Critical illness is characterized by inadequate tissue oxygen delivery leading to organ ischemia.1 Central venous oxygen saturation (ScvO2) and serum lactate are indicators of tissue hypoxia. However, they are invasive. Tissue oxygen saturation (StO2) measured using noninvasive near-infrared spectroscopy (NIRS) reflects tissue oxygenation and correlates with ScvO2 and mixed venous oxygen saturation.2,3 StO2 values have been found to be higher in survivors of septic shock compared with nonsurvivors.4 Regional cerebral oxygen saturation (rSO2) is a surrogate of cerebral blood flow (CBF), and has been correlated with serum lactate, ScvO2, and mean arterial pressure (MAP).5 However, there is a paucity of literature on the correlation between StO2 and rSO2 and prognostic value of StO2 and rSO2. The primary objective of this preliminary study was to correlate StO2 and rSO2 measurements, and the secondary objective to study their prognostic value in neurological patients with sepsis. We also investigated the relationship between StO2 and rSO2 with other global hemodynamic and metabolic parameters.
MATERIALS AND METHODS
This preliminary prospective, observational study was undertaken in a tertiary care neuroscience center after institutional ethics committee approval and informed consent from the patient or their relative. The study was registered with the Clinical Trials Registry—India (Registration number CTRI/2018/03/012826). Adult neurosurgical and neuromedical patients (aged 21 to 80 y) with a clinical diagnosis of sepsis or septic shock were recruited into the study. Exclusion criteria were hemodynamic instability on admission to the intensive care unit, established diagnosis of sepsis on intensive care unit admission, peripheral vascular disease, frontal lobe pathology, presence of intracranial aneurysm or arteriovenous malformation with bleed, cerebral infarction, cortical venous thrombosis, and brain death. Patients who became brain dead or died within 48 hours of study onset were considered dropouts.
Tissue oxygenation was measured using an EQUANOX oximeter (model 7600; NONIN Medicals, Plymouth, MA) with EQUANOX Advance Sensor (model 8004; CA). The readings were taken at predefined timepoints (described below). An average of 3 readings (taken at 5 min intervals) of StO2 and rSO2, measured from the thenar eminence of both hands and the forehead respectively, was recorded.
Demographic details were recorded at the time of diagnosis of sepsis/septic shock. Hemodynamic parameters (heart rate [HR], MAP, central venous pressure [CVP]), parameters of oxygenation (StO2, rSO2, ScvO2), serum lactate and base deficit, and illness severity scores (Acute Physiology and Chronic Health Evaluation [APACHE II] score, Sequential Organ Failure Assessment [SOFA] score, Glasgow Coma Scale [GCS]) were recorded at 0, 6, 12, 24, 36, and 48 hours after sepsis onset and, thereafter, once daily until sepsis resolved.
Sepsis and septic shock were defined according to the SEPSIS 3.0 guidelines.6 All patients were managed according to the Surviving Sepsis guidelines, with target endpoints of HR≤100/min, systolic blood pressure ≥90 mm Hg or MAP≥65 mm Hg, CVP of 8 to 12 mm Hg, urine output ≥0.5 mL/kg/h and ScvO2≥70%.7
Outcomes were sepsis-related in-hospital mortality and the Glasgow outcome score (GOS) at discharge from hospital. GOS 1 to 2 (1—good recovery, 2—moderate disability) was considered as a favorable outcome, and GOS 3 to 5 (3—severe disability, 4—vegetative state, 5—death) as unfavorable.
Analysis of qualitative data and quantitative data was undertaken using the χ2 and the student t test, respectively. A P≤0.05 was considered significant. A Pearson correlation coefficient was used to correlate oxygenation and hemodynamic variables. A repeated measures analysis of variance was used to compare the different parameters between survivors and nonsurvivors. The area under the receiver operating characteristic curves (AUC-ROC) was used to measure the discriminative ability of StO2 to distinguish survivors from nonsurvivors. We also divided the patients into those with sepsis and septic shock and used ROC curves to assess the predictive ability of rSO2. Friedman and Mann-Whitney U tests were used to compare rSO2 between survivors and nonsurvivors of septic shock, within the groups and between the groups, respectively. SPSS version 17 (SPSS Inc., Chicago, IL) was used for statistical analysis. As this was a preliminary study we undertook a post hoc power analysis. For a correlation value of 0.599 between StO2 and rSO2, the sample size required was 21 patients for a power of 0.9. With the available data of StO2 in survivors and nonsurvivors (68%±6% and 63%±7%, respectively), the power of the study was 0.858 (using G power software version 184.108.40.206).
Eighty patients were enrolled of which 23 were excluded; of the 57 eligible patients, 12 were dropouts (Fig. 1). Hence, 45 patients were analyzed and followed up for an average of 3.8 days for oxygenation parameters (StO2 and rSO2) until the resolution of sepsis and for 28 days for GOS and in-hospital mortality.
The demographics of the study cohort are shown in the supplementary material (Table 1, Supplemental Digital Content 1, http://links.lww.com/JNA/A84). Of the 29 patients who survived, 14 had favorable outcome, and 15 unfavorable outcome (7 had GOS 3, 8 had GOS 4) at hospital discharge.
Baseline hemodynamic parameters were comparable between survivors and nonsurvivors (Table 1). StO2 at baseline was significantly lower in nonsurvivors compared with survivors (68%±6% vs. 63%±7%, respectively; P=0.03). Median APACHE II score, mean SOFA score, mean lactate, and median GCS were also worse in nonsurvivors compared with survivors. rSO2 at baseline was not significantly different between survivors and nonsurvivors.
StO2 and rSO2
There was a moderately positive correlation between StO2 and rSO2 at baseline (r=0.599; P=0.001) (Fig. 2).
StO2 and rSO2 in survivors and nonsurvivors are shown in Figure 3, and serum lactate, APACHE II, and GCS in the supplementary material (Figs. 1–3, Supplemental Digital Content 2, http://links.lww.com/JNA/A85). There was a significant difference between survivors and nonsurvivors with respect to all variables except rSO2. The comparison of these same parameters in patients with favorable and unfavorable outcomes is shown in the supplementary material (Figs. 4–8, Supplemental Digital Content 3, http://links.lww.com/JNA/A86). There was a significant difference between patients with favorable and unfavorable outcomes in all these parameters, including rSO2.
ROC analysis (Fig. 4A) showed that StO2 differentiated between survivors and nonsurvivors, with highest accuracy being at 48 hours (AUC, 0.786; 95% confidence interval [CI], 0.628-0.943; P=0.003). StO2 of 68.5% had 77% sensitivity and 71% specificity for differentiation between survivors and nonsurvivors.
Sepsis Versus Septic Shock
There was no difference in rSO2 values between survivors and nonsurvivors in patients with sepsis. However, in patients with septic shock rSO2 increased in survivors (P=0.003) and decreased in nonsurvivors (P=0.004) over 48 hours. Also, between survivors and nonsurvivors of septic shock, there was a significant difference in rSO2 (P=0.032) (Table 2, Supplemental Digital Content 1, http://links.lww.com/JNA/A84). To discriminate survivors and nonsurvivors of septic shock, a ROC curve for rSO2 was constructed at 48 hours (Fig. 4B) and showed an AUC of 0.734 (95% CI, 0.526-0.942; P=0.033). An rSO2 value of 62.5% had a sensitivity of 83% and specificity of 67% to differentiate survivors and nonsurvivors (odds ratio, 0.894; 95% CI, 0.809-0.989; P=0.03).
Survivors and Nonsurvivors in Neurosurgical and Neuromedical Patients
In the neurosurgical population (n=30), survivors had a higher StO2 (P=0.027) and lower serum lactate (P=0.005) compared with nonsurvivors. However, there was a nonsignificant difference in rSO2 (P=0.075). We did not find any differences in these variables between survivors and nonsurvivors in the neuromedical cohort.
Correlation of StO2 and rSO2 With Other Parameters of Sepsis
The correlations between StO2 and rSO2 with other parameters of oxygenation and perfusion, that is, ScvO2, serum lactate, base deficit and MAP, are shown in Table 3 (Supplemental Digital Content 1, http://links.lww.com/JNA/A84). Although both StO2 and rSO2 had poor correlation with ScvO2, the correlation coefficient was slightly higher with StO2. Similarly, StO2 had a higher correlation with serum lactate than rSO2. There was an inverse correlation between oxygenation parameters and serum lactate. StO2, but not rSO2, correlated with base deficit and MAP.
Standard assessments of the adequacy of tissue oxygenation, such as ScvO2, serum lactate and base deficit, have potential limitations. NIRS-derived StO2 recorded from the thenar eminence is able to provide information on tissue oxygen supply and demand noninvasively and on a continuous basis. The advantage of choosing the thenar eminence for monitoring StO2 is that it has a relatively thin fat layer, is least affected by systemic edema formation, and it is almost always nonpigmented.8 Mulier et al3 observed that patients with severe sepsis had significantly lower thenar muscle StO2 values than healthy volunteers. Moreover, StO2 values have been shown to predict survival in patients with septic shock,4 and organ dysfunction and mortality in patients with trauma and critical illness.9–11 In our study, StO2 values were higher in survivors.
Cerebral perfusion is likely to suffer during septic shock, and rSO2 monitoring can be considered a surrogate measure of CBF. Decrease in rSO2 associated with systemic oxygen desaturation (reflected by reduced StO2) may be an early indicator of cerebral hypoperfusion. In a meta-analysis by Zorrilla-Vaca et al,12 intraoperative cerebral oximetry-guided management reduced the incidence of postoperative cognitive dysfunction at 1 week.
Individualized autoregulation curves can be constructed on a daily basis by plotting rSO2 against spontaneous fluctuations in MAP,13 to determine optimal MAP and cerebral perfusion pressure in an individual patient. In a study, patients in whom median cerebral perfusion pressure or MAP differed significantly from their respective “optimal” values were more likely to have unfavorable outcomes.14 rSO2 has also been shown to prognosticate outcome after return of spontaneous circulation following cardiac arrest.15
Litchtenstern et al16 correlated ScvO2 and rSO2 in patients with severe sepsis and concluded that ScvO2<70% was indicated by rSO2<56.5%, with a sensitivity and specificity of 75% and 100%, respectively. rSO2 has a negative correlation with lactic acid and positive correlation with ScvO2 and MAP in patients with shock.5 However, in our study, the correlation between rSO2 and ScvO2 was negligible (0.177; P=0.004), possibly because we included all patients with sepsis, not only those with septic shock.
We compared the StO2 and rSO2 values in survivors and nonsurvivors at baseline and over a period of time to determine their prognostic value in neurological patients with sepsis, and found that StO2, but not rSO2, was different between the 2 groups. There are several possible explanations for not identifying reduced cerebral oxygenation using NIRS in the face of systemic sepsis. Autoregulation maintains CBF in the face of hypotension, and, in milder forms of sepsis, decreased systemic perfusion may not be accompanied by cerebral oxygen desaturation; we found a decrease in rSO2 in nonsurvivors (at 48 h) only in patients with septic shock. Further, NIRS sensors placed on the forehead detects cerebral oxygenation up to a depth of 2 to 3 cm; any changes occurring at greater depth may not be detected by the monitor.
In a subgroup analysis, we found only a (nonsignificant) trend to a difference in rSO2 between survivors and nonsurvivors in neurosurgical patients. In contrast, Tayar and colleagues identified a significant difference in rSO2 between survivors and nonsurvivors, but that significance was only after 72 hour.4 A study in patients with septic shock found that those who died had more significant decreases in rSO2 than those who survived.17 We also found that cerebral oxygenation was reduced in nonsurvivors of septic shock. In our study, StO2>68.5% at 48 hour after sepsis onset had a good discriminative power to differentiate survivors from nonsurvivors. Also, an rSO2 value of 62.5% had a sensitivity of 83% and specificity of 67% to differentiate survivors and nonsurvivors of septic shock at 48 hours.
The strength of the current study is that we measured multiple parameters of sepsis and identified a correlation of these with StO2 and rSO2. However, our study also has some limitations. It was conducted in a single center and the study population was heterogenous. StO2 and rSO2 values were recorded at predefined timepoints and not continuously. Although we analyzed the data only during the first 48 hours of sepsis, the trends clearly indicate the potential for prognostication of the outcome of sepsis. Finally, this was a small, preliminary observational study, and a larger trial designed to investigate the effects of particular interventions aimed at improving tissue oxygenation is required.
In conclusion, we found a significant correlation between StO2 and rSO2 in neurological patients with sepsis. StO2 was able to offer prognostic information about survival and favorable outcomes in this patient cohort. The role of rSO2 in prediction of survival in milder form of sepsis is doubtful as only septic shock resulted in significant cerebral oxygen desaturation.
1. Shoemaker WC, Appel PL, Kram HB. Tissue oxygen debt as determinant of lethal and nonlethal postoperative organ failure. Crit Care Med. 1988;16:117–120.
2. Mesquida J, Masip J, Gili G, et al. Thenar oxygen saturation measured by near infrared spectroscopy as a noninvasive predictor of low central venous oxygen saturation in septic patients. Intensive Care Med. 2009;35:1106–1109.
3. Mulier KE, Skarda DE, Taylor JH, et al. Near-infrared spectroscopy
in patients with severe sepsis
: correlation with invasive hemodynamic measurements. Surg Infect. 2008;9:515–519.
4. Leone M, Blidi S, Antonini F, et al. Oxygen tissue saturation is lower in nonsurvivors than in survivors after early resuscitation of septic shock
. Anesthesiology. 2009;111:366–371.
5. Al Tayar A, Abouelela A, Mohiuddeen K. Can the cerebral regional oxygen saturation be a perfusion parameter in shock? J Crit Care. 2017;38:164–167.
6. Shankar Hari M, Phillips GS, Levy ML, et al. Sepsis
Definitions Task Force: developing a new definition and assessing new clinical criteria for septic shock
: for the Third International Consensus Definitions for Sepsis
and Septic Shock
-3). JAMA. 2016;315:775–787.
7. Dellinger RP, Levy MM, Rhodes A, et al. Surviving Sepsis
Campaign: International guidelines for management of severe sepsis
and septic shock
. Intensive Care Med. 2013;39:165–228.
8. Scheeren TWL, Schober P, Schwarte LA. Monitoring tissue oxygenation by near infrared spectroscopy (NIRS): background and current applications. J Clin Monit Comput. 2012;26:279–287.
9. Lima A, van Bommel J, Jansen TC, et al. Low tissue oxygen saturation
at the end of early goal-directed therapy is associated with worse outcome in critically ill patients. Crit Care. 2009;13(suppl 5):S13.
10. Cohn SM, Nathens AB, Moore FA, et al. Tissue oxygen saturation
predicts the development of organ dysfunction during traumatic shock resuscitation. J Trauma. 2007;62:44–54.
11. Veening A, Ince C, Bakker J. Incidence of low tissue oxygen saturation
in a mixed population of critically ill patients. Crit Care. 2010;14(suppl 1):P151.
12. Zorrilla-Vaca A, Healy R, Grant MC, et al. Intraoperative cerebral oximetry-based management for optimizing perioperative outcomes: a meta-analysis of randomized controlled trials. Can J Anaesth. 2018;65:529–542.
13. Rivera-Lara L, Zorrilla-Vaca A, Healy RJ, et al. Determining the upper and lower limits of cerebral autoregulation with cerebral oximetry autoregulation curves: a case series. Crit Care Med. 2018;46:e473–e477.
14. Rivera-Lara L, Zorrilla-Vaca A, Geocadin RG, et al. Cerebral autoregulation-oriented therapy at the bedside: a comprehensive review. Anesthesiology. 2017;126:1187–1199.
15. Storm C, Leithner C, Krannich A, et al. Regional cerebral oxygen saturation
after cardiac arrest in 60 patients—a prospective outcome study. Resuscitation. 2014;85:1037–1041.
16. Litchtenstern C, Koch C, Rohrig R, et al. Near-infrared spectroscopy
therapy : predictor of a low central venous oxygen saturation. Anaesthesist. 2012;61:883–891.
17. Funk DJ, Kumar A, Klar G. Decreases in cerebral saturation in patients with septic shock
are associated with increased risk of death: a prospective observational single center study. J Intensive Care. 2016;4:42–49.