed Abstract | Crossref Full Text | Google Scholar

6. Saugel B, Vincent JL. Cardiac output monitoring: how to choose the optimal method for the individual patient. Curr Opin Crit Care. (2018) 24(3):165–72. doi: 10.1097/MCC.0000000000000492

PubMed Abstract | Crossref Full Text | Google Scholar

7. Mehta Y. Newer methods of cardiac output monitoring. World J Cardiol. (2014) 6(9):1022. doi: 10.4330/wjc.v6.i9.1022

PubMed Abstract | Crossref Full Text | Google Scholar

8. Grensemann J. Cardiac output monitoring by pulse contour analysis, the technical basics of less-invasive techniques. Front Med. (2018) 5:64. doi: 10.3389/fmed.2018.00064

PubMed Abstract | Crossref Full Text | Google Scholar

9. Marik PE. Noninvasive cardiac output monitors: a state-of the-art review. J Cardiothorac Vasc Anesth. (2013) 27(1):121–34. doi: 10.1053/j.jvca.2012.03.022

PubMed Abstract | Crossref Full Text | Google Scholar

10. Oh C, Lee S, Oh P, Chung W, Ko Y, Yoon SH, et al. Comparison between fourth-generation FloTrac/Vigileo system and continuous thermodilution technique for cardiac output estimation after time adjustment during off-pump coronary artery bypass graft surgery: a retrospective cohort study. J Clin Med. (2022) 11(20):6093. doi: 10.3390/jcm11206093

PubMed Abstract | Crossref Full Text | Google Scholar

11. Romano SM, Pistolesi M. Assessment of cardiac output from systemic arterial pressure in humans. Crit Care Med. (2002) 30(8):1834–41. doi: 10.1097/00003246-200208000-00027

PubMed Abstract | Crossref Full Text | Google Scholar

12. Romagnoli S, Franchi F, Ricci Z, Scolletta S, Payen D. The pressure recording analytical method (PRAM): technical concepts and literature review. J Cardiothorac Vasc Anesth. (2017) 31(4):1460–70. doi: 10.1053/j.jvca.2016.09.004

PubMed Abstract | Crossref Full Text | Google Scholar

13. Scolletta S, Franchi F, Taccone FS, Donadello K, Biagioli B, Vincent JL. An uncalibrated pulse contour method to measure cardiac output during aortic counterpulsation. Anesth Analg. (2011) 113(6):1389–95. doi: 10.1213/ANE.0b013e318230b2de

PubMed Abstract | Crossref Full Text | Google Scholar

14. Greiwe G, Luehsen K, Hapfelmeier A, Rogge D, Kubik M, Schulte-Uentrop L, et al. Cardiac output estimation by pulse wave analysis using the pressure recording analytical method and intermittent pulmonary artery thermodilution: a method comparison study after off-pump coronary artery bypass surgery. Eur J Anaesthesiol. (2020) 37(10):920–5. doi: 10.1097/EJA.0000000000001227

PubMed Abstract | Crossref Full Text | Google Scholar

15. Greiwe G, Flick M, Hapfelmeier A, Winkler MS, Nitzschke R, Frings D, et al. Agreement between cardiac output measurements by pulse wave analysis using the pressure recording analytical method and transthoracic echocardiography in patients with veno-venous extracorporeal membrane oxygenation therapy: an observational method comparison. Eur J Anaesthesiol. (2023) 40(6):436. doi: 10.1097/EJA.0000000000001828

PubMed Abstract | Crossref Full Text | Google Scholar

16. Romagnoli S, Romano SM, Bevilacqua S, Ciappi F, Lazzeri C, Peris A, et al. Cardiac output by arterial pulse contour: reliability under hemodynamic derangements. Interact Cardiovasc Thorac Surg. (2009) 8(6):642–6. doi: 10.1510/icvts.2008.200451

PubMed Abstract | Crossref Full Text | Google Scholar

17. Critchley LAH, Critchley JAJH. A meta-analysis of studies using bias and precision statistics to compare cardiac output measurement techniques. J Clin Monit Comput. (1999) 15(2):85–91. doi: 10.1023/A:1009982611386

PubMed Abstract | Crossref Full Text | Google Scholar

18. Saugel B, Hapfelmeier A, Flick M, Rinehart J, Thiele RH, Vistisen ST, et al. Statistical analysis and reporting of cardiac output method comparison studies (COMPARE) statement. Anesthesiology. (2025) 143(3):518–32. doi: 10.1097/ALN.0000000000005559

PubMed Abstract | Crossref Full Text | Google Scholar

19. Silaschi M, Kruse J, Wittmann M, Osawa I, Goto T, Velten M, et al. Post-operative cerebral oximetry for detection of low-cardiac output syndrome after coronary artery bypass surgery. J Clin Monit Comput. (2025). doi: 10.1007/s10877-025-01359-y

PubMed Abstract | Crossref Full Text | Google Scholar

20. Critchley LA, Lee A, Ho AMH. A critical review of the ability of continuous cardiac output monitors to measure trends in cardiac output. Anesth Analg. (2010) 111(5):1180–92. doi: 10.1213/ANE.0b013e3181f08a5b

PubMed Abstract | Crossref Full Text | Google Scholar

21. Zangrillo A, Maj G, Monaco F, Scandroglio AM, Nuzzi M, Plumari V, et al. Cardiac index validation using the pressure recording analytic method in unstable patients. J Cardiothorac Vasc Anesth. (2010) 24(2):265–9. doi: 10.1053/j.jvca.2009.09.019

PubMed Abstract | Crossref Full Text | Google Scholar

22. Schmid ER, Schmidlin D, Tornic M, Seifert B. Continuous thermodilution cardiac output: clinical validation against a reference technique of known accuracy. Intensive Care Med. (1999) 25(2):166–72. doi: 10.1007/s001340050811

PubMed Abstract | Crossref Full Text | Google Scholar

23. Østergaard M, Nilsson LB, Nilsson JC, Rasmussen JP, Berthelsen PG. Precision of bolus thermodilution cardiac output measurements in patients with atrial fibrillation. Acta Anaesthesiol Scand. (2005) 49(3):366–72. doi: 10.1111/j.1399-6576.2005.00613.x

Crossref Full Text | Google Scholar

24. London GM, Blacher J, Pannier B, Guérin AP, Marchais SJ, Safar ME. Arterial wave reflections and survival in end-stage renal failure. Hypertension. (2001) 38(3):434–8. doi: 10.1161/01.HYP.38.3.434

PubMed Abstract | Crossref Full Text | Google Scholar

25. Laurent S, Katsahian S, Fassot C, Tropeano AI, Gautier I, Laloux B, et al. Aortic stiffness is an independent predictor of fatal stroke in essential hypertension. Stroke. (2003) 34(5):1203–6. doi: 10.1161/01.STR.0000065428.03209.64

PubMed Abstract | Crossref Full Text | Google Scholar

26. Parker KH. A brief history of arterial wave mechanics. Med Biol Eng Comput. (2009) 47(2):111–8. doi: 10.1007/s11517-009-0440-5

PubMed Abstract | Crossref Full Text | Google Scholar

27. Paz Ocaranza M, Riquelme JA, García L, Jalil JE, Chiong M, Santos RAS, et al. Counter-regulatory renin–angiotensin system in cardiovascular disease. Nat Rev Cardiol. (2020) 17(2):116–29. doi: 10.1038/s41569-019-0244-8

PubMed Abstract | Crossref Full Text | Google Scholar

28. Manolis AA, Manolis TA, Manolis AS. Neurohumoral activation in heart failure. Int J Mol Sci. (2023) 24(20):15472. doi: 10.3390/ijms242015472

PubMed Abstract | Crossref Full Text | Google Scholar

29. Herner A, Heilmaier M, Mayr U, Schmid RM, Huber W. Comparison of cardiac function index derived from femoral and jugular indicator injection for transpulmonary thermodilution with the PiCCO-device: a prospective observational study. PLoS One. (2018) 13(7):e0200740. doi: 10.1371/journal.pone.0200740

PubMed Abstract | Crossref Full Text | Google Scholar

30. Rook WH, Turner JD, Clutton-Brock TH. Analysis of damping characteristics of arterial catheter blood pressure monitoring in a large intensive care unit. South Afr J Crit Care. (2017) 33(1):8–10. doi: 10.7196/SAJCC.2017.v33i1.300

Crossref Full Text | Google Scholar

31. Lam S, Liu H, Jian Z, Settels J, Bohringer C. Intraoperative invasive blood pressure monitoring and the potential pitfalls of invasively measured systolic blood pressure. Cureus. (2021) 13(8):e17610. doi: 10.7759/cureus.17610

PubMed Abstract | Crossref Full Text | Google Scholar

32. Foti L, Michard F, Villa G, Ricci Z, Romagnoli S. The impact of arterial pressure waveform underdamping and resonance filters on cardiac output measurements with pulse wave analysis. Br J Anaesth. (2022) 129(1):e6–8. doi: 10.1016/j.bja.2022.03.024

PubMed Abstract | Crossref Full Text | Google Scholar

33. Gopal S, Do T, Pooni JS, Martinelli G. Reply to authors—pitfalls in hemodynamic monitoring techniques: troubles in research or in clinical practice? Minerva Anestesiol. (2015) 81(2):236–7.25338283

PubMed Abstract | Google Scholar

34. Franchi F, Taccone FS, Romagnoli S, Persona P, Donati A. Pitfalls of hemodynamic monitoring techniques: troubles in research or in clinical practice? Minerva Anestesiol. (2015) 81(2):234–5.25220557

PubMed Abstract | Google Scholar

35. Saugel B, Kouz K, Meidert AS, Schulte-Uentrop L, Romagnoli S. How to measure blood pressure using an arterial catheter: a systematic 5-step approach. Critical Care. (2020) 24(1):172. doi: 10.1186/s13054-020-02859-w

PubMed Abstract | Crossref Full Text | Google Scholar

36. Schramm S, Albrecht E, Frascarolo P, Chassot PG, Spahn DR. Validity of an arterial pressure waveform analysis device: does the puncture site play a role in the agreement with intermittent pulmonary artery catheter thermodilution measurements?. J Cardiothorac Vasc Anesth. (2010) 24(2):250–6. doi: 10.1053/j.jvca.2009.05.029

PubMed Abstract | Crossref Full Text | Google Scholar

37. Murata Y, Imai T, Takeda C, Mizota T, Kawamoto S. Agreement between continuous cardiac output measured by the fourth-generation FloTrac/Vigileo system and a pulmonary artery catheter in adult liver transplantation. Sci Rep. (2022) 12(1):11198. doi: 10.1038/s41598-022-14988-z

PubMed Abstract | Crossref Full Text | Google Scholar

38. Bortolotto LA, Hanon O, Franconi G, Boutouyrie P, Legrain S, Girerd X. The aging process modifies the distensibility of elastic but not muscular arteries. Hypertension. (1999) 34(4):889–92. doi: 10.1161/01.HYP.34.4.889

PubMed Abstract | Crossref Full Text | Google Scholar

39. Resnick LM, Militianu D, Cunnings AJ, Pipe JG, Evelhoch JL, Soulen RL, et al. Pulse waveform analysis of arterial compliance: relation to other techniques, age, and metabolic variables. Am J Hypertens. (2000) 13(12):1243–9. doi: 10.1016/S0895-7061(00)01219-X

PubMed Abstract | Crossref Full Text | Google Scholar

40. McEniery CM, Yasmin N, Hall IR, Qasem A, Wilkinson IB, Cockcroft JR. Normal vascular aging: differential effects on wave reflection and aortic pulse wave velocity: the Anglo-Cardiff Collaborative Trial (ACCT). J Am Coll Cardiol. (2005) 46(9):1753–60. doi: 10.1016/j.jacc.2005.07.037

PubMed Abstract | Crossref Full Text | Google Scholar

41. Gopal S, Do T, Pooni JS, Martinelli G. Validation of cardiac output studies from the Mostcare compared to a pulmonary artery catheter in septic patients. Minerva Anestesiol. (2014) 80(3):314–23.24398442

PubMed Abstract | Google Scholar

42. Bakker J. Clinical use of peripheral perfusion parameters in septic shock. Curr Opin Crit Care. (2021) 27(3):269–73. doi: 10.1097/MCC.0000000000000826

PubMed Abstract | Crossref Full Text | Google Scholar

43. Michard F. Towards the automatic detection and correction of abnormal arterial pressure waveforms. J Clin Monit Comput. (2024) 38(4):749–52. doi: 10.1007/s10877-024-01152-3

PubMed Abstract | Crossref Full Text | Google Scholar

44. Bramwell JC, Hill AV. The velocity of pulse wave in man. Proc R Soc Lond Ser B Biol Sci. (1922) 93(652):298–306. doi: 10.1098/rspb.1922.0022

Crossref Full Text | Google Scholar

45. Ding CCA, Dokos S, Bakir AA, Zamberi NJ, Liew YM, Chan BT, et al. Simulating impaired left ventricular–arterial coupling in aging and disease: a systematic review. Biomed Eng Online. (2024) 23(1):24. doi: 10.1186/s12938-024-01206-2

PubMed Abstract | Crossref Full Text | Google Scholar

46. Clermont G, Zenker S. The inverse problem in mathematical biology. Math Biosci. (2015) 260:11–5. doi: 10.1016/j.mbs.2014.09.001

PubMed Abstract | Crossref Full Text | Google Scholar

47. Quarteroni A, Manzoni A, Vergara C. The cardiovascular system: mathematical modelling, numerical algorithms and clinical applications. Acta Numerica. (2017) 26:365–590. doi: 10.1017/S0962492917000046

Crossref Full Text | Google Scholar

48. Charlton PH, Mariscal Harana J, Vennin S, Li Y, Chowienczyk P, Alastruey J. Modeling arterial pulse waves in healthy aging: a database for in silico evaluation of hemodynamics and pulse wave indexes. Am J Physiol Heart Circ Physiol. (2019) 317(5):H1062–85. doi: 10.1152/ajpheart.00218.2019

PubMed Abstract | Crossref Full Text | Google Scholar