Pitfalls in Hemodynamic Measurement of Oxygen Saturation in Right Heart Catheterization

Global tissue oxygenation is generally best represented by the oxygen saturation measured from the pulmonary artery during right heart catheterization (RHC), as it represents the true mixed venous saturation from the blood from the inferior vena cava (IVC), superior vena cava (SVC), and coronary sinus. However, potential pitfalls may arise when attempting to obtain an accurate oxygen saturation due to intracardiac or extracardiac shunts, incorrect catheter position during measurement, technique, and various other causes. One such example was manifested in our patient, where the alteration in cardiac hemodynamics after a transjugular intrahepatic portosystemic shunt (TIPS) procedure made the oxygen saturation from the SVC a more accurate representation of systemic oxygenation.

Case Presentation

Our patient was an 87-year-old male with heart failure and an ejection fraction (EF) of 20% with an implantable cardioverter defibrillator (ICD), hypertension, hyperlipidemia, coronary artery disease with prior coronary artery bypass graft surgery (CABG) and percutaneous coronary intervention (PCI), atrial fibrillation, and liver cirrhosis with a TIPS procedure several years prior. He was admitted to the hospital for worsening dyspnea and fluid overload as evidence of worsening heart failure. After optimizing his fluid status and medical therapy, he was discharged, and was readmitted four months later with progressively worsening dyspnea and volume overload. He underwent left and right heart catheterization to guide further management of his worsening heart failure. Right heart pressures, oximetry, and cardiac outputs were obtained via a Swan-Ganz catheter inserted through a sheath in the right femoral vein and serially advanced through the right heart chambers. The patient’s history of TIPS was unknown at the time of the RHC and consequently, mixed venous oxygen saturation was measured in the pulmonary artery and was 70%. There was no change in medical management at the time.

The patient was readmitted one month later with worsening heart failure symptoms, including dyspnea at rest and progression to New York Heart Association (NYHA) Class IV, Stage D heart failure with inability to tolerate optimal medical therapy due to hypotension and dizziness. He was not a candidate for a left ventricular assist device or heart transplant due to pulmonary artery hypertension, advanced age, and liver cirrhosis. He had a repeat RHC to assess whether he would benefit from home inotrope therapy on discharge. With the knowledge of his history of TIPS, the oxygen saturation was measured from the SVC and was 55% (Figure 1). He was discharged home on intravenous milrinone after showing positive response to milrinone as palliative treatment for end-stage heart failure with improvement in symptoms. The discrepancy between the oximetry measurement of 70% in the pulmonary artery and 55% in the SVC after TIPS raised the question of where best to measure the mixed venous oxygenation after TIPS, and what pitfalls may arise in obtaining and accurately interpreting oxygen saturation measurements during a RHC.

The Effect of TIPS on Cardiac Hemodynamics

The transjugular intrahepatic portosystemic shunt (TIPS) procedure inserts a shunt between the portal vein and the hepatic veins to decrease portal hypertension (Table 1). Our patient had TIPS several years prior in order to decrease portal hypertension secondary to liver cirrhosis. At baseline, cirrhosis is associated with a hyperdynamic circulation, characterized by increased cardiac output, systemic vasodilatation with low systemic vascular resistance, and increased arterial compliance.¹ TIPS often exacerbates the baseline hyperdynamic circulatory state in cirrhosis by decompressing the portal venous system and shunting blood to the hepatic veins that drain into the IVC. Consequently, there is increased preload, leading to an increase in cardiac output, and increased right atrial, pulmonary, and pulmonary capillary wedge pressure on RHC, as well as a compensatory vasodilatory response in the systemic and pulmonary circulations.²

Measurement of Systemic Oxygenation and the Influence of TIPS

There are different measures of systemic oxygenation, which include the mixed venous oxygen saturation (SvO2), for which normal values are 65-75%, and the central venous O2 saturation (ScvO2), for which normal values are 70-80%.³ The SvO2 is drawn from the tip of the pulmonary artery catheter and represents the true mixed venous oxygen, because all the venous blood combines from the IVC (blood from the gastrointestinal tract and lower extremities), SVC (venous blood from the head and arms), and coronary sinus blood (from the coronary veins).  SvO2 is the percentage of residual oxygen bound to hemoglobin in blood returning to the right side of the heart after tissue extraction of oxygen.⁴

The tissues require oxygen to make adenosine triphosphate (ATP) for cell function and survival. If the tissues do not receive adequate oxygen to meet their needs, because of increased demand or decreased supply, then the body compensates by first increasing cardiac output (CO) and next by extracting more oxygen from the arterial blood.⁵ There is a consequent reduction in the SvO2 as the blood returning to the right heart is less oxygenated. The decreased ScvO2 of 55% measured from the SVC in this decompensated heart failure patient is likely due to first attempting to increase cardiac output, which is limited in decompensated heart failure, and then due to increased extraction of oxygen from arterial blood to meet the tissues’ requirement. The low ScvO₂ indicates that the cardiac output is not high enough to meet tissue oxygen requirements.⁴ 

If TIPS increases the cardiac output, which increases the oxygen delivery, the SvO2 should increase, which may explain why the SvO2 was 70% on the first RHC, as measured in the pulmonary artery. The ScvO2 of 55% from the SVC, where the venous blood is largely from the cerebral circulation and upper extremities, seems more appropriate for a decompensated heart failure patient. The blood in the SVC was presumably less influenced by the shunted blood from the TIPS to the hepatic veins and IVC. The SvO2 measured in the pulmonary artery had blood from the IVC and shunted blood within it. 

The blood from the portal venous system bypassing the cirrhotic liver and being shunted to the hepatic veins and IVC is likely a contributing factor to the oxygen saturation being higher in the pulmonary artery than the SVC after TIPS. The portal venous oxygen saturation varies with gut activity and is approximately 85% in the fasting state⁶; otherwise the portal venous saturation is around 60-70%.⁷ Hepatic arterial oxygen saturation is around 95%.⁸ When there is decreased portal venous flow to the liver, the hepatic arterial flow increases as a compensatory mechanism.⁹ After TIPS, there is presumably an increase in hepatic arterial flow to the liver as a compensatory mechanism for the decreased portal venous flow. Therefore, after TIPS, the oxygen saturation in the pulmonary artery may have had a higher oxygen saturation of 70% than the 55% in the SVC. The pulmonary artery contained blood from the IVC, which was shunted from the portal vein bypassing the liver, with resultant reduced hepatic oxygen extraction and increased hepatic arterial blood.  

Indication for Right Heart Catheterization 

A RHC has many diagnostic and therapeutic uses. The indication for a RHC in this patient was to determine whether he would benefit from intravenous inotrope therapy as palliative treatment for class IV, Stage D heart failure, since our patient was not a left ventricular assist device or heart transplant candidate. 

Obtaining Oxygen Saturation During Routine RHC and Potential Pitfalls 

During routine RHC, oxygen saturation from the pulmonary artery is obtained as an estimate of global oxygenation, because complete mixing of venous blood does not occur until the blood from the SVC, IVC, and coronary sinus mix in the pulmonary artery. However, there are circumstances such as intracardiac or extracardiac shunts (such as TIPS, arteriovenous fistula), heart failure, or techniques during RHC that complicate obtaining an accurate oxygen saturation measurement representative of systemic oxygenation. Common sources of inaccuracy should be considered, such as the use of sedation and supplemental oxygen during the procedure. Another example was the discrepancy between the oxygen saturation from the SVC and PA in our patient with advanced heart failure after TIPS. Following are some additional potential pitfalls in obtaining accurate oxygen saturation during a RHC.

Wedging Too Deeply

Catheter position is important. Wedging the pulmonary artery catheter too deeply can be a cause of obtaining an inaccurate mixed venous oxygen saturation, as it pulls oxygenated blood from the pulmonary venous system and may give an overestimation of the true mixed venous oxygen saturation. Initially, when a mixed venous saturation of 70% was obtained from the pulmonary artery in this patient, the catheter was pulled back and repeat values were similar. Correct wedging by the balloon floatation catheter can be validated by withdrawing blood from the distal port during balloon inflation to ensure the sample matches the systemic arterial saturation. Pressure measurements at the site can confirm catheter tip position. 

Absence of a Steady State

The absence of a steady state while obtaining the samples is a potential source of error and is possibly caused by a prolonged time to obtain serial values. Start with the pulmonary capillary wedge and proceed serially to the pulmonary artery, right ventricle, right atrium, and finally to the SVC and IVC. 

Shunts

The patient’s clinical presentation often raises suspicion for an intracardiac shunt prior to a RHC. However, unexplained arterial desaturation during a RHC raises the possibility of a right-to-left shunt while unexpectedly high oxygen saturation from the pulmonary artery makes a left-to-right shunt possible.¹⁰

Left-to-Right Intracardiac Shunts

During a standard RHC where no shunt is suspected, O2 saturation is measured from the pulmonary artery.¹¹ If the oxygen saturation obtained from the pulmonary artery is greater than 75% on repeated measurement and is not well explained, for example, by a high output state or arteriovenous fistula, an oximetry run should be performed. In our patient, the history of TIPS was unknown at the time of the first RHC. Therefore, a saturation run was not performed. In an oximetry run, oxygen saturations are obtained from the SVC and IVC, right atrium, right ventricle, and pulmonary artery in order to quantify and detect possible left-to-right shunts. A step up in oxygen saturation of 5-7% between cardiac chambers or a great vessel is considered significant, and the left-to-right shunt should be localized.¹⁰

The optimal site for measurement of venous oxygen saturation in the presence of a left-to-right shunt is variable. Some operators recommend averaging the saturations from the SVC and the IVC.¹² Others recommend weighing the sample one-third SVC to two-thirds IVC, because approximately two-thirds of venous return is through the IVC and one-third from the SVC.¹³ Some use the oxygen saturation from the SVC to approximate the mixed venous saturation, because the high saturation of the IVC blood is thought to be offset by the low saturation from the coronary sinus.¹⁴

In a study by Gutgesell et al, hemodynamic data was obtained from 35 patients with no hemodynamic abnormalities on RHC such as intracardiac shunts.¹⁴ Sequential oxygen saturations were obtained during an oximetry run from the pulmonary artery to the SVC and IVC. The pulmonary artery saturation was accepted as true mixed venous saturation. The pulmonary artery saturation was compared to the saturation in the IVC, SVC, (SVC + IVC )/2, (SVC + 2 IVC)/3, (3 SVC + IVC)/4. The results showed that the SVC sample was the closest approximation of the pulmonary artery oxygen saturation and was, therefore, the best indicator of mixed venous saturation. The SVC saturation was an average of 0.6% lower than the pulmonary artery saturation, arithmetically closer to the pulmonary artery saturation more often than the IVC sample or combinations of the two. 

The inclusion of the IVC saturation in the estimate of mixed venous saturation makes the result less accurate, because of a variable amount of highly saturated renal venous blood and unsaturated hepatic venous return. The IVC sample is also extremely dependent upon catheter position.¹⁴ Kjellberg et al found that a simple rotation of the catheter in the IVC changed the saturation an average of 5.8%.¹⁵ In patients with left-to-right atrial shunts, the IVC sample may, in addition, be contaminated by oxygenated blood spilling across the atrial septum. 

The best approximation of the mixed venous saturation weights the SVC heavily and IVC little. The use of the SVC value alone is a simple and accurate estimate of mixed venous saturation.¹⁴

Right-to-Left Shunt

Right-to-left intracardiac shunts can be suspected in the context of unexpected arterial desaturation to less than 95%. The most common causes include alveolar hypoventilation associated with physiologic right-to-left shunt including chronic obstructive pulmonary disease (COPD) or other pulmonary disease, oversedation, and pulmonary edema, which is worsened by the supine position during the catheterization.¹³ Most cases improve with the patient sitting upright, taking deep breaths, coughing, and supplemental oxygen. If arterial blood oxygen saturation is not improved significantly with supplemental oxygen, a right-to-left shunt is presumed, and localization and quantification can be determined by RHC.

Conclusion

The discrepancy between the oximetry measurement of 70% from the pulmonary artery and 55% from the SVC in this patient after TIPS raised the question of where the appropriate location is to measure the mixed venous oxygenation after TIPS. The oximetry measured from the SVC was likely a more accurate representation of the mixed venous oxygenation after TIPS in this patient with end-stage heart failure whose SvO2 was less likely to be 70%. The expected decrease in oxygenation in the pulmonary artery as compared to the SVC is surmised to be secondary to myocardial oxygen extraction and mixing of venous blood from the IVC, coronary sinus, and Thebesian veins. In addition, in low-output states such as in decompensated heart failure, the oxygen saturation in the SVC is expected to be higher than the oxygenation measured in the pulmonary artery, due to redistribution of blood flow away from femoral, splanchnic, and renal circulation, and preferential preservation of cerebral blood flow.¹⁶ In this decompensated heart failure patient, the SvO2 measured from the pulmonary artery was higher than the oxygenation measured in the SVC. This is not consistent with the slight, expected step down in oxygen from the SVC to the PA, due to myocardial oxygen extraction, or the slightly increased oxygenation measured in the SVC as expected with congestive heart failure from preferential preservation of cerebral blood flow. The shunting of blood to the hepatic veins and IVC after TIPS, which increases the preload and oxygen delivery to the heart, is presumed to lead to the increased SvO2 measurement in the pulmonary artery. Additionally, the oxygen saturation in the pulmonary artery may have been higher than the SVC after TIPS, because the pulmonary artery contained blood from the IVC that was shunted from the portal vein, bypassing the liver with resultant reduced hepatic oxygen extraction and increased hepatic arterial blood. We feel that the most accurate site for measurement of systemic oxygenation during right heart catheterization after a TIPS procedure is the SVC. Consequently, we reviewed some potential pitfalls that may arise in a RHC that may complicate obtaining a value that most accurately represents the global venous oxygenation, including intracardiac or extracardiac shunts, incorrect catheter position during measurement, and technique.

References

1. Colombato LA, Spahr L, Martinet JP, et al. Haemodynamic adaptation two months after transjugular intrahepatic portosystemic shunt (TIPS) in cirrhotic patients. Gut. 1996 Oct; 39(4): 600-604.
2. Saugel B, Mair S, Meidert AS, et al. The effects of transjugular intrahepatic portosystemic stent shunt on systemic cardiocirculatory parameters. J Crit Care. 2014 Dec; 29(6): 1001-1005.
3. Marino PL. Systemic oxygenation. In: The ICU Book. Philadelphia, PA: Lippincott Williams and Wilkins; 2016: 93-105.
4. Maddirala S, Khan A. Optimizing hemodynamic support in septic shock using central and mixed venous oxygen saturation. Crit Care Clin. 2010 Apr;26(2):323-33, table of contents. 
5. Morgan B. SvO2 (mixed venous oxygen saturation) or ScvO2 (central venous oxygen saturation). July 24, 2012. EDUBRIEFS in CCTC. London Health Sciences Centre. Available online at https://www.lhsc.on.ca/Health_Professionals/CCTC/edubriefs/svo2.htm. Accessed April 20, 2018. 
6. Kam P, Power I. Principles of Physiology for the Anaesthetist. Boca Raton, FL: CRC Press; 2015.
7. Pandey CK, Nath SS, Tripathi M. Hepatic and Biliary Diseases: Anesthesiologists’ Perspective. New Delhi, India: Jaypee Brothers Medical Publishers; 2012.
8. Eipel C, Abshagen K, Vollmar B. Regulation of hepatic blood flow: The hepatic arterial buffer response revisited. World J Gastroenterol. 2010 Dec 28;16(48): 6046-6057.
9. Mücke I, Richter S, Menger MD, Vollmar B. Significance of hepatic arterial responsiveness for adequate tissue oxygenation upon portal vein occlusion in cirrhotic livers. Int J Colorectal Dis. 2000 Nov; 15(5-6): 335-341.
10. Baim DS, Grossman W, eds. Grossman’s Cardiac Catheterization, Angioplasty, and Intervention. 7th ed. Philadelphia, PA: Lippincott Williams and Wilkins; 2006.
11. Patel R, Rajamanickam A, Nair A. Hemodynamic Assessment. In: Kini A, Sharma S, Narula J, eds. Practical Manual of Interventional Cardiology. London, England: Springer-Verlag London; 2014.
12. Yang SS, Bentivoglio LG, Maranhao V, Goldberg H. From Cardiac Catheterization Data to Hemodynamic Parameters. Philadelphia, PA: F.A. Davis Company; 1972.
13. Weidman WH, Swan HJC, DuShane JW, Wood EH. A hemodynamic study of atrial septal defects and associated anomalies involving the atrial septum. J Lab Clin Med. 1957 Aug; 50(2): 165-185.
14. Gutgesell HP, Williams RL. Caval samples as indicators of mixed venous oxygen saturation: implications in atrial septal defect. Cardiovasc Dis. 1974; 1(3): 160-164.
15. Kjellberg SR, Mannheimer E, Rudhe U, Jonsson B. Diagnosis of Congenital Heart Disease. Chicago, IL: The Yearbook Publishers, Inc; 1959.
16. Scheinman M, Brown M, Rapaport E. Critical assessment of use of central venous oxygen saturation as a mirror of mixed venous oxygen in severely ill cardiac patients. Circulation. 1969 August; 40(2): 165-172.