Figura 1. Linear regression analysis. Heart rate, as quantified by right heart catheterization and ...

Figura 2. Bland-Altman plot for heart rate between right heart catheterization and impedance cardio...

Figura 3. Comparison of ideal and actual agreement between heart rates measured by right heart cathe...

Figura 4. Bland-Altman plot for cardiac index between right heart catheterization and impedance car...

Figura 5. Bland-Altman plot for cardiac index between right heart catheterization and impedance car...

Figura 6. Comparison of ideal and actual agreement between cardiac index (CI) measured by right hear...

Figura 7. Linear regression analysis. Systemic vascular resistance (SVR), as quantified by right he...

Figura 8. Bland-Altman plot for systemic vascular resistance between right heart catheterization and...

Tabla 1. Demographic characteristics of the patient sample, including subjects with and without pu...

Tabla 2. Hemodynamic characteristics obtained through right heart catheterization and impedance car...

Tabla 3. Correlation between hemodynamic variables.

Tabla 4. Agreement between hemodynamic variables.

Tabla 5. Linear regression

Figura 9. Comparison of ideal and actual agreement between systemic vascular resistance measured by...

Introduction

Pulmonary hypertension (PH) is a pathophysiological alteration characterized by elevation in the pressure of pulmonary circulation linked to multiple clinical entities. In general, it tends to complicate most cardiovascular and respiratory diseases1.

Its identification is based on the determination of a mean pulmonary artery pressure (mPAP) >20 mmHg measured by right-heart catheterization (RHC). This value, along with the estimation of capillary or wedge pressure (WP) and pulmonary vascular resistance (PVR), allows for the classification of PH into the categories below. 1) Pre-capillary PH: PVR >2 Wood Units (WU) and WP ≤15 mmHg. 2) Isolated post-capillary PH: PVR < 2 WU and WP >15 mmHg. 3) Combined PH (both pre- and post-capillary): PVR >2 WU and WP >15 mmHg2.

RHC not only allows for the diagnosis and classification of PH but also assesses its severity and determines its prognosis. In this context, RHC allows us to assess the response to volume overload (saline solution), perform pulmonary angiography, and finally, conduct a pulmonary vasoreactivity test3.

The pulmonary vasoreactivity test is performed during right-heart catheterization in patients with pulmonary arterial hypertension to assess acute responders who may be candidates for treatment with high-dose calcium channel blockers.

RHC is an invasive, complex procedure that requires operator experience. Therefore, it should be performed in centers with adequately trained operators to obtain precise and reproducible information while minimizing its risks. It is an invasive method that might cause adverse effects while being conducted.

Impedance cardiography (IC) is a non-invasive method whereby an electrical representation of cardiac flow is obtained placing 6 electrodes on the patient’s chest.

It is a method frequently used in patients with heart failure and/or arterial hypertension4.

There is limited evidence correlating data obtained from RHC and IC for monitoring patients with PH.

The aim of this work was to prospectively assess the correlation and agreement of systemic vascular resistance (SVR) and cardiac index (CI) measurements obtained by means of RHC and IC in a population of patients with PH, while also assessing the correlation between intrathoracic fluid volume and WP obtained by both methods.

Materials and methods

Population. The study was conducted in a prospective cohort of patients referred to a private healthcare center in the city of Trelew, Chubut, for RHC due to suspected PH or for PH follow-up, from September 2021 to April 2023. Patients included in the study were over 18 years old and had to sign an informed consent form. All patients with inconclusive PH results on RHC who could not have an IC within 4 hours after RHC or who had poor IC signal were excluded.

Variables. As described below, there were various hemodynamic data collected, along with clinical data such as age, gender, comorbidities, PH type, and use of specific PH medication (phosphodiesterase-5 inhibitors, endothelin inhibitors, and/or prostanoids).

Right-heart catheterization. Hemodynamic assessment was conducted in supine position and under room temperature conditions. A 7-Fr introducer was inserted via the right jugular or femoral vein, through which a Swan-Ganz catheter with four lumens and a balloon tip (Arrow™) was advanced. A multiparametric monitor was used to control intracavity pressure and advance to the pulmonary artery. This was confirmed when a pressure curve was registered in the pulmonary artery. The hemodynamic assessment included the measurement of right atrial pressure (RAP), mean pulmonary artery pressure (mPAP), pulmonary artery wedge pressure (PAWP), and cardiac output (CO).

The cardiac index (CI) was calculated as:

CI = CO/body surface area

Pulmonary vascular resistance (PVR) was calculated as:

PVR = (mPAP – PAWP)/CO

Impedance cardiography. CO measurements by impedance cardiography used a Z-Logic™ (Exxer, Argentina) device. They were all conducted by the same operator, who was unaware of RHC results. Two electrodes, one transmitter and one receiver, were placed towards the left at the base of the neck, while two more were placed along the xiphoid area. Impedance variations were recorded after a weak (3.8 mA) but high-frequency (75 Hz) electric current was passed. They represent variations in blood flow in the chest. The following equation was used to determine stroke volume according to impedance variation over time:

SV = LVET [(dZ/dt) max/Z0]

Where TEGV is the ejection time from the left ventricle, K is a constant based on patient sex, age, and size, Z0 is the baseline impedance, and dZ/dt is the variation in impedance over time. The signal was recorded for 20 minutes, and then the results were averaged.

Statistical analysis. Variables were evaluated for distribution using the Shapiro-Wilk test, the skewness and kurtosis normality test, and graphical methods (histograms and standardized normal probability plots). According to their distribution, variables were reported as mean and standard deviation (SD) for normal distribution, and median and interquartile range (IQR) for non-normal distribution. In the case of discrete variables, their relative frequency was determined.

Different continuous variables, according to their normal or non-normal distribution, were compared using Student’s t-test or the signed-rank test for paired data with respect to hemodynamic and impedance cardiography measurements A p-value < 0.05 was considered statistically significant.

Researchers analyzed the correlation between continuous variables (stroke volume, cardiac index, systemic vascular resistance, compliance, thoracic fluid content (TFC), and wedge pressure, and TFC and right atrial pressure for both methods (RHC and IC). Additionally, heart rate (HR), systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean arterial pressure (MAP) were correlated between the two measurement methods. Pearson’s test was used for normally distributed variables, and Spearman’s test was used for non-normal distributions. A p-value < 0.05 was considered statistically significant.

Agreement between methods was assessed using Bland-Altman plots and Lin’s concordance test, with agreement considered when p < 0.05.

In cases when the distribution of the difference between the bivariable averages showed a non-normal distribution pattern, a logarithmic (ln) transformation of the data was performed.

Statistical analysis was carried out using the Stata 14 (StataCorp LLC, Texas, USA) software.

Results

Selected patients. This study included 17 patients, with a mean age of 55 years; 82.35% of subjects were women. The median body surface area was 1.84 m2 for the whole cohort. Of the 17 patients, 12 had group I PH, 2 had group II PH, 2 had group III PH, and 1 patient had group IV PH. There were no patients with group V PH. Regarding classic cardiovascular risk factors, 29.4% had hypertension, 17.65% had diabetes, and 23.53% had dyslipidemia. Regarding New York Heart Association (NYHA) classification, 58.8% were in functional class III, 23.5% in functional class II, and 17.65% in functional class I. Regarding treatment with phosphodiesterase-5 inhibitors (PDE-5), 52.94% used sildenafil, and 5.88% used tadalafil. Additionally, 35.39% of PH patients used macitentan, 5.88% used ambrisentan, and 5.88% used intravenous epoprostenol only.

Table 1 summarizes the clinical and epidemiological characteristics of study patients.

Comparison between methods. The CI, SVR, and compliance values measured by hemodilution by RHC and by IC were: CI 3.01 L/m/m2 BSA (IQR: 2.36-3.49) vs. 2.4 L/m/m2 BSA (IQR: 2.2-3.1) (p=0.0109); PVR 1200 dyn/s/cm5 (IQR: 857-1631) vs. 1201 dyn/s/cm5 (IQR: 870-1865) (p=0.0627), and compliance 2.26 mL/mmHg (IQR: 1.69-4.47) vs. 1.53 mL/mmHg (IQR: 1.27-2.35) (p=0.0758). Regarding heart rate, the values quantified by right-heart catheterization were 80.76 bpm±15.31 bpm, and 75.82 bpm±15.82 bpm (p=0.1745).

Table 2 shows the differences between both groups for other hemodynamic variables measured by right heart catheterization and impedance cardiography.

Analysis of hemodynamic variables

Heart rate. HR values measured by RHC and by IC were 80.76±15.31 vs. 75.82±15.31, respectively. The correlation between both variables is significant (Pearson coefficient 0.56, p=0.019). The linear regression equation between these methods has a slope of 0.5616 (95% confidence interval [CI]: 0.1062043-1.017012) and an intercept of 38.1816 (p=0.019). The coefficient of determination (R2) was 0.31 (Figure 1).

The Bland-Altman analysis for all CI measurements obtained by RHC and IC showed a mean difference of 4.94 (95% CI: -2.430395-12.31275) with agreement limits between -23.15 and 33.04 (Figure 2). Lin’s concordance coefficient (rho_c) was 0.5302, demonstrating statistically significant agreement between both methods (p=0.016).

The comparison between the obtained concordance model and an ideal concordance model between both techniques showed a similar distribution of variables and parallel lines (Figure 3).

Cardiac index. CI values measured by RHC and by IC were 3.01 L/min/m2 (IQR: 2.36-3.49) and 2.4 L/min/m2 (IQR: 2.2-3.1), respectively (p= 0.019. The correlation between both variables is significant (Spearman rho coefficient: 0.60; p=0.009). The linear regression equation between these methods has a slope of 0.4633 (95% CI: 0.0862-0.8404) and an intercept of 0.6758. The coefficient of determination (R²) was 0.25 (p=0.039) (Figure 4).

The Bland-Altman analysis for all CI measurements obtained by RHC and IC showed a mean difference of 0.18 (logarithmic transformation, Shapiro-Wilk normality test for Y-X; p=0.0027) with agreement limits between -0.53 and 0.91 (Figure 5). Lin’s concordance coefficient was 0.4459, demonstrating statistically significant agreement between both methods (p=0.016).

The comparison between the obtained concordance model and an ideal concordance model between both techniques showed a similar distribution of variables and parallel lines (Figure 6).

Systemic vascular resistance. The median SVR by thermodilution via RHC was 1200 dyn/s/cm5, with IQR 857-1631, while the median SVR by IC was 1201 dyn/s/cm5, with IQR 870-1865. These variables tended to correlate, but such correlation did not reach statistical significance (Spearman coefficient 0.4608, p=0.0627). The linear regression equation between these methods has a slope of 0.2785 (95% CI: 0.0205-0.5364) and an intercept of 863.19 (p=0.036). The coefficient of determination (R²): was 0.1566 (Figure 7)

The Bland-Altman analysis for all CI measurements obtained by RHC and IC showed a difference between averages of -0.1393 (logarithmic transformation, Shapiro-Wilk normality test for Y-X; p=0.0111) with agreement limits between -1.05 and 0.77 (Figure 8). Lin’s concordance coefficient was 0.4195, demonstrating statistically significant agreement between both methods (p=0.038).

The comparison between the obtained concordance model and an ideal concordance model between both techniques showed a similar distribution of variables and almost parallel lines (Figure 9).

Wedge pressure and intrathoracic fluid content. Wedge pressure (WP) measured by RHC was 10.47 mmHg±5.63, while TFC measured by IC was 39.6 ohm±7.08. The correlation between both variables was not statistically significant (Pearson coefficient -0.1033; p=0.6933). The linear regression equation between these methods has a slope of -0.8211 (95% CI: -0.517396 to 0.3531679) and an intercept of 13.7223. No linear relationship was demonstrated between the two techniques (p=0.69) (Figure 10).

Table 3 lists all the correlations between variables, Table 4 shows all the concordance calculations for all compared variables, and Table 5 presents the values obtained from simple linear regression performed in all comparisons.

Discussion

Impedance cardiography is a highly validated method in heart failure and arterial hypertension. It is not extensively studied in patients with pulmonary hypertension.

In decompensated heart failure, the increase in thoracic fluid content during pulmonary edema or pleural effusion would explain the decrease in total thoracic impedance in CI, as biological fluids are excellent conductors of electricity5.

The utility of CI in predicting acute decompensation due to heart failure caused by pulmonary edema was evaluated in a group of 33 patients with chronic heart failure, NYHA functional class III and IV. This study found a significant reduction in baseline impedance values before hospitalization in all patients who experienced decompensation and, more importantly, that this reduction started an average of 15 days before the onset of decompensation symptoms, which started on average 3 days before hospitalization6.

The utility of CI in identifying the risk of clinical decompensation in patients with chronic heart failure was also assessed in the PREDICT study, which analyzed the value of various clinical variables, as well as of hemodynamic parameters determined by CI, in predicting acute decompensation in 212 patients with stable heart failure and NYHA functional class II-IV. This study demonstrated statistically significant differences in a large number of hemodynamic parameters estimated by CI between the control group and patients who required hospitalization for acute decompensation. The three hemodynamic parameters that showed the greatest predictive value for acute decompensation included increased thoracic fluid content index, decreased velocity index, and decreased ventricular ejection time. The prognostic value of these indices, when considered together, was even greater than that for other variables such as the NYHA functional class and systolic blood pressure levels7.

As for arterial hypertension, several studies have been conducted to assess whether a new approach to managing arterial hypertension based on the use of CI improves the achievement of control goals in hypertensive patients8.

By allowing for the assessment of stroke volume and other derived hemodynamic parameters (cardiac index, myocardial contractility, and peripheral vascular resistance), CI helps us reach a hemodynamic characterization of hypertensive patients and therefore enables individualized pharmacological management9.

In a recent clinical trial, Taler et al. randomized 104 patients with difficult-to-control arterial hypertension to either conventional management or CI-guided hypertension management. After three months of pharmacological treatment, in addition to a significant reduction in blood pressure values, blood pressure control goals were achieved more frequently in the CI-guided management group than in the control group, which was managed by a hypertension specialist10.

A post hoc analysis of multicenter CONTROL study reached similar results: 164 patients with poorly controlled hypertension were randomized to conventional clinical management or a CI-guided therapeutic strategy. When comparing both groups after three months of treatment, blood pressure control was achieved significantly more frequently in the CI-guided treatment group, even when control goals were defined with very strict blood pressure values11.

Considering the evidence provided by various clinical studies that have demonstrated the usefulness of CI in the management of arterial hypertension by differentiating the different hypertensive phenotypes (hyperdynamic, vascular hyperreactivity, volume overload, mixed), individualized management of arterial hypertension based on altered hemodynamic indices could be considered, allowing for the withdrawal of unnecessary medication.

In pulmonary hypertension, there is limited evidence on the use of impedance cardiography for the management of this challenging group of patients. In 2018, Marion Dupuis et al. presented a prospective study of 75 patients at University of Toulouse, France. They included patients over 18 years old with Group I or Group IV PH. They compared RHC with CI using the PHYSIOFLOW device and the Fick method. Their aim was to assess the correlation between cardiac output by both methods, achieving a positive result with a regression equation that had a slope of 0.7±0.1 and an intercept of 1.7±0.6, correlated with r=0.365 (p < 0.001). This study did not seek to assess systemic vascular resistance or attempt to correlate total body water measured by CI with central venous pressure or capillary pressure12.

As for our study, our main finding was that there is indeed a correlation and agreement between cardiac index, systemic vascular resistance, and heart rate measured by RHC and CI, non-invasively.

Because diagnosing PH truly requires RHC (an invasive study) to assess hemodynamic parameters, we understand that CI could not replace it in this task.

Complications associated with the use of right heart catheterization are reported in 2% to 17% of cases. The most frequent events are local complications at the puncture site, but more severe complications such as pseudoaneurysms, malignant arrhythmias, and arterial ruptures are also described13.

Patients with PH require periodic monitoring of hemodynamic parameters to assess systemic vascular resistance and cardiac index to monitor changes in treatment, for example. This scenario could, according to our experience, present a niche for IC. The mean difference between cardiac index measurements obtained by right heart catheterization and impedance cardiography was 0.6 L/min/m2.

The correlation of systemic vascular resistance measurements was positive.

There was no correlation between wedge pressure and total body water measured by CI. This supports the idea that the non-invasive method is not aimed at replacing right heart catheterization for PH diagnosis, as the measurement of wedge pressure is crucial to defining whether pulmonary hypertension is pre- or post-capillary.

These results provide valuable information about the relationship between different variables measured by right heart catheterization and impedance cardiography, as well as about the agreement between the measurements obtained by these two methods.

Limitations

The main limitation of this study is its small sample size, in addition to a lack of sample size calculation prior to recruitment. Furthermore, more patient epidemiological data (e.g., comorbidities) could have been collected so as to adjust more variables in regression models to assess the correlation of the studied measurements.

Conclusions

Impedance cardiography used in pulmonary hypertension appears as a method with good correlation to right heart catheterization in parameters such as cardiac index, systemic vascular resistance, and heart rate.

There is no correlation between wedge pressure and total body water.

Impedance cardiography could be used for follow-up and monitoring of patients with a previously established diagnosis of pulmonary arterial hypertension.

Acknowledgments

To my family, always a fundamental support in my career; to my former colleagues who collaborated with me in the creation of part of this material, and to my co-authors for their help and experience.

Key points

Current knowledge: There is little evidence on impedance cardiography in pulmonary hypertension.

Contribution of this article to current knowledge: Our work provides a starting point for research on impedance cardiography for the follow-up of patients with pulmonary hypertension, so as to monitor cardiac index, systemic vascular resistance, and heart rate by non-invasive means.

1. Humbert M, Kovacs G, Hoeper MM, et al. 2022 ESC/ERS Guidelines for the diagnosis and treatment of pulmonary hypertension. Eur Heart J. 2022 Oct 11;43(38):3618-3731.

2. Simonneau G, Montani D, Celermajer DS, et al. Haemodynamic definitions and updated clinical classification of pulmonary hypertension. Eur Respir J. 2019 Jan 24;53(1):1801913.

3. Coronel LM, Diez M, Lema LR, et al. Recomendaciones para la realización de cateterismo cardíaco derecho en hipertensión pulmonar 2022. Rev Argent Cardiol 2022;90 (Suplemento 7):1-22.

4. Myers J, Wong M, Adhikarla C, et al. Cardiopulmonary and noninvasive hemodynamic responses to exercise predict outcomes in heart failure. J Card Fail. 2013 Feb;19(2):101-7.

5. Vollmann D, Nagele H, Schauerte P, et al. Clinical utility of intrathoracic impedance monitoring to alert patients with an implanted device of deteriorating chronic heart failure. Eur Heart J 2007;28:1835-40.

6. Yu CM, Wang L, Chau E, et al.Intrathoracic impedance monitoring in patients with heart failure: correlation with fluid status and feasibility of early warning preceding hospitalization. Circulation 2005;112:841-8.

7. Packer M, Abraham WT, Mehra MR, et al. Utility of impedance cardiography for the identification of short-term risk of clinical decompensation in stable patients with chronic heart failure. J Am Coll Cardiol 2006;47:2245-52.

8. Parrott CW, Quale C, Lewis DL, Ferguson S, Brunt R, Glass S. Systolic blood pressure does not reliably identify vasoactive status in chronic heart failure. Am J Hypertens 2005;18:82S-6S.

9. Ferrario CM, Flack JM, Strobeck JE, Smits G, Peters C. Individualizing hypertension treatment with impedance cardiography: a meta-analysis of published trials. Ther Adv Cardiovasc Dis 2010;4:5-16

10. Taler SJ, Textor SC, Augustine JE. Resistant hypertension: comparing hemodynamic management to specialist care. Hypertension 2002;39:982-8.

11. Smith RD, Levy P, Ferrario CM. Value of noninvasive hemodynamics to achieve blood pressure control in hypertensive subjects. Hypertension 2006;47:771-7.

12. Dupuis, M., Noel-Savina, E., Prévot, et al. (2018). Determination of cardiac output in pulmonary hypertension using impedance cardiography. Respiration 2018,96(6),500-506.

13. Hoeper MM, Lee SH, Voswinckel R, et al. Complications of right heart catheterization procedures in patients with pulmonary hypertension in experienced centers. J Am Coll Cardiol. 2006 Dec 19;48(12):2546-52.

Para descargar el PDF del artículo
Correlation of hemodynamic variables measured by impedance cardiography and right heart catheterization in pulmonary hypertension: a prospective cohort study

Haga click aquí