Comparison of cardiac index values between bioimpedance and pulse contour analysis in patients undergoing cardiac surgery under cardiopulmonary bypass

Abstract

Background and Aims:

The performance of non-invasive techniques remains inconsistent in cardiac surgical patients. This study assessed the level of agreement and trending abilities of cardiac index (CI) values obtained using Noninvasive Cardiac System (NICaS™) and FloTrac™ (version 4.0) in adult patients undergoing cardiac surgery.

Methods:

This retrospective observational study was conducted over 6 months at a tertiary care centre and included 16 patients undergoing elective cardiac surgery with cardiopulmonary bypass. CI was measured simultaneously using FloTrac™ from the radial (FCIrad) and femoral (FCIfem) arteries and using NICaS™ at corresponding time points (NCIrad and NCIfem). CI measurements were recorded at six time points: baseline (T1), post-induction (T2), post-sternotomy (T3), post-CPB weaning (T4), post-protamine (T5), and post-sternal closure (T6). Agreement between methods was assessed using Bland–Altman analysis, while trending ability was evaluated using four-quadrant and polar plot analyses. A ±30% change in CI relative to FloTrac™ values was considered the acceptable limit of agreement.

Results:

A total of 90 paired measurements were available for FCIrad–NCIrad and 73 for FCIfem–NCIfem. Bland–Altman analysis showed a bias of 0.33 L/min/m² (limits of agreement: -1.81 to 2.48 L/min/m²) for FCIrad–NCIrad and 0.52 L/min/m² (-2.03 to 3.07 L/min/m²) for FCIfem–NCIfem. Four-quadrant plots demonstrated concordance rates of 69% (radial) and 63.4% (femoral). Polar plot analysis showed angular concordance rates of 46% (radial) and 31% (femoral).

Conclusion:

NICaS™ showed limited agreement and trending ability compared with FloTrac™ at radial and femoral sites, indicating limited interchangeability between the two methods during cardiac surgery.

INTRODUCTION

Accurate assessment of cardiac index (CI) is essential for optimal haemodynamic management during cardiac surgery. Intermittent thermodilution using a pulmonary artery catheter remains the gold standard for cardiac output (CO) measurement, but its invasive nature and associated complications have led to an increased interest in non-invasive monitoring methods. These non-invasive monitors perform well in stable patients but show reduced accuracy in patients with unstable haemodynamics or during surgeries associated with major haemodynamic changes.[1]

Noninvasive Cardiac System (NICaS™) (NI Medical; Hod-Hasharon, Israel) is one such non-invasive technology which can measure the continuous CO using whole-body impedance cardiography. Even though this technology has been validated in heart failure patients and non-cardiac surgical patients, it has shown inconsistent results in critically ill patients and surgeries associated with major haemodynamic changes.[2,3,4,5] There is limited literature on NICaS™-derived haemodynamic parameters and their validation with invasive techniques in cardiac surgery.[6]

The latest-generation FloTrac/Vigileo™ system (version 4.0; Edwards Life Sciences, United Staes of America) measures continuous CI using arterial waveform analysis, with calibration updated every 20 seconds to adapt to changes in peripheral vascular resistance. This improvement enhances the system’s responsiveness and accuracy in rapidly changing haemodynamic conditions. FloTrac™ has been widely used for monitoring critically ill patients in operating rooms and the intensive care unit (ICU). Recent evidence shows that the trending ability of CI measurements derived from the FloTrac™ is comparable to that of CI measurements obtained by other invasive techniques. Hence, we selected FloTrac™ as the reference method to compare the NICaS™-measured CI values.[7,8,9]

This retrospective study was considered as a pilot study to assess the level of agreement and trending abilities of CI values obtained using NICaS™ with the fourth-generation FloTrac™ system in adult patients undergoing cardiac surgery under cardiopulmonary bypass (CPB).

METHODS

After obtaining institutional ethics committee approval (EC/NIMS/2685/2021, dated 28.07.2021) prior to the study, we retrospectively analysed prospectively collected data from 18 adult patients (aged 18–60 years) who underwent elective cardiac surgery using CPB between February and March 2019. We could include only 16 patients because the equipment was available for only a brief evaluation period. General consent from the patient was taken before surgery. The study was carried out using the principles of the Declaration of Helsinki, 2013 and good clinical practice guidelines.

Perioperative anaesthetic and surgical management were provided as per the standard institutional protocols. After shifting the patient inside the operation theatre, non-invasive monitors, five-lead electrocardiography (ECG), peripheral oxygen saturation (SpO₂), non-invasive blood pressure (NIBP), and two sensors for the NICaS™, one on a wrist and the other on the contralateral ankle, were applied. After local anaesthetic infiltration, an 18G intravenous cannula and a 20G radial arterial catheter were placed, followed by insertion of a 7 Fr triple-lumen central venous catheter into the right internal jugular vein.

The arterial line was connected to the FloTrac™ sensor. Baseline CI values of NICaS™ and FloTrac™ were obtained. The patient was induced with fentanyl 5 µg/kg, titrated doses of propofol or etomidate (if the ejection fraction was less than 45%), and vecuronium bromide 0.1 mg/kg. Anaesthesia was maintained with sevoflurane or isoflurane, titrated to maintain a bispectral index value between 40 and 60, along with oxygen (40–50%) and air. Muscle relaxation was maintained with vecuronium, and fentanyl was administered at 1–2 µg/kg/h to ensure adequate analgesia. After the induction of anaesthesia, a 16G or 18G femoral arterial catheter and an additional large-bore intravenous cannula were placed. It is part of our institutional protocol to insert an additional femoral arterial line because the radial artery trace is often unreliable immediately after CPB.[10] The femoral waveform provides a more stable signal, which helps guide the use of vasoactive drugs and facilitates smooth weaning from CPB.[11,12] The femoral arterial line was also connected to the same FloTrac™ sensor using a three-way stopcock to measure the CI. Later, the patient was handed over to the surgeons.

CI values were measured simultaneously using FloTrac™ from the radial artery (FCIrad), femoral artery (FCIfem), and NICaS™ (NCIrad, measured at the time of FCIrad and NCIfem, measured at the time of FCIfem) at specified time intervals. Each time, the average of three consecutive measurements was taken as a final reading. NICaS™ and FloTrac™ measurements obtained from both radial and femoral arterial sites were synchronised by recording and averaging three consecutive readings over the same time window at each predefined time point. As obtaining three consecutive readings strictly at end expiration was not feasible in all time-points, measurements were recorded during periods of haemodynamic stability under controlled mechanical ventilation. All patients were also monitored with transoesophageal echocardiography (TEE) during the surgery.

CI measurements recorded at baseline (T1), after induction of anaesthesia (T2), after sternotomy (T3), after weaning from CPB (T4), after protamine administration (T5), and after sternal closure (T6) were included for statistical analyses. In addition to CI, heart rate, arterial blood pressure, central venous pressure (CVP), and nasopharyngeal temperature were recorded at these time points.

Patients with body mass index >30%, arrhythmias, use of intra-aortic balloon pump (IABP), aortic regurgitation, and other aortic pathologies (coarctation of aorta, aneurysm, and dissection), severe peripheral oedema, severe peripheral vascular disease, and congenital heart disease with shunt were excluded. In patients who required intraoperative IABP support, CI values were not recorded thereafter. For patients with an open sternum after surgery, the T6 value was not recorded.

During CPB, all patients were cooled to 34°C and blood cardioplegia was given to all patients as per surgical requirement. A minimum haemoglobin (Hb) of 7g/dL was maintained for all patients during CPB. After surgical correction, haemostasis and adequate rewarming (nasopharyngeal temperature 36 to 36.5°C), heparin was reversed with protamine and the sternum was closed. All patients received nitroglycerin (0.2 µg/kg/min) once rewarming was initiated, and inotropes (epinephrine, dobutamine, or norepinephrine) were initiated as needed. IABP support was initiated in patients with a high Vasoactive–Inotropic Score (>25–30) accompanied by clinical or haemodynamic indicators of myocardial dysfunction, including a distended left ventricle and/or elevated CVP or pulmonary capillary wedge pressure. After surgery, all patients were transferred to the cardiothoracic ICU and placed on mechanical ventilation for further management.

Statistical analysis was performed using Jamovi version 2.6 (Sydney, Australia), MedCalc - version 23.4.5 (MedCalc Software Ltd, Belgium), and SigmaPlot software version 11 (Systat Software, United States of America). No formal sample size calculation was performed as this study was conducted as a pilot investigation during a limited period of device availability. Data are presented as mean [standard deviation (SD)] for continuous variables and as n (%) for categorical variables. Paired data analysis was performed separately for radial and femoral arterial recordings.

Agreement between NICaS™ and FloTrac™-derived CI values was evaluated using Bland–Altman analysis. Percentage error was calculated as (1.96 × SD of bias/mean CI by FloTrac™) × 100, with values ≤30% considered clinically acceptable for interchangeability, as proposed by Critchley and Critchley.[13,14] Bias, limits of agreement (LOA), and percentage error between NCIrad, NCIfem, and FCIrad, FCIfem were calculated for all paired measurements. As a sensitivity analysis to account for repeated measurements within individuals, patient-level mean differences in CI between NICaS™ and FloTrac™ were calculated by averaging all paired measurements across perioperative phases for each patient, and Bland–Altman analysis was repeated using one observation per patient.

Trending ability was analysed using four-quadrant and polar-plot methods, with a concordance rate >90% and an angular bias within ± 5° considered acceptable.[14,15] The polar plot method was used to assess agreement in the direction and magnitude of CI changes, with angular bias indicating systematic directional differences, angular concordance representing the proportion of changes within ± 30°, and radial LOA reflecting consistency in the magnitude of change.[15] Mountain plots were used to visually assess the distribution, symmetry, and spread of differences between the methods. Within-subject variability was quantified using the within-subject standard deviation and coefficient of variation for radial and femoral CI measurements.

RESULTS

Eighteen patients undergoing elective cardiac surgery with CPB were initially enroled. Two patients were excluded because intraoperative TEE revealed significant aortic regurgitation, although their preoperative transthoracic echocardiograms had shown only mild regurgitation. The demographic characteristics of the study population are presented in Table 1. Radial and femoral mean arterial pressures (MAPs) measured at various time points are shown in Table 2. Although radial MAP was slightly higher than femoral MAP at T2, it remained consistently lower at subsequent time points. Radial and femoral arterial CI measurements obtained by both methods, along with the temperature at various time points, are presented in Table 2. CI values from T4 to T6 were not recorded in case 2, who required intraoperative IABP support after CPB, and the T6 value was not recorded in one patient whose sternum was left open postoperatively. CI recordings could not be obtained for case 2 at the T2 time point, case 3 at the T4 time point, and case 12 at the T3 time point due to equipment calibration issues. With all these exclusions, 90 paired radial and 73 paired femoral datasets were available for analysis.

Table 1.

Patient and surgical characteristics

Variable	Value (n=16)
Age (years)	53.56 (12.75) (46.8, 60.4)
Male/Female	10 (62.5%)/6 (37.5%)
Height (cm)	160.06 (10.34) (155, 166)
Weight (kg)	57.94 (16.69) (49, 66.8)
Ejection fraction (%)	57.13 (11.52) (51, 63.3)
Diabetes mellitus	6 (37.5%)
Hypertension	7 (43.7%)
Coronary artery bypass grafting	9 (56.2%)
Valve replacement (aortic or mitral)	4 (25%)
Coronary artery bypass grafting + Valve replacement	3 (18.8%)

Data are presented as mean (standard deviation) (95% confidence interval) or as number (percentage), as appropriate. n=Number of patients

Table 2.

Temperature, mean arterial pressure, and FloTrac™ and NICaS™ cardiac index measurements from radial and femoral artery recordings

	T1	T2	T3	T4	T5	T6
Temp	36.1 (1.15) (35.5, 36.7)	35.5 (0.82) (35.1, 36)	35.2 (0.64) (34.9, 35.6)	35.4 (1.17) (34.8, 36.1)	34.7 (1.42) (33.9, 35.5)	34.6 (1.32) (34.1, 35.6)
RadMAP (mm Hg)	102 (15.5) (93.9, 110)	82.2 (9.45) (77.2, 87.2)	85.5 (14.4) (77.6, 93.5)	71.1 (14.4) (62.8, 79.4)	76.8 (14.1) (69, 84.6)	77.4 (14.3) (69.2, 85.7)
FemMAP (mm Hg)	-	81.4 (11.2) (75.2, 87.6)	88.3 (13.9) (80.6, 96)	77.9 (15.2) (69.1, 86.7)	83.0 (14.0) (75.2, 90.8)	82.0 (14.4) (73.7, 90.3)
FCIrad (L/min/m2)	3.67 (0.95) (3.16, 4.18)	2.75 (0.86) (2.41, 3.20)	3.00 (0.85) (2.67, 3.53)	3.50 (0.94) (2.98, 3.89)	3.53 (1.55) (2.74, 4.32)	3.14 (1.35) (2.50, 4.)
NCIrad (L/min/m2)	3.14 (0.89) (2.67, 3.62)	2.91 (1.20) (2.27, 3.55)	2.94 (1.03) (2.37, 3.51)	2.93 (0.84) (2.44, 3.42)	3.05 (1.20) (2.39, 3.72)	2.80 (1.13) (2.15, 3.45)
FCIfem (L/min/m2)	-	2.91 (0.78) (2.48, 3.35)	2.92 (0.77) (2.49, 3.55)	3.89 (0.89) (3.37, 4.40)	3.99 (1.17) (3.34, 4.63)	3.74 (0.97) (3.18, 4.30)
NCIfem (L/min/m2)	-	3.02 (1.18) (2.37, 3.67)	2.98 (0.99) (2.43, 3.53)	3.03 (0.68) (2.63, 3.43)	2.96 (0.97) (2.42, 3.50)	2.80 (1.09) (2.18, 3.42)

Data are presented as mean (standard deviation) (95% confidence interval), FCIfem=FloTrac™-derived cardiac index from the femoral artery; FCIrad=FloTrac™-derived cardiac index from the radial artery; NCIfem=NICaS™-derived cardiac index measured at the time of FCIfem, NCIrad=NICaS™-derived cardiac index measured at the time of FCIrad; MAP=mean arterial pressure; T1=baseline; T2=after induction of anaesthesia; T3=after sternotomy; T4=after weaning from cardiopulmonary bypass; T5=after protamine administration; and T6=after sternal closure; Temp=temperature

Bland–Altman analysis between NICaS™ and FloTrac™-derived CI values across all time points (T1–T6) revealed variable agreement depending on the perioperative phase. Across all six perioperative time points (T1–T6), FloTrac™ consistently reported higher CI values than NICaS™ at both radial and femoral sites, with the disparity increasing after CPB [Table 3 and Figure 1]. The bias between radial Flotrac™ and NICaS™ CI measurements is less than that of the femoral measurements. Sensitivity analysis to account for repeated measurements within individuals followed the similar pattern [Table 3].

Table 3.

Bland–Altman Analysis of FloTrac™ and NICaS™ Cardiac Index Across Pooled, Pre-Bypass, and Post-Bypass Periods

Variable	Radial (n=90) FloTrac™ Vs NICaS™	Femoral (n=73) FloTrac™ Vs NICaS™
Bias (L/min/m2)	0.33	0.52
SD	1.1	1.30
Limits of agreement (95% CI)	-1.81 to 2.48	-2.03 to 3.07
Percentage error	66.4	74.7
Pre-bypass data
	Radial (n=47) FloTrac™ Vs NICaS™	Femoral (n=30) FloTrac™ Vs NICaS™
Bias (L/min/m2)	0.19	-0.08
SD	1.08	1.23
Limits of agreement (95% CI)	-1.91 to 2.30	-2.32 to 2.49
Percentage error	66.35	82.56
Post-bypass data
	Radial (n=43) FloTrac™ Vs NICaS™	Femoral (n=43) FloTrac™ Vs NICaS™
Bias (L/min/m2)	0.48	0.94
SD	1.11	1.19
Limits of agreement (95% CI)	-1.69 to 2.67	-1.39 to 3.27
Percentage error	63.61	60.26
Sensitivity analysis using patient-level averaged data
	Radial (n=16) FloTrac™ Vs NICaS™	Femoral (n=16) FloTrac™ Vs NICaS™
Bias L/min/m2	0.31	0.43
SD	0.72	1.05
Limits of agreement (95% CI)	-1.10 to 1.72	-1.64 to 2.49
Percentage error	45.2	64.1

CI=Confidence interval; SD=the standard deviation of the differences between the two cardiac index measurements; n indicates the number of matched paired cardiac index measurements for pooled, pre-bypass, and post-bypass analyses. In the sensitivity analysis, n denotes the number of patients, with one averaged observation per patient

Figure 1.

Bland–Altman analysis of cardiac index measurements comparing FloTrac™ and NICaS™. a; Bland–Altman plot for cardiac index values obtained using radial FloTrac™ and NICaS™ (before and after cardiopulmonary bypass), b; Bland–Altman plot for cardiac index values obtained using femoral FloTrac™ and NICaS™ (before and after cardiopulmonary bypass), c; Bland–Altman plot for cardiac index values obtained using radial FloTrac™ and NICaS™ before cardiopulmonary bypass, d; Bland–Altman plot for cardiac index values obtained using femoral FloTrac™ and NICaS™ before cardiopulmonary bypass, e; Bland–Altman plot for cardiac index values obtained using radial FloTrac™ and NICaS™ after cardiopulmonary bypass, f; Bland–Altman plot for cardiac index values obtained using femoral FloTrac™ and NICaS™ after cardiopulmonary bypass. The solid line represents mean bias, and the dashed lines represent the limits of agreement (mean bias ± 1.96 standard deviation)

For both four-quadrant plot analysis and polar plot analysis, central zone data [mean change in cardiac output (ΔCO) <15% and < 10%, respectively] representing any statistical noise component were excluded.[14] Four-quartile plot analysis of data points outside the 15% exclusion zone showed a concordance rate of 69% for radial measurements and 63.4% for femoral measurements [Figure 2]. It demonstrated limited agreement between NICaS™ and FloTrac™ in tracking directional changes in CI. Although the radial site showed slightly better concordance than the femoral site, both demonstrated limited trending ability, as indicated by concordance rates below the acceptable threshold of 90%. Polar plot analysed outside the 10% exclusion zone showed an angular bias of 7.2° (radial), 8.5° (femoral), and angular concordance rates of 46% for radial and 31% for femoral [Figure 2]. The wide angular biases exceeding ± 5°and the low angular concordance rates at 30° suggest a limited ability of NICaS™ to track changes in CI relative to FloTrac™.

Figure 2.

Four quadrant and polar plot analysis of trending ability for pooled cardiac index data. Four-quadrant plots depict directional changes in cardiac index (ΔCI) measured by NICaS™ and FloTrac™ for pooled radial and femoral data, with a central ±15% exclusion zone applied to reduce measurement noise. Polar plots assess agreement in both direction and magnitude of ΔCI, with a ±10% exclusion zone around the origin. Angular bias represents the mean polar angle, and angular concordance was defined as the proportion of data points within ±30°

Mountain plots demonstrated that the paired differences at both radial and femoral sites were symmetrically distributed, with peaks shifted slightly left of zero, confirming a modest systematic underestimation by NICaS™ [Figure 3]. However, the broad bases of the folded distributions showed substantial dispersion. Within-subject variability analysis showed mean within-subject standard deviations of 0.75 and 0.77 L/min/m² for radial NICaS™ and FloTrac™, respectively, and 0.72 and 0.69 L/min/m² for femoral NICaS™ and FloTrac™. The corresponding coefficients of variation were 25.4%, 23.3%, 24.1%, and 19.4%, respectively.

Figure 3.

Mountain plots showing the difference between FloTrac™ and NICaS™ cardiac index measurements from radial and femoral artery recordings

DISCUSSION

Our results indicate limited agreement between NICaS™-derived and Flotrac™-derived CI values. FloTrac™ consistently showed higher CI values compared with NICaS™ at almost all time intervals. The difference was more during the post-CPB period (T4–T6). The wide LOA (> ±1.0–1.25 L/min/m²) and high percentage error (>30%) observed for both radial and femoral measurements indicate limited interchangeability of absolute CI values between the two methods in this setting.[3]

Based on our findings, NICaS™ showed limited agreement with FloTrac™ measurements, reflected by a consistent mean bias with underestimation of CI values. Precision was also limited, as indicated by wide LOA, suggesting that NICaS™ did not reproducibly track FloTrac™ measurements. However, these results should be interpreted with caution for the following reasons. FloTrac™ is not a gold standard method. The ≤30% percentage error threshold proposed by Critchley and Critchley assumes comparison with a true reference method and should therefore be interpreted cautiously when both techniques have intrinsic measurement limitations.[13] In the present study, the high percentage errors observed may have been influenced by low mean CI values and marked physiological variability after CPB. Accordingly, percentage error was interpreted alongside absolute bias and LOA in relation to clinically meaningful changes in CI. In addition, when repeated measurements were averaged at the patient level, and the Bland–Altman analysis was repeated using one observation per subject, the direction and magnitude of bias remained consistent with the primary analysis, although the LOA remained wide, likely reflecting the imprecision associated with a small pilot sample.

Although the FloTrac™ system version 4.0 algorithm recalibrates every 20 seconds to account for changes in vascular tone, its accuracy can still be affected by alterations in vascular compliance, which are frequently encountered during cardiac surgery.[16] NICaS™ estimates stroke volume using whole-body bioimpedance and is sensitive to changes in fluid distribution and tissue resistivity. During CPB, haemodilution, hypothermia, fluid shifts, and peripheral oedema alter bioimpedance, leading to underestimation of CI.[17,18] These acute physiological alterations explain the progressively widening bias between NICaS™ and FloTrac™ from T4 through T6. Low tidal volume ventilation, cardiac arrhythmias, low CO states, and interference from surgical diathermy may also influence NICaS™ measurements.

Previous validation studies on non-invasive technologies such as bioimpedance have shown good correlation with thermodilution in stable or heart failure populations but reduced accuracy and agreement during haemodynamic instability and in critically ill patients.[18,19,20,21] These findings support our observation that NICaS™ performs better under stable pre-CPB conditions but demonstrates reduced agreement and trending performance in the dynamic post-CPB period.

The smaller bias observed between radial FloTrac™ and NICaS™, compared with femoral FloTrac™ and NICaS™, can be attributed to physiological and algorithmic factors inherent to arterial waveform analysis. The FloTrac™ algorithm incorporates demographic adjustments and derives stroke volume from peripheral arterial waveform characteristics, integrating information related to arterial diameter, elastance, and vascular resistance.[22] Overall, the interplay of vascular tone, central waveform amplification, and bioimpedance sensitivity to fluid and temperature shifts introduces significant limitations to both technologies during CPB. Clinicians must recognise these physiological and technical fallacies to avoid misinterpretation of CI trends and ensure appropriate haemodynamic decision-making in cardiac surgery.

Another important focus of this pilot study was to evaluate the trending ability of NICaS™. Four-quadrant and polar plot analyses showed concordance rates below predefined acceptance thresholds, indicating a limited ability of NICaS™ to reliably track the direction and magnitude of CI changes relative to FloTrac™ during dynamic perioperative conditions. In clinical practice, trending ability is often more important than single absolute measurements during rapid haemodynamic changes as poor trending limits the ability to distinguish true physiological changes from measurement variability.

This study has limitations, including a small sample size and the absence of a formal sample size calculation. Repeated measurements within patients may have contributed to the unstable Bland–Altman LOA, inflated percentage errors, and overestimation of discordance in trending analyses. Accordingly, the findings should be interpreted as exploratory and hypothesis-generating. Furthermore, while some evidence suggests minimal impact, thoracic opening and skin disruption may specifically confound bioimpedance measurements and hence, the results should be interpreted with caution.[23,24,25] Finally, the inclusion of postoperative measurements could have provided additional valuable insights into device performance during recovery.

CONCLUSION

This pilot study demonstrates that NICaS™-derived CI measurements show limited agreement and trending ability when compared with FloTrac™ in patients undergoing cardiac surgery. As both techniques have inherent methodological limitations and neither represents a gold standard, the observed systematic bias, wide LOA, high percentage errors, and low concordance across perioperative phases suggest limited interchangeability between the two methods. Larger multicentre studies using thermodilution are needed for further validation.

Study data availability

De-identified data may be requested with reasonable justification from the authors (email to the corresponding author) and shall be shared after approval as per the authors’ Institution policy.

Disclosure of use of artificial intelligence (AI)-assistive or generative tools

No artificial intelligence tools or large language models were utilised in the preparation of this manuscript, except for the use of software for grammar correction.

Author contributions

AP Design, literature search, conduct of cases, data collection and analysis, review and approval. SSA Concepts, design, definition of intellectual content, literature search, conduct of cases, data acquisition, data analyses, manuscript preparation, editing, review and approval. DRR Concepts, definition of intellectual content, literature search, conduct of cases, data acquisition, data analyses, manuscript preparation, editing, review and approval. GR Concepts, design, definition of intellectual content, literature search, conduct of cases, data analyses, manuscript preparation, editing, review and approval. All the authors have participated in the review, drafting and final approval of the manuscript.

Conflicts of interest

Dr. Gopinath Ramachandran who is one of the authors of this article is a member of the editorial team of Indian Journal of Anaesthesia. However, he was not involved at any time in the decision-making process for the publication of this study.

Also, the NICaS™ device and sensors used in this study were provided by Sandor Medicaids Pvt. Ltd., the Indian distributor of the NICaS™ haemodynamic monitoring system. The FloTrac™ monitor and transducer were institutional equipment and were not sponsored. There are no other conflicts of interest.

Funding Statement

No financial grant or funding was received for this study.