Accuracy of Swan‒Ganz catheterization-based assessment of right ventricular function: Validation study using high-fidelity micromanometry-derived values as reference.

Right ventricular (RV) function critically affects the outcomes of patients with pulmonary hypertension (PH). Pressure wave analysis using Swan‒Ganz catheterization (SG-cath) allows for the calculation of indices of RV function. However, the accuracy of these indices has not been validated. In the present study, we calculated indices of systolic and diastolic RV functions using SG-cath-derived pressure recordings in patients with suspected or confirmed PH. We analyzed and validated the accuracies of three RV indices having proven prognostic values, that is, end-systolic elastance (Ees)/arterial elastance (Ea), β (stiffness constant), and end-diastolic elastance (Eed), using high-fidelity micromanometry-derived data as reference. We analyzed 73 participants who underwent SG-cath for the diagnosis or evaluation of PH. In this study, Ees/Ea was calculated via the single-beat pressure method using [1.65 × (mean pulmonary arterial pressure) - 7.79] as end-systolic pressure. SG-cath-derived Ees/Ea, β, and Eed were 0.89 ± 0.69 (mean ± standard deviation), 0.027 ± 0.002, and 0.16 ± 0.02 mmHg/ml, respectively. The mean differences (limits of agreement) between SG-cath and micromanometry-derived data were 0.13 (0.99, -0.72), 0.002 (0.020, -0.013), and 0.04 (0.20, -0.12) for Ees/Ea, β, and Eed, respectively. The intraclass correlation coefficients of the indices derived from the two catheterizations were 0.76, 0.71, and 0.57 for Ees/Ea, β, and Eed, respectively. In patients with confirmed or suspected PH, SG-cath-derived RV indices, especially Ees/Ea and β, exhibited a good correlation with micromanometry-derived reference values.

INTRODUCTION

The right ventricular (RV) function is critically important in the management of pulmonary hypertension (PH). , A variety of indices are used to assess RV function, such as cardiac output, tricuspid annular plane systolic excursion, and RV ejection fraction (EF). However, these indices are subject to change under different pre‐ and after‐loads and, thus do not reflect the “intrinsic” RV function.

The intrinsic RV function comprises contractility and diastolic function (relaxation and compliance). , Gold standard indices for these functions are end‐systolic elastance (Ees) for contractility, tau for relaxation, β and end‐diastolic elastance (Eed) for compliance/elastance. Another emerging element of RV function is RV‐pulmonary arterial (PA) coupling, represented as Ees/[arterial elastance (Ea)]. Previous studies, including ours, have reported the promising prognostic roles of these indices. , , ,

In most previous studies, these RV indices were calculated using Swan‒Ganz catheterization (SG‐cath)‐derived pressure and its derivatives, including isovolumic pressure (Piso). However, a Swan‒Ganz catheter is a compliant water‐filled catheter that has innate inaccuracies, such as bending and distortion, pressure blurring due to air bubbles, and attenuation of high frequencies. , Consequently, some RV indices, calculated using SG‐cath, may not be free from catheter‐related errors. In fact, it is commonly observed that pressure waves recorded via SG‐cath show fluctuations or noises that may impede the reliability of the obtained pressure and its derivatives.

In Japan, a thin modified pressure catheter used for high‐fidelity micromanometry (Pressure‐cath) became available in 2017. This enables accurate RV pressure (RVP) measurements with less invasiveness and a shorter examination time than its previous model. This modified pressure catheter enables successive measurements of RVP using both SG‐ and Pressure‐cath with a clinically acceptable examination time and invasiveness. Notably, the gold standard method for calculating indices of RV function requires a simultaneous application of pressure and conductance catheterizations, which enable drawings of multiple pressure‐volume loops. However, Pressure‐cath is a reliable modality that enables more accurate RVP recording and calculations of RV indices, at least, compared to SG‐cath.

In this study, we aimed to validate the accuracy of SG‐cath‐derived indices of RV function. Specifically, we aimed to determine whether RV indices having prognostic values, such as Ees/Ea and β, were significantly correlated with the corresponding Pressure‐cath‐derived values, and whether the correlation coefficients were acceptably high.

METHODS

Study participants

We collected data from a series of patients with suspected or confirmed PH who underwent right heart catheterization at our hospital between May 2020 and December 2021. In patients with confirmed PH, the diagnosis, classification, and management of PH were performed according to the guidelines for PH published in 2015. Patients also underwent cardiac magnetic resonance imaging (CMRI) when possible.

This study was approved by the institutional review board of our institution. Owing to the retrospective nature of the study, informed consent was obtained on an opt‐out basis, via our institution website. Patient identity was concealed, and all data were compiled according to the requirements of the Japanese Ministry of Health, Labor, and Welfare. Moreover, this study was conducted in accordance with the 1964 Declaration of Helsinki and its later amendments.

Right heart catheterization

Assessment of pulmonary hemodynamics via SG‐cath

Participants underwent SG‐cath according to the methods recommended in the guidelines for PH. Briefly, with the patient in the supine position, a 6‐French sheath was placed through the right internal jugular or femoral vein, and a Swan‒Ganz catheter was inserted and advanced into the pulmonary artery. The catheter was connected to a polygraph (RMC‐4000; Nihon Kohden Corporation). PA pressure, PA wedge pressure, CO, RVP, and right atrial pressure were recorded. The RV end‐diastolic pressure (RVEDP) was measured immediately before a rapid upstroke in RVP. The zero level was determined at the mid‐level of the chest wall. All pressure measurements were performed at the end of normal expiration (without breath holding), while additional diastolic RVP measurements of at least five cardiac cycles were performed with breath holding at the end of normal expiration. These recommendations related to breathing were in accordance with the guidelines of PH management. Cardiac output was measured using the thermodilution method, and the mean of at least three measurements was used as the representative value.

Measurement of RVP using a high‐fidelity micromanometer

Immediately after SG‐cath, the Swan‒Ganz catheter was removed and a 5‐F or 6‐F guiding catheter (Mach 1™ Coronary Guide Catheter, Boston Scientific) was inserted and advanced to the right ventricle. Subsequently, a pressure catheter (Micro‐Cath™ pressure catheter, Millar Co., Ltd.) was inserted through the guiding catheter into the right ventricle. The pressure catheter was connected to an AV converter (PowerLab®, ADInstruments), and then to a personal computer wherein a dedicated software (LabChart Pro®, ADInstruments) was installed to allow for real‐time pressure recording at 1000 Hz. The zero level was determined before the pressure catheter was inserted into the guiding catheter, by placing the tip of the catheter immediately below the surface of warm water in a cup. RVP was recorded at natural end‐expiration. At least five consecutive RVP curves were recorded and analyzed. These processes of Pressure‐cath were completed within approximately 15 min, without any significant changes in circulatory and respiratory conditions.

Assessment of RV morphology using CMRI

CMRI was performed using a 1.5‐T Achieva scanner (Philips Medical Systems) with a five‐channel coil and master gradients (maximum gradient amplitude, 33 mT/m; maximum slew rate, 100 mT/m/ms). Furthermore, CMRI was performed within 14 days from the date of right heart catheterization, during which there were no significant changes in either hemodynamic status or PH treatment. Image acquisition and analysis were performed using a previously described protocol, with high intra‐ and interobserver reproducibilities. Briefly, 12 axial slices were acquired using a steady‐state free‐precession pulse sequence (repetition time, 2.8 ms; echo time, 1.4 ms; flip angle, 60; acquisition matrix, 192 × 256; field of view, 380 ms; slice thickness, 10 mm; 0 mm inter‐slice gap; and 20 phases/cardiac cycle). Images were analyzed using commercially available analysis software (Extended MR Work Space ver. 2.6.3; Philips Medical Systems). In the axial datasets, the endocardial contours of the right ventricle were manually traced, and the RV and left ventricular end‐diastolic volume (EDV) and end‐systolic volume (ESV) were computed. RV and left ventricular stroke volume (SV) and EF were calculated as SV = EDV − ESV and EF = SV/EDV × 100%, respectively.

Assessment of RV function

Analysis of SG‐cath‐ and Pressure‐cath‐derived RVP indices

RVP data recorded via SG‐ and Pressure‐cath were analyzed offline using LabChart Pro®.

Five stable, consecutive RVP curves were selected, and the mean of five maximum RVPs (systolic RVP [SRVP]), minimum RVP (RVP‐min), RVEDP, maximum dp/dt (dp/dt‐max), minimum dp/dt (dp/dt‐min), tau, and Piso were calculated. RVEDP was determined using a dedicated program for RVP analysis bundled in LabChart Pro®. Piso calculation was performed using LabChart Pro® based on the single‐beat method reported by Bellofiore et al. In this method, sine‐wave curves were fitted (curve fitting) to the early systolic and early diastolic portions of the RVP curve, and the peak pressure value of the sine‐wave for one cardiac cycle was read as Piso. The mean of five consecutive Piso values was used as the representative value. Tau was calculated using the Weiss and Logistic methods. ,

Calculations of Ees, Ees/Ea, β, and Eed

Ees was calculated using the following formula (single‐beat method) , : Here, Pes represents end‐systolic pressure. The gold standard method to determine Pes is the multiple pressure‐volume (PV)‐loop analysis (multiple beat method) in which conductance and pressure catheterization are simultaneously applied. However, in the present study, we did not conduct multiple PV‐loop analysis; contrariwise, we used three surrogate parameters for Pes, that is, mean PA pressure (MPAP), calculated pressure using MPAP (MPAPcalc, calculated as 1.65 × MPAP – 7.79), and SRVP, as reported in previous studies. , , , , , , , When MPAP was used for Pes, Ees was represented as Ees (MPAP); when MPAPcalc was used for Pes; Ees was represented as Ees (MPAPcalc); and when SRVP was used for Pes, Ees was represented as Ees (SRVP). SG‐cath‐derived MPAP was used in the calculations of Ees (MPAP) and Ees (MPAPcalc). Alternatively, SG‐cath‐ and Pressure‐cath‐derived SRVP were used for the calculations of SG‐cath‐ and Pressure‐cath‐derived Ees/Ea (SRVP), respectively. RVSV represents the RV SV calculated from CMRI.

Ees/Ea is the ratio of Ees to Ea, where Ea is estimated as the ratio of Pes to CMRI‐derived RVSV. Ees/Ea was then calculated from Equations (1) and (2).

This method of Ees/Ea calculation is called the “pressure method” because only pressure data are used for calculating Ees/Ea. Here, Ees/Ea ratios calculated using MPAP, MPAPcalc, and SRVP as Pes were represented as Ees/Ea (MPAP), Ees/Ea (MPAPcalc), and Ees/Ea (SRVP), respectively.

Another method for calculating Ees/Ea is the “volume method,” in which only volume data are used. In this method, Ees is represented as Pes/RVESV. Thus, Ees/Ea can be calculated by [Pes/RVESV]/[Pes/RVSV], which leads to the following equation. β was calculated as a solution of the following two simultaneous equations, as reported by Rain et al. Here, RVV represents RV volume, and the following three points of pressure and volume were used to calculate α and β: (RVP, RVV); (0, 0), (RVP‐min, RVESV), and (RVEDP, RVEDV). RVP‐min was normalized at 1 mmHg and, only in this calculation, RVEDP was modified by a formula of [1 + (RVEDP – RVP‐min)], to avoid measurement errors.

Eed was calculated as the slope of the diastolic pressure‐volume relationship at end‐diastole, using the formula previously described by Trip et al.

All RV indices were calculated using both SG‐ and Pressure‐cath.

Data analysis

Categorical variables are expressed as absolute numbers (percentages) and continuous variables as means ± standard deviations (SDs) or medians (interquartile ranges), as appropriate. Using Bland‐Altman analysis and intraclass correlation coefficients (ICCs), we evaluated SG‐cath‐derived RVP and RV indices (dp/dt‐max, dp/dt‐min, tau [Weiss], tau [Logistic], Ees [MPAP], Ees [MPAPcalc], Ees [SRVP], Ees/Ea [MPAP], Ees/Ea [MPAPcalc], Ees/Ea [SRVP], β, and Eed) using the corresponding Pressure‐cath‐derived variables as reference. The 95% limits of agreement were calculated by mean ± 1.96 × SD. A high ICC ranging between 0 and 1 indicated high similarity between SG‐ and Pressure‐cath‐derived data. The sample size was not calculated and all consecutive participants who were eligible for the present study during the study period were included. JMP Pro version 14 (SAS Institute Inc.) was used for statistical analyses.

RESULTS

Table 1 shows the demographic characteristics of the 73 participants, 57 of whom were diagnosed with PH. The other 16 individuals underwent SG‐cath with a suspicion of PH; however, they did not fulfill the criteria for PH, as their MPAP and pulmonary vascular resistance were ≤20 mmHg and ≤3 Wood units, respectively, and so were analyzed as controls.

Table 1

Characteristics of the study population

	Total subjects (n = 73)	Pulmonary hypertension (n = 57)	Controls (n = 16)
Age (years)	58 ± 16	56 ± 17	62 ± 10
Sex (female/male)	49/24	35/22	14/2
Body mass index (kg/m2)	23 ± 5	23 ± 5	22 ± 6
Pulmonary hypertension group (1/2/3/4/5)		33/0/11/12/1	NA
WHO functional class (I/II/III/IV)		12/15/28/2	NA
Use of pulmonary vasodilator(s)			NA
None		17
Single drug		10
Double combination		18
Triple combination		12

Abbreviations: NA, not applicable; WHO, World Health Organization.

Table 2 shows the results of CMRI and SG‐cath of the 73 participants. CMRI was performed in 50 of the 73 patients.

Table 2

Results of Swan‒Ganz catheterization and cardiac magnetic resonance imaging

Swan‒Ganz catheterization
	Total subjects	Pulmonary hypertension	Controls
n	73	57	16
PAWP (mmHg)	7.6 ± 2.7	8.0 ± 2.7	6.3 ± 2.4
MPAP (mmHg)	28.9 ± 11.2	32.5 ± 10.0*	16.1 ± 3.9
RVEDP (mmHg)	7.1 ± 3.0	7.9 ± 2.7*	4.4 ± 2.6
RAP (mmHg)	3.8 ± 2.4	4.2 ± 2.4	2.3 ± 2.0
CO (L/min)	4.5 ± 1.0	4.5 ± 1.1	4.5 ± 1.0
CI (L/min/m2)	2.8 ± 0.6	2.8 ± 0.6	2.9 ± 0.5
PVR (Wood unit)	5.2 ± 3.5	6.0 ± 3.6*	2.2 ± 0.9

Note: Data are presented as mean ± standard deviation (SD).

Abbreviations: CI, cardiac index; CO, cardiac output; Ees/Ea (Vol), Ees/Ea calculated with the volume method ([RV stroke volume]/[RV end‐systolic volume]); MPAP, mean pulmonary arterial pressure; PAWP, pulmonary arterial wedge pressure; PVR, pulmonary vascular resistance; RAP, right atrial pressure; RV, right ventricular; RVEDP, RV end‐diastolic pressure.

p < 0.05 versus Controls by Wilcoxon's test.

Figure 1 shows representative images of RVP waves recorded via SG‐ and Pressure‐cath in a 43‐year‐old woman with heritable PA hypertension. Notably, SG‐cath‐derived RVPs showed fluctuations or notches near end‐diastole (bottom arrow) and after the steep increase in the systolic phase (top arrow; Figure 1a), whereas Pressure‐cath‐derived RVPs showed no or slight fluctuations/notches (Figure 1b). In addition, SG‐cath‐derived RVP and calculated indices of RV function, including Ees/Ea (MPAP), Ees/Ea (MPAPcalc), dp/dt‐max, β, and Eed, were higher than those obtained via Pressure‐cath.

Figure 1

Representative image of right ventricular pressure recorded via Swan–Ganz catheterization and high‐fidelity micromanometry. SG‐cath‐derived RVP showed fluctuations or notches near end‐diastole and after the steep increase in the systolic phase (arrows) (a), whereas Pressure‐cath‐derived RVPs showed no or slight fluctuations/notches (b). In addition, SG‐cath‐derived RVP and calculated indices of RV function, including Ees (MPAP), Ees (MPAPcalc), Ees/Ea (MPAP), Ees/Ea (MPAPcalc), β, and Eed, were higher than those obtained by Pressure‐cath. dp/dt‐max, maximum dp/dt; dp/dt‐min, minimum dp/dt; Ea, arterial elastance; Eed, end‐diastolic elastance; Ees, end‐systolic elastance; MPAP, mean pulmonary arterial pressure; MPAPcalc, calculated end‐systolic pressure using MPAP; Piso, isovolumic pressure; Pressure‐cath, high‐fidelity micromanometry; RVEDP, right ventricular end‐diastolic pressure; RVP, right ventricular pressure; SG‐cath, Swan–Ganz catheterization; SRVP, systolic right ventricular pressure

Table 3 shows comparisons of pulmonary hemodynamic parameters between SG‐ and Pressure‐cath. Bland‐Altman analysis showed no systematic trends in the values of the parameters. All limits of agreement included 0, and there were no significant differences between SG‐cath‐ and Pressure‐cath‐derived values (Figure 2). ICCs between SG‐ and Pressure‐cath‐derived data are shown in Table 3. The time difference between SG‐ and Pressure‐cath RVP recordings was 12 (11, 16) min.

Table 3

Comparison of right ventricular pressure and indices of right ventricular function evaluated with Swan‒Ganz catheterization and high‐fidelity micromanometry

	Swan‒Ganz catheterization	High‐fidelity micromanometry	Bland–Altman plot, mean difference (limits of agreement)	Intraclass correlation coefficient
Total subjects (n = 73)
HR (/min)	69.8 ± 1.4	71.1 ± 1.4	−1.3 (7.7, −10.3)	0.92
SRVP (mmHg)	47.5 ± 2.1	43.0 ± 2.0	4.5 (11.3, −2.2)	0.95
RVP‐min (mmHg)	2.3 ± 0.3	0.8 ± 0.3	1.5 (5.8, −2.8)	0.62
RVEDP (mmHg)	6.6 ± 0.4	4.4 ± 0.3	2.2 (7.2, −2.9)	0.49
Piso (mmHg)a	67.9 ± 2.9	62.0 ± 2.1	6.0 (32.9, −21.0)	0.77
Ees/Ea (MPAP)b	1.46 ± 0.08	1.28 ± 0.06	0.18 (1.20, −0.84)	0.60
Ees/Ea (MPAPcalc)c	0.89 ± 0.69	0.76 ± 0.61	0.13 (0.99, −0.72)	0.76
Ees/Ea (SRVP)d	0.48 ± 0.04	0.53 ± 0.04	−0.05 (0.53, −0.62)	0.65
dp/dt‐max (mmHg/s)	502.6 ± 20.6	427.1 ± 13.7	75.5 (325.4, −174.4)	0.57
dp/dt‐min (mmHg/s)	−467.5 ± 18.4	−445.4 ± 18.0	−22.1 (119.2, −163.3)	0.89
Tau (s) (Weiss)	43.5 ± 2.4	44.0 ± 2.0	−0.5 (25.0, −26.1)	0.76
Tau (s) (Logistic)	27.8 ± 1.5	35.5 ± 1.7	−7.7 (24.8, −40.2)	0.22
Subjects with CMRI study (n = 50)
Ees (MPAP) (mmHg/ml)e	0.78 ± 0.06	0.63 ± 0.04	0.14 (0.74, −0.46)	0.63
Ees (MPAPcalc) (mmHg/ml)f	0.55 ± 0.37	0.41 ± 0.30	0.14 (0.74, −0.46)	0.54
Ees (SRVP) (mmHg/ml)g	0.34 ± 0.04	0.39 ± 0.05	0.06 (0.59, −0.48)	0.65
β h	0.027 ± 0.002	0.024 ± 0.001	0.002 (0.02, −0.013)	0.71
Eed (mmHg/ml)i	0.16 ± 0.02	0.12 ± 0.01	0.04 (0.2, −0.12)	0.57

Note: Data are shown as mean ± standard error (SE).

Abbreviations: CMRI, cardiac magnetic resonance imaging; dp/dt‐max, maximum dp/dt; dp/dt‐min, minimum dp/dt; Ea, arterial elastance; Eed, end‐diastolic elastance; Ees, end‐systolic elastance; HR, heart rate; MPAP, mean pulmonary arterial pressure; MPAPcalc, calculated end‐systolic pressure using MPAP; Piso, isovolumic pressure; RVEDP, right ventricular end‐diastolic pressure; RVEDV, right ventricular end‐diastolic volume; RVP‐min, minimum right ventricular pressure; RVSV, right ventricular stroke volume; RVV, right ventricular volume; SRVP, systolic right ventricular pressure.

Calculated using the single‐beat method.

Calculated with the following formula: (Piso/MPAP) − 1.

Calculated with the following formula: (Piso/[1.65 × MPAP − 7.79]) − 1.

Calculated with the following formula: (Piso/SRVP) − 1.

Calculated with the following formula: (Piso − MPAP)/(CMRI‐derived RVSV).

Calculated with the following formula: (Piso − MPAPcalc)/(CMRI‐derived RVSV).

Calculated with the following formula: (Piso − SRVP)/(CMRI‐derived RVSV).

Calculated by solving the equations: RVP = α(eRVV·β − 1) (n = 48).

Calculated using the following formula: Eed = α·β·eRVEDV·β (n = 48).

Figure 2

Bland‒Altman plots of 13 SG‐ and Pressure‐cath‐derived indices of RVP. There were no systematic trends based on the values of the 13 indices. In all indices, the limits of agreement included 0, indicating that there were no significant differences between SG‐ and Pressure‐cath‐derived indices. dp/dt‐max, maximum dp/dt; dp/dt‐min, minimum dp/dt; Ea, arterial elastance; Eed, end‐diastolic elastance; Ees, end‐systolic elastance; HR, heart rate; MPAP, mean pulmonary arterial pressure; MPAPcalc, calculated end‐systolic pressure using MPAP; Piso, isovolumic pressure; Pressure‐cath, high‐fidelity micromanometry; RV, right ventricular; RVEDP, RV end‐diastolic pressure; RVP‐min, minimum RV pressure; SG‐cath, Swan–Ganz catheterization; SRVP, systolic RV pressure

Table 4 shows the SG‐ and Pressure‐cath‐derived data in controls and in patients with PH.

Table 4

Right ventricular pressure and indices of right ventricular function in controls and in patients with pulmonary hypertension

Data from the entire subjects (n  = 73)
	Swan–Ganz catheterization‐derived data		High‐fidelity micromanometry‐derived data
	Controls	Patients with pulmonary hypertension	Controls	Patients with pulmonary hypertension
n	16	57	16	57
SRVP (mmHg)	29.1 ± 5.0	52.7 ± 16.6*	25.7 ± 5.0	47.8 ± 16.1**
RVP‐min (mmHg)	0.8 ± 2.6	2.7 ± 2.7*	0.0 ± 2.2	1.0 ± 3.0
RVEDP (mmHg)	4.5 ± 3.0	7.2 ± 3.0*	3.3 ± 2.0	4.8 ± 2.8**
Piso (mmHg)a	47.6 ± 13.8	73.6 ± 23.9*	44.1 ± 8.2	67.0 ± 17.1**
Ees/Ea (MPAP)b	2.00 ± 0.86	1.31 ± 0.50*	1.80 ± 0.47	1.13 ± 0.44**
Ees/Ea (MPAPcalc)c	1.68 ± 0.90	0.67 ± 0.40*	1.51 ± 0.69	0.55 ± 0.37**
Ees/Ea (SRVP)d	0.64 ± 0.41	0.44 ± 0.36*	0.73 ± 0.20	0.47 ± 0.34**
dp/dt‐max (mmHg/s)	396.4 ± 87.0	532.4 ± 183.6*	329.3 ± 65.4	454.5 ± 113.4**
dp/dt‐min (mmHg/s)	−292.9 ± 74.3	−516.5 ± 138.9*	−280.5 ± 58.7	−491.7 ± 139.3**
Tau (s) (Weiss)	37.5 ± 19.2	45.2 ± 20.4	38.9 ± 11.0	45.4 ± 18.1
Tau (s) (Logistic)	27.7 ± 12.1	27.8 ± 12.8	37.6 ± 25.5	34.9 ± 9.7

Note: Data are shown as mean ± standard error (SE).

Abbreviations: CMRI, cardiac magnetic resonance imaging; dp/dt‐max, maximum dp/dt; dp/dt‐min, minimum dp/dt; Ea, arterial elastance; Eed, end‐diastolic elastance; Ees, end‐systolic elastance; HR, heart rate; MPAP, mean pulmonary arterial pressure; MPAPcalc, calculated end‐systolic pressure using MPAP; Piso, isovolumic pressure; RVEDP, right ventricular end‐diastolic pressure; RVEDV, right ventricular end‐diastolic volume; RVP‐min, minimum right ventricular pressure; RVSV, right ventricular stroke volume; RVV, right ventricular volume; SRVP, systolic right ventricular pressure.

Calculated using the single‐beat method.

Calculated with the following formula: (Piso/MPAP) − 1.

Calculated with the following formula: (Piso/[1.65 × MPAP − 7.79]) − 1

Calculated with the following formula: (Piso/SRVP) − 1.

Calculated with the following formula: (Piso − MPAP)/(CMRI‐derived RVSV).

Calculated with the following formula: (Piso − [1.65 × MPAP − 7.79])/(CMRI‐derived RVSV).

Calculated with the following formula: (Piso − SRVP)/(CMRI‐derived RVSV).

Calculated with the following formula: MPAP/(CMRI‐derived RVSV).

Calculated with the following formula: MPAPcalc/(CMRI‐derived RVSV).

Calculated with the following formula: SRVP/(CMRI‐derived RVSV).

Calculated by solving the equations: RVP = α (eRVV·β − 1) (n = 48 [controls, n = 10; patients with pulmonary hypertension, n = 38]).

Calculated using the following formula: Eed = α·β·eRVEDV·β (n = 48 [controls, n = 10; patients with pulmonary hypertension, n = 38]).

<0.05 versus Controls (Swan–Ganz catheterization‐derived data) by Wilcoxon's test.

0.05 versus Controls (High‐fidelity micromanometry‐derived data) by Wilcoxon's test.

DISCUSSION

The present study is the first to validate the accuracy of SG‐cath‐derived indices of RV function, using Pressure‐cath‐derived data as reference, in patients with confirmed or suspected PH. The results indicated that SG‐cath‐derived RV indices were similar to the corresponding Pressure‐cath‐derived indices, with neither significant differences nor obvious systematic trends as shown by Bland‐Altman analysis findings. In addition, SG‐cath‐derived RV indices with proven prognostic value (Ees/Ea and β) exhibited good correlations (ICC ≥ 0.6) compared with the corresponding Pressure‐cath‐derived indices.

Bachman et al. assessed RVP and calculated dp/dt‐max and dp/dt‐min via SG‐cath and found significant correlations with the corresponding Pressure‐cath‐derived values in 13 participants suspected of having PH. Their study indicated a promising role of SG‐cath‐derived pressure analysis in the evaluation of RV function. However, Piso was not calculated in their study. Consequently, the suitability of using SG‐cath data to calculate Ees and Ees/Ea was not examined. In addition, Bachman et al. calculated dp/dt‐min but did not calculate tau, β, and Eed. Thus, their study did not fully examine the accuracy of emerging SG‐cath‐derived indices of systolic and diastolic RV functions, as was performed in this study in a larger population.

Among the three RVP‐derived indices that are associated with the systolic RV function (Ees, Ees/Ea, and dp/dt‐max), Ees/Ea is considered to have the highest clinical value; moreover, it is significantly associated with the outcome of patients with PH. , , , Notably, there are several variations in the calculation methods of Ees/Ea. For example, in previous reports (including ours) , , MPAP was used as Pes, and thus Ees/Ea was calculated using the formula Piso/MPAP − 1. Alternatively, Pes is approximately equal to SRVP in patients with severe PH; hence, using SRVP as Pes may be a reasonable option to calculate Ees/Ea. Furthermore, a recent study reported that modified MPAP calculated using the formula (1.65 × MPAP − 7.79) is another surrogate candidate for Pes. With this background, in the present study, we calculated Ees/Ea (MPAP), Ees/Ea (MPAPcalc), as well as Ees/Ea (SRVP), and demonstrated insignificant differences in both SG‐ and Pressure‐cath‐derived indices. In addition, the ICCs of SG‐ and Pressure‐cath‐derived data were good for all three Ees/Ea ratios: Ees/Ea (MPAP) (ICC = 0.60), Ees/Ea (MPAPcalc) (ICC = 0.76), and Ees/Ea (SRVP) (ICC = 0.65). This indicates that Ees/Ea can be used as an accurate parameter that reflects RV‐PA coupling in patients with PH, particularly when MPAPcalc is used as Pes.

Ees/Ea can also be calculated using the “volume method” in which only volume data are used. Brewis et al. reported that volume‐method‐derived Ees/Ea (RVSV/RVESV) was an independent predictor of outcomes of patients with PA hypertension. However, volume‐method‐derived Ees/Ea is a mathematical transformation of RVEF [RVEF/(1 − RVEF)]. Further, it is considered to underestimate the true Ees/Ea, particularly in patients with PH. The clinical relevance of pressure‐ and volume‐method‐derived Ees/Ea ratios requires further investigation in future clinical studies.

Indices of diastolic RV function derived from RVP analysis include dp/dt‐min, tau, β, and Eed. Among these, β and Eed reflect RV stiffness, and are associated with the outcomes of patients with PH. , In the present study, SG‐cath‐derived β was comparable to Pressure‐cath‐derived β, whereas SG‐cath‐derived Eed tended to be higher than Pressure‐cath‐derived Eed, although the difference was not statistically significant. Meanwhile, SG‐cath‐derived β exhibited higher ICC (0.71) than SG‐cath‐derived Eed (0.57) when compared with the corresponding Pressure‐cath‐derived values. These indicate that β might be more suitable than Eed for the assessment of RV stiffness when SG‐cath recordings are used.

The following reasons may explain the differences in RV indices between SG‐ and Pressure‐cath‐based evaluations. First, the effects of whipping and/or bending of a Swan–Ganz catheter may have caused the differences. Notably, as shown in Figure 1, the pressure waves obtained by SG‐cath exhibited more fluctuations and notches than those of Pressure‐cath recordings. Such artifacts might cause differences in RVP indices between SG‐ and Pressure‐cath. Moreover, the artifacts can cause a recording of steeper incremental and decremental RVP changes, thereby increasing dp/dt‐max and decreasing dp/dt‐min. In addition, this could theoretically cause an overestimation of Piso, which would lead to overestimations of Ees and Ees/Ea. Second, the measurement of pressure using a water‐filled catheter is known to have a limited frequency response, which might cause inaccuracies. Finally, an arbitrary setting of the zero level in SG‐cath might confound recordings. In our study, the pressure transducer was set to zero at the mid‐thoracic line for SG‐cath, as recommended. However, the obtained pressure could be directly affected by the zero leveling method. These suggest that a careful setup, including zero leveling, and efforts to diminish catheter flipping and bending during data acquisition are required to minimize possible inaccuracies when SG‐cath is employed.

This study has some limitations. First, we used Pressure‐cath for the validation of SG‐cath‐derived data, although the true gold standard method for calculating RV indices is multiple pressure‐volume loop analysis (“multiple beat” method). Pressure‐cath provides more accurate pressure recordings than SG‐cath and is less invasive than the multiple beat method. However, it should be noted that we used a simplified method (“single‐beat” method) for calculating Pes and RV indices. Second, the SG‐ and Pressure‐cath were not performed simultaneously. However, the two catheterizations were performed consecutively, mostly within 15 min, and participants were instructed to maintain a stable breath hold in the same manner for both catheterizations. Third, CMRI was not performed in 23 of the 73 participants; thus, Ees, β, and Eed were calculated only in the remaining 50 participants. Fourth, different examiners performed SG‐ and Pressure‐cath, which might have affected the results; nevertheless, the use of different examiners reduced the examination duration (within 15 min in most cases) and examination‐related risks. Finally, the present study only aimed to validate the accuracy of SG‐cath‐derived indices of RV function. The relevance of using SG‐cath‐derived indices in clinical settings needs to be examined in future prospective studies on a sufficiently large number of patients.

In conclusion, this study examined the accuracy of SG‐cath‐derived indices of RV function in patients with suspected or confirmed PH, and showed that clinically relevant RV indices, such as Ees/Ea and β, were similar and exhibited good correlations with the reference values obtained by Pressure‐cath. It is expected that a suitable application of RV indices will promote a better understanding of the RV function, optimal management, and improve outcomes in patients with PH.

AUTHOR CONTRIBUTIONS

Hideki Shima and Toshitaka Nakaya generated the study concept, collected and analyzed data, and wrote the manuscript draft. Ichizo Tsujino co‐generated the study concept and collected and analyzed data. Junichi Nakamura and Hiroshi Ohira performed Swan–Ganz and pressure catheterizations and recorded right ventricular pressure. Hideki Shima, Ayako Sugimoto, Takahiro Sato, and Taku Watanabe interpreted the data and revised the manuscript. Masaru Suzuki and Masaru Kato collected data on respiratory and connective tissue disease and revised the manuscript. Isao Yokota reviewed the statistical analysis. Satoshi Konno and Ichizo Tsujino supervised the entire study, including manuscript writing.

CONFLICTS OF INTEREST

Dr. Ichizo Tsujino, Dr. Takahiro Sato, and Dr. Satoshi Konno belong to an endowed department (Division of Respiratory and Cardiovascular Innovative Research) supported by Nihon Shinyaku Co., Ltd.; Nippon Boehringer Ingelheim Co., Ltd.; and Mochida Pharmaceutical Co., Ltd. Dr. Isao Yokota reports having received funding from the Japan Society for the Promotion of Science KAKENHI and from the Health, Labour, and Welfare Policy Research Grants. Dr. Yokota also received speaker fees from Chugai Pharmaceutical Co., AstraZeneca Japan, Tobacco Pharmaceutical Division, and Nippon Shinyaku Co., outside the submitted work.

ETHICS STATEMENT

This study was approved by the Institutional Review Board of Hokkaido University Hospital for Clinical Research (No. 016‐0461). Owing to the retrospective nature of the study, informed consent was obtained on an opt‐out basis, via the website of Hokkaido University Hospital (https://www.huhp.hokudai.ac.jp/).