Correlation of Near-Infrared Spectroscopy Oximetry and Corresponding Venous Oxygen Saturations in Children with Congenital Heart Disease

Loomba RS, Rausa J, Sheikholeslami D, Dyson AE, Farias JS, Villarreal EG, Flores S, Bronicki RA.

Pediatr Cardiol. 2022 Jan;43(1):197-206. doi: 10.1007/s00246-021-02718-7. Epub 2021 Aug 30.PMID: 34459948


Commentary from Dr. Kesava Ramakrishnan (Abu Dhabi, UAE), section editor of Congenital Heart Anesthesia and Intensive Care Journal Watch








Take Home Points:

  • There is a strong correlation between SVC saturation and cerebral NIRS oximetry that is not affected by the size of the patient, univentricular heart or the make of the monitor used
  • There is a strong correlation between IVC saturation and renal NIRS oximetry
  • NIRS monitors offer a method for continuous monitoring of tissue oxygenation
  • The limits of agreement between NIRS oximetry and SVC/IVC saturation are wide, hence NIRS oximetry should be used as a trend monitor.
  • To increase the sensitivity of NIRS monitoring to detect low cardiac output state, multisite monitoring with the cerebral and somatic sites is recommended.



A basic tenet of critical care medicine is to ensure adequate tissue oxygenation1. Nevertheless, clinicians often struggle to assess the adequacy of oxygen delivery to essential organs. Clinical examination, hemodynamic measures (e.g., heart rate, blood pressure, central venous pressure, urine output etc.) and laboratory parameters (e.g., lactate, acid-based status) are often insensitive and late markers of tissue oxygenation2. According to the Fick equation, oxygen consumption (VO2) equals Cardiac Output (CO) times the difference in arterio-venous oxygen content (CaO2- CvO2)1. As oxygen is poorly soluble in plasma, the amount of oxygen in a dissolved state is inconsequential. Hence, ignoring the dissolved oxygen, one can derive that the arterial (SaO2) to mixed-venous saturation (MVO2) difference reflects the balance between VO2 and oxygen delivery (DO2). Anaerobic metabolism and increased lactate production do not occur until tissue hypoxemia ensures. But, as the DO2 decreases, the difference between SaO2-MVO2 widens, reflecting the imbalance of VO2/DO21. This provides an opportunity to intervene before tissue hypoxia ensues.


Measuring mixed venous saturation requires pulmonary artery catheters, which are seldom used in children. Besides, this is not useful when there is a left-to-right shunt, which may be the case in children with congenital heart disease. An alternative to MVO2 is central venous oxygen saturation (ScVO2) from SVC or IVC. The frequent use of central venous catheters (CVC) has enabled intermittent monitoring of ScVO2. But this may miss detecting tissue hypoxia in a child whose clinical condition can deteriorate quickly, e.g., critical congenital heart disease like hypoplastic left heart syndrome (HLHS)). CVC catheters of appropriate size for children incorporating fibreoptic technology for continuous measurement of ScVO2 are available. The mean bias compared to ScVO2 measured by co-oximetry was acceptable, but the level of agreement was – 10% to +8% and -15% to + 20% in two studies raising concern about accuracy3,4.


Near-InfraRed Spectroscopy (NIRS) oximetry offers a continuous non-invasive tissue oxygenation monitoring method. Near-infrared spectrum (700-1300 nm) can penetrate tissue, including bone, to several centimetres. NIRS monitors measure tissue oxygen saturation using a modification of Beers-Lamberts law which describes the relationship between the concentration of a light absorbing substance and the attenuation of the light intensity. Oxy and de-oxy haemoglobin have distinct absorption wavelengths in the infrared spectrum. An isosbestic wavelength is used to measure the total haemoglobin level. Most NIRS oximetry monitors use wavelengths between 700 to 850 to avoid other substances in the tissue which can absorb light. The path of the light is modelled to be elliptical based on computer predictions with a depth of penetration, roughly 1/3 of the distance between the sensors. Two sensors at different distances from the light source enable spatial resolution between the superficial and deeper tissues. Subtraction algorithms are used to calculate the difference between superficial and deeper tissue to measure oxygen saturation of deeper tissue such as the brain. Unlike pulse oximetry, NIRS oximetry measures venous weighted oxygen saturation of the tissue. Due to technical limitations, NIRS oximetry cannot provide an absolute saturation value; hence the recommendation is to use it as a trend monitor.


The study’s primary objective was to do a pooled analysis to determine the correlation between the NIRS oximetry and corresponding ScVO2. The secondary objective was to assess the impact of other variables such as age, weight, make and model of NIRS monitor etc., on the correlation.



Keywords search was done in multiple databases to identify relevant studies for inclusion in the metanalysis. The final date of the search was 1st September 2020.

The study selection criteria:


  • Children < 18 who underwent cardiac surgery or catheterisation
  • The study should report simultaneous NIRS oximetry and ScVO2 measurement
  • Explicitly state the site of measurement of NIRS oximetry and ScVO2 measurements.
  • The study should have used one of the two NIRS oximetry monitors: Somanetics INVOS or Casmed ForeSight.


Based on abstracts, studies reporting simultaneous measurements of NIRS oximetry and ScVO2 were selected for full-text review. Two authors reviewed the full text to assess if the study fits the inclusion criteria. In disagreements between the two independent reviews, a third author reviewed the discrepancies and reached a consensus.


The Studies were assessed for quality using the national heart, lung, and blood institute (NHLBI) quality assessment tool for case series studies. Data were extracted from studies that met the quality assessment by two authors independently using an electronic data sheet. Resolution of discrepancies was done similarly to study selection.



The manuscript does not state if a subject matter expert was involved in the literature search. The rationale for limiting the study population to cardiac patients and two of many FDA-approved NIRS oximetry monitors is unclear. The criterion for inclusion based on the study quality is not explicitly stated.


Statistical analysis:

The individual r-values from the studies were pooled after transformation to Fischer’s Z. Heterogeneity was assessed using the I-squared value. An I-squared value of more than 50% was indicative of significant heterogeneity. In the presence of significant heterogeneity,

a random-effects model was used. Otherwise, a fixed-effects model was used. Meta-regression was conducted to determine the impact of age, weight, clinical setting, functionally univentricular circulation, and NIRS make and model on the correlation. Due to a low number of source studies from which data could be extracted, meta-regression could only be performed on one combination of NIRS and venous oximetry site.



A total of 16 studies involving 640 patients were included in the final analysis. 13 out of 16 used Medtronic Invos oximetry, and 11 included patients undergoing cardiac surgery. The mean age and weight of the participants were 36.6 months and 11.4 kgs.

Summary of Correlations:


Cerebral near-infrared


Renal near-infrared


Superior vena caval (SVC) saturation Strong correlation

0.70 (0.55 to 0.80)

13 studies; 562 patients

No data available
Inferior vena caval (IVC)


Weak correlation

0.38 (0.15 to 0.57)

4 studies; 78 patients

Strong correlation

0.70 (0.56 to 0.81)

4 studies; 86 patients

Very strong correlation if r-value between 0.90 and 1.00; strong correlation if r-value between 0.70 and 0.89; moderate correlation if r-value between 0.50 and 0.69; weak correlation if r-value between 0.30 and 0.49; negligible correlation if r-value 0.29 or less
Data presented as r-value (95% confidence interval)


Significant heterogeneity between the studies reporting a correlation between SVC and cerebral NIRS oximetry (I2=84%) was found. Meta-regression did not show a significant effect of age, weight, univentricular heart, NIRS make and model on the correlation. A weak correlation between IVC saturation and cerebral NIRS oximetry is expected as they measure different vascular beds. A strong correlation was found between IVC saturation and renal NIRS oximetry. Meta-regression could not be performed for this pair of correlations due to the small number of patients included in the studies.



The important finding in this study is a strong correlation between SVC saturation and cerebral NIRS oximetry and between IVC saturation and Renal NIRS oximetry. The correlation between the cerebral NIRS and SVC saturation was not affected by the size of the patient, type of congenital cardiac condition or the make of the monitor. It is important to recognise that other studies have reported a significantly wide limit of agreement (+/- 20%)5. The authors hence suggest that cerebral and renal oximetry monitoring be used as trend monitors.


Also, the cerebral vascular bed is better protected by autoregulation, therefore cerebral oximetry may be better preserved in shock. Cerebral oximetry is also affected by vasodilation in response to hypercarbia. In order to increase the sensitivity of detection of low cardiac output multisite monitoring with cerebral and somatic site NIRS.


  1. Bronicki R. A. (2011). Venous oximetry and the assessment of oxygen transport balance. Pediatric critical care medicine : a journal of the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies12(4 Suppl), S21–S26.
  2. Tibby, S. M., Hatherill, M., Marsh, M. J., & Murdoch, I. A. (1997). Clinicians’ abilities to estimate cardiac index in ventilated children and infants. Archives of disease in childhood77(6), 516–518.
  3. Mohseni-Bod, H., Frndova, H., Gaitaro, R., Holtby, H., & Bohn, D. (2011). Evaluation of a new pediatric continuous oximetry catheter. Pediatric critical care medicine : a journal of the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies12(4), 437–441.
  4. odice, F. G., Ricci, Z., Haiberger, R., Favia, I., & Cogo, P. (2014). Fiberoptic monitoring of central venous oxygen saturation (PediaSat) in small children undergoing cardiac surgery: continuous is not continuous. F1000Research3, 23.
  5. Altun, D., Doğan, A., Arnaz, A., Yüksek, A., Yalçinbaş, Y. K., Türköz, R., & Sarioğlu, T. (2020). Noninvasive monitoring of central venous oxygen saturation by jugular transcutaneous near-infrared spectroscopy in pediatric patients undergoing congenital cardiac surgery. Turkish journal of medical sciences50(5), 1280–1287.