Secondary Logo

Journal Logo

Clinical Science Aspects

Influence of Arterial Dissolved Oxygen Level on Venous Oxygen Saturation

Don’t Forget the PaO2!

Legrand, Matthieu*†; Vallée, Fabrice*; Mateo, Joaquim*; Payen, Didier*†

Author Information
doi: 10.1097/SHK.0000000000000162
  • Free



One major goal of critically ill patient’s management is to ensure appropriate oxygen supply to the organs to avoid organ dysfunction (1). An adequate tissue oxygenation is obtained when the tissue oxygen consumption (VO2) is covered by sufficient oxygen delivery (DO2) associated with a normal range of oxygen (O2) extraction (2). Central venous oxygen saturation (ScvO2) measurement is considered a guide for hemodynamic optimization in circulatory failure conditions (3). Cardiac output (CO), plasma hemoglobin (Hb) concentration, arterial Hb oxygen saturation, and oxygen consumption are the most important factors influencing ScvO2. In textbooks and reviews, dissolved oxygen (i.e., unbound to Hb) is often neglected as a part of DO2 and consequently as a determinant of ScvO2. This is mostly caused by the very low oxygen solubility coefficient (i.e., 0.003). However, this statement may lead to a potential misinterpretation of ScvO2 values across FiO2 changes inducing large changes in PaO2. With this study, we aimed to study the influence of PaO2 and FiO2 on ScvO2 in ventilated critically ill patients. We hypothesized that dissolved oxygen (PadissolvO2) can significantly contribute to ScvO2 values when PaO2 increases.


The study population

This study was approved by our local ethics committee (Comité d’Évaluation de l’Éthique des projets de Recherche Biomédicale), and written informed consent was waived because of the noninterventional design of the study. Patients were eligible for inclusion if their blood arterial pressure was invasively monitored, if they were under sedation (with midazolam and fentanyl) with a Ramsay score of 5 to 6, mechanically ventilated with a PaO2/FiO2 > 200 with FiO2 ≤ 0.5, and with hemodynamic stability (i.e., no need for fluid loading or change of vasopressor infusion rate over the last 6 h). Patients had to have a central venous line and to be monitored for CO with a transoesophageal Doppler (Deltex Medical, Chicester, UK). A second cohort of brain-injured patients with jugular vein oxygen saturation (SjvO2) monitoring was also included (cohort SjvO2) to assess whether the concept holds true when applied to regional circulation (4).


As a standard of care in our unit, once daily, all patients monitored with ScvO2 had both arterial and venous blood gas analyses at the FiO2 that ensure a SaO2 greater than 95% (baseline FiO2) and PaO2 approximately 100 mmHg and at FiO2 = 1 to assess pulmonary shunting. The blood gas analysis was performed 15 min after changing the FiO2. Blood gases for analyses performed at both FiO2 were collected when patients were resting without any external stimulation. No changes in vasopressor infusion rate, fluid bolus, or other therapeutic intervention between the two measurements were required for data analysis. The following parameters were measured in cohort A: hemodynamic parameters (CO, blood arterial pressure) and blood gas analysis using a Rapidlab 1265 Series automat (Siemens, Camberley, UK) including arterial oxygen pressure (PaO2), venous oxygen pressure (PvO2), measured Hb arterial oxygen saturation (SaO2), Hb venous oxygen saturation measured on blood gas analysis (ScvO2) and Hb concentration. A chest x-ray confirmed the adequate position at entrance of the right atrium of the tip of the central venous catheter. In cohort B (SjvO2), cerebral oxygenation monitoring was performed using a retrograde internal jugular catheter. The correct placement of the catheter tip in the jugular bulb was confirmed with x-ray.

Calculation of oxygenation parameters

Formulae used for calculated oxygenation parameters are presented in Table 1. Formula 5 in Table 1 was used for predicting ScvO2 variation (Delta ScvO2) after change in FiO2 (≈0.4 vs. 1), considering oxygen consumption was stable between the two measurements. Impact of dissolved oxygen content (CadissolvO2) on ScvO2 (Delta dScvO2) was evaluated by subtracting the ScvO2 at FiO2 approximately 0.4 and 1 (Table 1). All were paired measures.

Table 1:
Calculation of oxygenation parameters*

Statistical analysis

Values are reported as median (25–75 percentiles). Paired Wilcoxon test analysis was used for intragroup comparisons for ScvO2 and SjvO2 between FiO2 at approximately 0.4 and 1. A Wilcoxon test was used for paired comparisons. Value of P < 0.05 was considered statistically significant. Statistical analysis was performed with GraphPad Prism version 4.0 for Windows (GraphPad Software, San Diego, Calif). We estimated that we needed to include 16 patients to show an increase of ScvO2 of 9% (SD, 6%) after increasing PaO2 from 100 to 500 mmHg, assuming a median Hb level of 10 g/dL and median oxygen consumption of 200 mL/min, with an α risk of 5% and a power of 90%.


Characteristics of patients

Twenty patients formed the cohort ScvO2, aged 63 (46–80) years, Simplified Acute Physiology Score (SAPS2) score 54 (42–58), 5 women and 15 men. Eleven had sepsis, five were trauma/burn patients, two had hemorrhagic shock, and two had cardiogenic shock. Twelve patients were under norepinephrine 0.5 (0.35–0.95) μg/kg per min, three received epinephrine 0.41 (0.35–0.47) μg/kg per min, and two received dobutamine 10 (10–10) μg/kg per min. Serum lactate level was 3.2 (1.8–4.3) mmol/L, platelets 166 (85–352) × 103/L, serum bilirubin 14 (9–25) μmol/L, serum creatinine 125 (72–240) μmol/L. Eight patients formed the cohort SjvO2, aged 32 (20–54) years, SAPS2 score 46 (34–57), five had a traumatic brain injury, two had stroke, and one had subarachnoid hemorrhage.

Oxygenation parameters

We observed that increasing FiO2 from 0.4 (0.3–0.5) to 1 (1–1) leads to an increase of PaO2 from 100 (85 – 124) mmHg to 387 (360–449) mmHg and a significant increase of ScvO2 from 71% (69%–76%) to 84% (78%–88%) (Fig. 1). We found similar results with SjvO2 when increasing FiO2 in patients with brain injury (Fig. 1). The rise of ScvO2 was not explained by changes in CO or Hb levels (5 [3.9–6.1] L/min vs. 5 (3.9–6.2) L/min and 11.1 (9.8–12.8) vs. 10.9 (9.8–12.7) g/dL, respectively, ns). Hemoglobin arterial oxygen saturation slightly increased from 0.97% (0.96%–0.98%) to 0.99% (0.98%–1.0%) (P < 0.001). Dissolved oxygen arterial content (CadissolvO2) rose significantly after increasing the FiO2 from 0.30 (0.26–0.37) to 1.16 (1.08–1.35) mL and from 0.32 (0.23–0.45) to 1.47 (1.20–1.72) mL in the ScvO2 group and the SjvO2 group, respectively (Table 2; Fig. 2). However, estimated participation of the dissolved oxygen component (based on formula 6 in Table 1) revealed that it largely accounted for the rise of ScvO2 (8.0% [5.9%–11.9%]) (Tables 1 and 2). Arterial carbon dioxide partial pressure (PaCO2), venous carbon dioxide partial pressure (PvCO2), arterial pH, and venous pH were not different between FiO2 approximately 0.4 and 1. Fourteen patients had a third time point available, performed at FiO2 0.5 or 0.6, allowing to plot ScvO2 values against PvO2 values and showing a sigmoidal relationship (Fig. 3).

Table 2:
Oxygenation parameters between FiO2 40% and FiO2 100% in the cohort ScvO2
Fig. 1:
Evolution of measured ScvO2 (A) and SjvO2 (B) with increase of FiO2. Data presented in box and whisper plots (min-max).
Fig. 2:
Evolution of the dissolved fraction of oxygen after increasing FiO2 from 0.4 to 1 in the ScvO2 cohort (A) and the SjvO2 cohort (B). Data presented in box and whisper plots (min-max).
Fig. 3:
Plot of PvO2 and ScvO2 across a range of FiO2 from 0.4 to 1 in 20 patients showing a sigmoidal relationship (54 measures). These data highlight the expected sharp rise in ScvO2 for modest changes in PvO2 when PaO2 increases.


Our study shows that dissolved oxygen can significantly influence the ScvO2 value. Similar results were observed using SjvO2 in brain-injured patients, suggesting that it can also be applied at a regional circulation level. Our study highlights a well-known basic physiological principle, which has been overlooked in most reviews and textbooks that may lead to misinterpretation of ScvO2. These misinterpretations of the determinant of ScvO2 might arise from the oversimplification of the determinants of ScvO2 with PaO2 omission (leaving CO, SaO2, and Hb level as sole determinants). We therefore believe that the results of our study can reach an important educational target in providing evidence for careful reading of the determinant of ScvO2: PaO2 should not be neglected when interpreting the ScvO2 value.

The ScvO2 has been widely included in goal-directed therapy algorithms both in critically ill patients and patients undergoing major surgery. These algorithms propose to optimize systemic hemodynamics to reach ScvO2 greater than 70% and improve oxygen delivery (5). Because most textbooks and reviews proposed to neglect dissolved oxygen as part of O2 content, PaO2 is not included in most formulae showing the main determinants of ScvO2 {i.e., ScvO2 = SaO2 − [VO2 /(CO × Hb × 1.34)]} (2,6). Likewise, most observational studies did not mention the PaO2 level in their results. For example, an observational study from Hernandez et al. showed a ScvO2 increase from 61.8% ± 12.6% to 68.9% ± 12.2% when in critically ill patients were intubated and mechanically ventilated (7). The authors did not find any correlation between increase of ScvO2 and changes in SaO2, although blood gas were obtained 15 min after mechanical ventilation with FiO2 = 1, a situation where high PaO2 values might have accounted for the rise of ScvO2. Because PaO2 was not reported, it became difficult to conclude on this point. This is true for many studies testing and measuring ScvO2 as an algorithm target (1, 8–10).

The range of ScvO2 changes associated with the increase in FiO2 (∼9%–10%) can be seen modest when compared with those associated with CO or SaO2 level. Indeed, because of the sigmoid shape of the oxygen-Hb dissociation curve, modest changes in PvO2 can induce significant changes in ScvO2. However, this range of ScvO2 variation appears relevant in the light of current recommendation for resuscitation of patients with severe sepsis (i.e., to reach ScvO2 >70%) (5). Furthermore, it is reasonable to expect a more pronounced effect of dissolved O2 content when CO and Hb level are low (11).

The aim of the study was not to evaluate the potential benefit of increasing PaO2 for improving tissue oxygenation but to appreciate the influence of variations in PaO2 on ScvO2 to raise awareness of intensivists while using the ScvO2 as a hemodynamic target. However, different studies have used high inspiratory fraction of O2 and dissolved O2 as factors to improve tissue oxygen tension, illustrating that the amount of dissolved O2 content may participate in tissue oxygenation. Yu et al. (12) observed a significant increase of tissue PO2 (PtO2) in critically ill patients after an oxygen challenge test (i.e., increase FiO2 to 100%), illustrating the better oxygen delivery to tissue after increasing FiO2. Likewise, Dyson et al. (13) observed a significant rise of PtO2 in animal breathing at FiO2 = 1 compared with room air. Although this study was not designed to assess the role of high FiO2 to increase DO2, this question merits to be prospectively addressed.

Our study has several limitations. The cohort size is small. The sizing was, however, fitting well with the sample size calculation, and the results are sustained by a strong physiological basis. It is unlikely that increasing the sample size would modify the results. Second, VO2 was not measured. However, our patients were considered stable, nonstimulated, and under sedation, making a significant change in VO2 between measurements unlikely.

Our study has direct clinical implications. In addition to CO and Hb levels as key determinant of ScvO2 or organ SvO2, our study shows that PaO2, when varying significantly, is also to be taken into account when interpreting changes in ScvO2. It can therefore be recommended to assess the impact of each therapeutic intervention (e.g. fluid loading, inotropes or vasopressors infusion, blood transfusion) at a constant FiO2 and avoid significant variations in PaO2.

In conclusion, our study shows that increasing dissolved oxygen content can notably influence ScvO2 and SjvO2 values in intensive care unit patients.


1. Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, Peterson E, Tomlanovich MEarly Goal-Directed Therapy Collaborative Group: Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 345 (19): 1368–1377, 2001.
2. Walley KR: Use of central venous oxygen saturation to guide therapy. Am J Respir Crit Care Med 184 (5): 514–520, 2011.
3. Reinhart K, Rudolph T, Bredle DL, Hannemann L, Cain SM: Comparison of central-venous to mixed-venous oxygen saturation during changes in oxygen supply/demand. Chest 95 (6): 1216–1221, 1989.
4. Clavier N, Schurando P, Raggueneau JL, Payen DM: Continuous jugular bulb venous oxygen saturation validation and variations during intracranial aneurysm surgery. J Crit Care 12 (3): 112–119, 1997.
5. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, Sevransky JE, Sprung CL, Douglas IS, Jaeschke R, et al.: Surviving Sepsis Campaign Guidelines Committee including The Pediatric Subgroup. Intensive Care Med 39 (2): 165–228, 2013.
6. Teboul J-L, Hamzaoui O, Monnet X: SvO2 to monitor resuscitation of septic patients: let’s just understand the basic physiology. Crit Care 15 (6): 1005, 2011.
7. Hernandez G, Peña H, Cornejo R, Rovegno M, Retamal J, Navarro JL, Aranguiz I, Castro R, Bruhn A: Impact of emergency intubation on central venous oxygen saturation in critically ill patients: a multicenter observational study. Crit Care 13 (30): R63, 2009.
8. Trzeciak S, Dellinger RP, Parrillo JE, Guglielmi M, Bajaj J, Abate NL, Arnold RC, Colilla S, Zanotti S, Hollenberg SMMicrocirculatory Alterations in Resuscitation and Shock Investigators: Early microcirculatory perfusion derangements in patients with severe sepsis and septic shock: relationship to hemodynamics, oxygen transport, and survival. Ann Emerg Med 49 (1): 88–98, 2007.
9. Bracht H, Hänggi M, Jeker B, Wegmüller N, Porta F, Tüller D, Takala J, Jakob SM: Incidence of low central venous oxygen saturation during unplanned admissions in a multidisciplinary intensive care unit: an observational study. Crit Care 11 (1): R2, 2007.
10. Jones AE, Shapiro NI, Trzeciak S, Arnold RC, Claremont HA, Kline JAEmergency Medicine Shock Research Network (EMShockNet) Investigators: Lactate clearance vs central venous oxygen saturation as goals of early sepsis therapy: a randomized clinical trial. JAMA 303 (8): 739–746, 2010.
11. He H-W, Liu D-W, Long Y, Wang X-T, Chai W-Z, Zhou X: The transcutaneous oxygen challenge test: a noninvasive method for detecting low cardiac output in septic patients. Shock 37 (2): 152–155, 2012.
12. Yu M, Morita SY, Daniel SR, Chapital A, Waxman K, Severino R: Transcutaneous pressure of oxygen: a noninvasive and early detector of peripheral shock and outcome. Shock 26 (5): 450–456, 2006.
13. Dyson A, Stidwill R, Taylor V, Singer M: The impact of inspired oxygen concentration on tissue oxygenation during progressive hemorrhage. Intensive Care Med 35 (10): 1783–1791, 2009.

Oxygen saturation; oxygen pressure; hemoglobin; cardiac output; shock; resuscitation

© 2014 by the Shock Society