Journal of Critical Care xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Journal of Critical Care journal homepage: www.jccjournal.org
Noninvasive assessment of hemodynamic variables using near-infrared spectroscopy in patients experiencing cardiogenic shock and individuals undergoing venoarterial extracorporeal membrane oxygenation☆ Petr Ostadal, MD, PhD ⁎, Andreas Kruger, MD, Dagmar Vondrakova, MD, Marek Janotka, MD, Hana Psotova, MD, Petr Neuzil, MD, PhD Cardiovascular Center, Na Homolce Hospital, Prague, Czech Republic
a r t i c l e
i n f o
Keywords: Cardiogenic shock Extracorporeal membrane oxygenation Hemodynamics Near-infrared spectroscopy Oximetry
a b s t r a c t Purpose: The relationship between near-infrared spectroscopy cerebral oximetry (CrSO2), peripheral oximetry (PrSO2) and hemodynamic variables is not fully understood. Methods: The relationship between CrSO2/PrSO2 and cardiac index (CI), systemic vascular resistance index (SVRI) and mean arterial pressure (MAP) in patients experiencing cardiogenic shock and those undergoing venoarterial extracorporeal membrane oxygenation (ECMO) was retrospectively analyzed; in patients on ECMO, total circulatory index (TCI) was calculated from the sum of CI and extracorporeal blood flow index. Results: In patients experiencing cardiogenic shock (n = 10), significant correlations between PrSO2 values and CI (Spearman r = 0.81; P b .0001), SVRI (r = −0.45; P b .0001), and MAP (r = 0.58; P b .0001) were found. Significant correlations between CrSO2 and CI (r = 0.55; P b .0001) and SVRI (r = −0.47; P b .0001), but not MAP, were observed. Linear regression analysis revealed that CI could be calculated using the following equation: CI = PrSO2/24.0. In patients on VA ECMO (n = 12), significant correlations were found between PrSO2 and TCI (r = 0.68; P b .0001), SVRI (r = − 0.47; P b .0001), and MAP (r = 0.27; P = .025). Significant correlations were also found between CrSO2 and TCI (r = 0.68; P b .0001) and SVRI (r = −0.51; P b .0001), but not MAP. Conclusions: Results of the present study suggest that CrSO2 and PrSO2 in particular can be used for noninvasive estimation and monitoring of global circulatory status in patients experiencing cardiogenic shock and individuals undergoing ECMO. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Measurement of hemodynamic variables is often valuable in many critical conditions, especially in patients experiencing cardiogenic shock. However, current methods for cardiac output (CO) measurement are often invasive and limited in their ability to continuously monitor; furthermore, they are not applicable in some cases such as patients undergoing venoarterial extracorporeal membrane oxygenation (VA ECMO). The criterion standard for CO measurement in intensive care—thermodilution using pulmonary artery catheterization (PAC)—does not allow proper measurement of global hemodynamic status in patients on VA ECMO because (i) a substantial portion of blood flow bypasses the lungs and (ii) thermodilution with PAC often losses accuracy in these patients when pulmonary artery flow falls
☆ The present study was supported by a grant from the Czech Ministry of Health, No. 12153. ⁎ Corresponding author. Department of Cardiology, Cardiovascular Center, Na Homolce Hospital, Roentgenova 2, 15000 Prague, Czech Republic. Tel.: +420 257272208; fax: +420 257272974. E-mail address: [email protected] (P. Ostadal).
below 2 L/min. Frequently used minimally invasive or noninvasive methods such as pulse-wave analysis are not suitable in substantial part of these individuals, for example, because of continual blood flow. Furthermore, transpulmonary thermodilution is inappropriate for CO measurements in patients with extracorporeal circulation. Similarly, although echocardiography provides important data regarding cardiac functions, it does not enable continuous monitoring and assessment of global circulatory status in patients undergoing VA ECMO. Near-infrared spectroscopy (NIRS) oximetry has recently emerged as a promising method for the noninvasive measurement of cerebral and tissue oxygenation and perfusion. Several studies have demonstrated its use in the assessment of perioperative brain perfusion in cardiac [1] and noncardiac [2] surgery, in carotid interventions [3], and for monitoring brain and tissue perfusion in several critical conditions [4–7]. To date, however, there are only limited data regarding measurement of cerebral regional oxygen saturation (CrSO2) and peripheral regional oxygen saturation (PrSO2) using NIRS in patients experiencing cardiogenic shock and those undergoing VA ECMO [8,9]. In the present study, we tested the hypothesis that NIRS oximetry can be used to estimate global circulatory status in patients experiencing cardiogenic shock and in patients with cardiogenic shock undergoing
0883-9441/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcrc.2014.02.003
Please cite this article as: Ostadal P, et al, Noninvasive assessment of hemodynamic variables using near-infrared spectroscopy in patients experiencing cardiogenic shock and in..., J Crit Care (2014), http://dx.doi.org/10.1016/j.jcrc.2014.02.003
2
P. Ostadal et al. / Journal of Critical Care xxx (2014) xxx–xxx
VA ECMO, in whom standard methods of hemodynamic measurement cannot be easily used. 2. Materials and methods 2.1. Study overview The present study was approved by the Na Homolce Hospital Ethics Committee (Prague, Czech Republic). Written informed consent for data collection from patient’s medical records was not required for this retrospective study, and all data were collected and analyzed in a de-identified form. We performed a retrospective analysis of medical records from patients experiencing cardiogenic shock hospitalized at the Cardiology Intensive Care Unit, Na Homolce Hospital in Prague, in whom both CrSO2 and PrSO2 4-channel NIRS oximetry (INVOS, Covidien, Dublin, Ireland) was monitored simultaneously with hemodynamic monitoring using PAC (CCOmbo Pulmonary Artery Catheter and Vigilance II Monitor; Edwards Lifesciences, Irvine, California). The indication for PAC was insufficient response to inotropic and vasopressor therapy (defined as norepinephrine dose N0.2 μg kg −1 min −1 and dobutamin dose N5 μg kg −1 min −1). Since 2010, CrSO2 and PrSO2 NIRS oximetry monitoring has been routinely used at the Na Homolce Hospital in most patients experiencing cardiogenic shock and in almost all patients undergoing VA ECMO. Cerebral oximetry has been performed using two forehead sensors, and peripheral oximetry using 2 sensors placed at the calves. Cerebral oximetry (N 50%) became also one of the major therapeutic goals in these patients, besides mixed venous oxygen saturation and mean arterial pressure (MAP). The medical records of these patients contained oximetry and hemodynamic data collected at 1-hour intervals. In the present analysis, we included values obtained at the beginning of measurement and subsequently at 6-hour intervals for a maximum of 2 days (ie, a maximum of 9 data sets per patient). The reason for selection only data recorded at 6-hour intervals for the analysis was to reflect more different global circulatory status. In patients experiencing cardiogenic shock, values of CO, MAP, central venous pressure, and body surface area were obtained from the medical records; cardiac index (CI) and systemic vascular resistance index (SVRI) were then calculated. In patients undergoing VA ECMO, total circulatory index (TCI) was calculated according to the following formula: TCI ¼ CI þ extracorporeal blood flow index ðEBFIÞ; EBFI ¼ extracorporeal blood flow=BSA In VA ECMO patients, SVRI was then calculated from TCI instead of CI:
recently underwent simultaneous monitoring of NIRS oximetry and hemodynamics were identified. Up to May 2013, 69 nonsurgical patients who experienced cardiogenic shock with persisting or progressing tissue hypoperfusion despite standard therapeutic approaches or refractory cardiac arrest underwent successful VA ECMO (Levitronix Centrimag pump (Thoratec Corporation, Pleasanton, California) using the Quadrox oxygenator (Maquet Cardiopulmonary AG, Hirrlingen, Germany) or Cardiohelp system (Maquet Cardiopulmonary AG, Hirrlingen, Germany) at the Na Homolce Hospital. In 2010, the hospital began using cerebral and peripheral oximetry in these patients, and to date, 44 patients on VA ECMO have been monitored using this noninvasive method. From this group, 12 individuals who underwent simultaneous monitoring of NIRS oximetry and hemodynamics using PAC were identified. In all of these individuals, PAC was placed already before ECMO introduction. In patients with limb ischemia caused by arterial outflow cannula (oximetry b 40% with absolute difference N10% with contralateral limb), distal perfusion bypass was introduced using 5F or 6F arterial sheath. 2.3. Statistical analysis A mean value was calculated for each pair of NIRS oximetry data (left and right brain hemisphere or left and right lower limb); this value was used for further analysis. Spearman correlation and linear regression were performed to analyze the relationship between oximetry and hemodynamic values using GraphPad Prism 6 software (GraphPad Software, Inc, San Diego, California). Stepwise multiple regression analyses of PrSO2 and CrSO2 against the set of hemodynamic parameters were conducted using the MedCalc 12 software (MedCalc Software bvba, Ostend, Belgium). P b .05 was considered statistically significant. 3. Results The baseline characteristics of individual patients are shown in Table 1. In patients experiencing cardiogenic shock, a strong, highly significant correlation was observed between PrSO2 and CI (r = 0.81; 95% confidence interval [95% CI], 0.71-0.87; P b .0001; Fig. 1A), and a weaker yet still highly significant correlation was found between
Table 1 Baseline characteristics of individual patients
Cardiogenic shock
SVRI ¼ ðMAP–CVPÞ 80=TCI In patients on ventilation support, ventilator parameters were set to maintain a target partial pressure of oxygen in arterial blood (paO2) of 10.0 to 13.0 kPa and partial pressure of carbon dioxide (paCO2) of 4.0 to 6.5 kPa. Similarly, in patients undergoing VA ECMO, oxygenator parameters were set to maintain the same target values in the arterial (ie, outflow) line. Blood pressure was monitored invasively from the radial or femoral artery in all patients. 2.2. Study population Most patients with cardiogenic shock experienced acute myocardial infarction or decompensated chronic heart failure. All were treated with inotropes and vasopressors, and the majority (80%) with an intra-aortic balloon pump. All patients who experienced acute myocardial infarction underwent primary percutaneous coronary intervention. For the analysis, 10 consecutive individuals who
VA ECMO
Patient
Cause of circulatory failure
Age (y)
Sex
Alive
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 11 12
AMI AMI, MR DCMP DCMP AMI ICMP AMI AMI, MR DCMP AMI CA, AMI, MR Myocarditis AMI AMI DCMP DCMP DCMP AMI AMI CA, AMI AMI CA, AMI
53 52 68 60 72 66 67 58 69 71 65 66 66 67 70 67 61 63 65 24 58 54
M M F M M M F M F M M F M F M M M M M M M M
No Yes Yes Yes No Yes No Yes Yes Yes Yes No Yes No Yes Yes Yes Yes No No Yes Yes
AMI indicates acute myocardial infarction; CA, cardiac arrest; DCMP, dilated cardiomyopathy; F, female; ICMP, ischemic cardiomyopathy; M, male; MR, severe mitral regurgitation. Patients marked “Alive: Yes” were either discharged from hospital or alive at 30 days.
Please cite this article as: Ostadal P, et al, Noninvasive assessment of hemodynamic variables using near-infrared spectroscopy in patients experiencing cardiogenic shock and in..., J Crit Care (2014), http://dx.doi.org/10.1016/j.jcrc.2014.02.003
P. Ostadal et al. / Journal of Critical Care xxx (2014) xxx–xxx
3
Fig. 1. Correlation between NIRS oximetry and hemodynamic variables in patients experiencing cardiogenic shock. Dependent values from individual patients are shown in different colors. A, Correlation of peripheral oximetry and CI. B, Correlation of peripheral oximetry and SVRI. C, Correlation of peripheral oximetry and MAP. D, Correlation of cerebral oximetry and CI. E, Correlation of cerebral oximetry and systemic vascular resistance. F, Correlation of cerebral oximetry and MAP.
PrSO2 and SVRI (r = − 0.45; 95% CI, − 0.62 to − 0.23; P b .0001; Fig. 1B) and between PrSO2 and MAP (r = 0.58; 95% CI, 0.38-0.71; P b .0001; Fig. 1C). Moreover, a significant correlation between CrSO2 and CI (r = 0.55; 95% CI, 0.37-0.70; P b .0001; Fig. 1D) and between CrSO2 and SVRI (r = − 0.47; 95% CI, − 0.63 to − 0.27; P b .0001; Fig. 1E) was found, but not between CrSO2 and MAP (r = 0.12; 95% CI, − 0.12 to 0.34; P b .3031; Fig. 1F). An equation relating PrSO2 and CI was determined using linear regression: CI ¼ PrSO2 =24:0ðslope 95%CI 23:0 to 25:0; Pb:0001Þ: Stepwise multiple regression analysis revealed that the PrSO2 values could be predicted by CI (multiple correlation coefficient r =
0.60, P b .0001), MAP (r = 0.43, P = .0002), and SVRI (r = − 0.24, P = .0454), whereas the CrSO2 values were predicted by MAP (r = 0.45, P = .0001) and SVRI (r = − 0.42, P = .0002). In patients undergoing VA ECMO, a significant correlation between PrSO2 and TCI (r = 0.68; 95% CI, 0.53-0.79; P b .0001; Fig. 2A) and between PrSO2 and SVRI (r = − 0.47; 95% CI, − 0.63 to − 0.26; P b .0001; Fig. 2B) was found, and a weak yet still significant correlation between PrSO2 and MAP (r = 0.27; 95% CI, 0.03-0.47; P = .025; Fig. 2C) was found. Furthermore, a significant correlation between CrSO2 and TCI (r = 0.68; 95% CI, 0.52-0.79; P b .0001; Fig. 2D) and between CrSO2 and SVRI (r = − 0.51; 95% CI, − 0.67 to − 0.31; P b .0001; Fig. 2E) was found, but not between CrSO2 and MAP (r = 0.09; 95% CI, − 0.15 to 0.32; P = .46; Fig. 2F). An
Please cite this article as: Ostadal P, et al, Noninvasive assessment of hemodynamic variables using near-infrared spectroscopy in patients experiencing cardiogenic shock and in..., J Crit Care (2014), http://dx.doi.org/10.1016/j.jcrc.2014.02.003
4
P. Ostadal et al. / Journal of Critical Care xxx (2014) xxx–xxx
Fig. 2. Correlation between NIRS oximetry and hemodynamic variables in patients undergoing ECMO. Dependent values from individual patients are shown in different colors. A, Correlation of peripheral oximetry and TCI. B, Correlation of peripheral oximetry and SVRI. C, Correlation of peripheral oximetry and MAP. D, Correlation of cerebral oximetry and TCI. E, Correlation of cerebral oximetry and systemic vascular resistance. F, Correlation of cerebral oximetry and MAP.
equation relating PrSO 2 and TCI was determined using linear regression:
.0014), whereas the CrSO2 values were predicted by TCI only (r = 0.65, P b .0001).
TCI ¼ PrSO2 =21:3ðslope 95%CI 20:0 to 22:5; Pb:0001Þ:
4. Discussion
In patients undergoing VA ECMO, CI can be estimated using the formula: CI ¼ ðPrSO2 =21:3Þ – EBFI: Stepwise multiple regression analysis revealed that in patients on VA ECMO, the PrSO2 values could be predicted by TCI (multiple correlation coefficient r = 0.67, P b .0001) and MAP (r = 0.38, P =
The major observation from the present study was the significant correlation found between NIRS oximetry and hemodynamic variables, primarily CI, in patients experiencing cardiogenic shock and in those undergoing VA ECMO. Moreover, our results indicate that PrSO2 values in particular can be used for the estimation of CI in these patients. These data, therefore, propose a method for rapid and noninvasive assessment of circulatory status in patients experiencing cardiogenic shock and patients undergoing VA ECMO, which also enables continuous monitoring. The advantages of this approach may
Please cite this article as: Ostadal P, et al, Noninvasive assessment of hemodynamic variables using near-infrared spectroscopy in patients experiencing cardiogenic shock and in..., J Crit Care (2014), http://dx.doi.org/10.1016/j.jcrc.2014.02.003
P. Ostadal et al. / Journal of Critical Care xxx (2014) xxx–xxx
be particularly useful in patients undergoing VA ECMO, in whom evaluation of circulatory status is limited, and in individuals with nonpulsatile blood flow. Currently, NIRS oximetry is widely used for the noninvasive measurement of brain oxygen saturation, and the maintenance of CrSO2 values within target limits is a clear therapeutic goal supported by evidence in the literature [7,10–12]. However, we emphasize the advantages of PrSO2 NIRS monitoring, which likely reflect a better estimation of circulatory status than CrSO2. We observed a correlation between both CrSO2 and PrSO2 values and CI (or TCI); however, a stronger relationship was found between PrSO2 and CI in patients experiencing cardiogenic shock. We speculate that the peripheral circulation is less influenced by the various reflexive and adaptive mechanisms than the cerebral circulation. This hypothesis is also supported by the fact that, contrary to PrSO2 and MAP, we did not find any correlation between CrSO2 and MAP in either patient group. The observed correlation between PrSO2 and MAP was also weaker in patients on VA ECMO. These results may reflect the efficacious blood exchange in oxygenator and the perfusion of lower extremities predominantly from extracorporeal blood flow. Nevertheless, based on the stepwise multiple regression, MAP remains the significant predictor for PrSO2, even in patients on VA ECMO. Our observations, however, only reflect hemodynamic alterations in a specific population of patients with cardiogenic shock. In contrast to our results, Lima et al [6] did not find a correlation between tissue oximetry (measured at the thenar eminence) and hemodynamic variables in a mixed population of critically ill patients. Similarly, Uilkema and Groeneveld [13] did not observe a correlation between thenar eminence tissue oximetry and centrally measured circulatory variables in patients after cardiac surgery. Georger et al [14] did not report a correlation between thenar eminence tissue oximetry and CI or MAP in severely hypotensive septic patients. Contrary to these studies, we measured PrSO2 using sensors placed at the calves. This is the standard approach at our institution because it enables rapid detection of lower limb ischemia, which represents one of the most common complications in patients undergoing VA ECMO. However, it is not clear whether different sensor positions (ie, thenar eminence, arms, or body) are equally sensitive to detect changes in circulation status in patients experiencing cardiogenic shock. Our data also contrast with the study by Hershenson et al [15], who reported a significant correlation between CrSO2 and MAP during supraventricular tachycardia. We did not observe a significant relationship between CrSO2 and MAP in either patients experiencing cardiogenic shock or patients undergoing VA ECMO; these differences may be explained by the diverse patient population and the different methods used for oximetry measurements. Our study had several limitations. First, the limited number of patients and the retrospective design may have introduced various biases. However, the present analysis was a pilot study, and we acknowledge that these results must be corroborated in a large, welldesigned trial. Furthermore, our results cannot be generalized to all cases of cardiogenic shock; however, assessment of target organ perfusion with NIRS oximetry provides an opportunity to evaluate global circulatory status, even in patients with complex circulatory disorders in whom other feasible methods for continuous monitoring are not available (ie, cardiogenic shock with left-to-right shunt and nonpulsatile blood flow on VA ECMO). Another major limitation is that the analysis combines within oneplot repeated measures of the same parameter in more than 1 individual. However, this is the pilot study, and the results should be
5
confirmed by a larger clinical trial before applying in clinical practice. The dependent values from individual patients are shown in the graphs by different colors. We cannot also exclude the possibility that the oximetry values were affected by the different paO2 levels; however, we used the same target paO2 values (10.0-13.0 kPa) in all participants. Similarly, paCO2 may significantly influence the cerebral circulation; once again, however, target values for paCO2 (4.0-6.5 kPa) were the same in all patients. Nevertheless, it should be pointed out that the described correlations may only apply within the constrained blood gas parameters.
5. Conclusions Our results suggest that NIRS oximetry is a method that enables rapid, noninvasive assessment and monitoring of circulatory status in patients with cardiogenic shock and those undergoing VA ECMO. Cardiac index can be estimated by dividing the PrSO2 value by 24.0 in patients experiencing cardiogenic shock, and TCI can be assessed by dividing the PrSO2 value by 21.3 in patients undergoing VA ECMO. Nevertheless, a large prospective trial is needed to confirm these observations and to validate the proposed formulas.
References [1] Murkin JM, Adams SJ, Novick RJ, et al. Monitoring brain oxygen saturation during coronary bypass surgery: a randomized, prospective study. Anesth Analg 2007;104:51–8. [2] Casati A, Fanelli G, Pietropaoli P, et al. Continuous monitoring of cerebral oxygen saturation in elderly patients undergoing major abdominal surgery minimizes brain exposure to potential hypoxia. Anesth Analg 2005;101:740–7 [table of contents]. [3] Mille T, Tachimiri ME, Klersy C, et al. Near infrared spectroscopy monitoring during carotid endarterectomy: which threshold value is critical? Eur J Vasc Endovasc Surg 2004;27:646–50. [4] Crookes BA, Cohn SM, Bloch S, et al. Can near-infrared spectroscopy identify the severity of shock in trauma patients? J Trauma 2005;58:806–13 [discussion 813–806]. [5] 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. [6] Lima A, van Bommel J, Sikorska K, et al. The relation of near-infrared spectroscopy with changes in peripheral circulation in critically ill patients. Crit Care Med 2011;39:1649–54. [7] Lipcsey M, Woinarski NC, Bellomo R. Near infrared spectroscopy (NIRS) of the thenar eminence in anesthesia and intensive care. Ann Intensive Care 2012;2:11. [8] Nanas S, Gerovasili V, Dimopoulos S, et al. Inotropic agents improve the peripheral microcirculation of patients with end-stage chronic heart failure. J Card Fail 2008;14:400–6. [9] Wong JK, Smith TN, Pitcher HT, et al. Cerebral and lower limb near-infrared spectroscopy in adults on extracorporeal membrane oxygenation. Artif Organs 2012;36:659–67. [10] Murkin JM, Adams SJ, Pardy E, et al. Monitoring brain oxygen saturation during coronary bypass surgery improves outcomes in diabetic patients: a post hoc analysis. Heart Surg Forum 2011;14:E1–6. [11] Scheeren TW, Schober P, Schwarte LA. Monitoring tissue oxygenation by near infrared spectroscopy (NIRS): background and current applications. J Clin Monit Comput 2012;26:279–87. [12] Ghosh A, Elwell C, Smith M. Review article: cerebral near-infrared spectroscopy in adults: a work in progress. Anesth Analg 2012;115:1373–83. [13] Uilkema RJ, Groeneveld AB. Correlates of thenar near-infrared spectroscopyderived tissue O2 saturation after cardiac surgery. Interact Cardiovasc Thorac Surg 2007;6:265–9. [14] Georger JF, Hamzaoui O, Chaari A, et al. Restoring arterial pressure with norepinephrine improves muscle tissue oxygenation assessed by near-infrared spectroscopy in severely hypotensive septic patients. Intensive Care Med 2010;36:1882–9. [15] Hershenson JA, Ro PS, Miao Y, et al. Changes in hemodynamic parameters and cerebral saturation during supraventricular tachycardia. Pediatr Cardiol 2012;33:286–9.
Please cite this article as: Ostadal P, et al, Noninvasive assessment of hemodynamic variables using near-infrared spectroscopy in patients experiencing cardiogenic shock and in..., J Crit Care (2014), http://dx.doi.org/10.1016/j.jcrc.2014.02.003
Contents lists available at ScienceDirect
Journal of Critical Care journal homepage: www.jccjournal.org
Noninvasive assessment of hemodynamic variables using near-infrared spectroscopy in patients experiencing cardiogenic shock and individuals undergoing venoarterial extracorporeal membrane oxygenation☆ Petr Ostadal, MD, PhD ⁎, Andreas Kruger, MD, Dagmar Vondrakova, MD, Marek Janotka, MD, Hana Psotova, MD, Petr Neuzil, MD, PhD Cardiovascular Center, Na Homolce Hospital, Prague, Czech Republic
a r t i c l e
i n f o
Keywords: Cardiogenic shock Extracorporeal membrane oxygenation Hemodynamics Near-infrared spectroscopy Oximetry
a b s t r a c t Purpose: The relationship between near-infrared spectroscopy cerebral oximetry (CrSO2), peripheral oximetry (PrSO2) and hemodynamic variables is not fully understood. Methods: The relationship between CrSO2/PrSO2 and cardiac index (CI), systemic vascular resistance index (SVRI) and mean arterial pressure (MAP) in patients experiencing cardiogenic shock and those undergoing venoarterial extracorporeal membrane oxygenation (ECMO) was retrospectively analyzed; in patients on ECMO, total circulatory index (TCI) was calculated from the sum of CI and extracorporeal blood flow index. Results: In patients experiencing cardiogenic shock (n = 10), significant correlations between PrSO2 values and CI (Spearman r = 0.81; P b .0001), SVRI (r = −0.45; P b .0001), and MAP (r = 0.58; P b .0001) were found. Significant correlations between CrSO2 and CI (r = 0.55; P b .0001) and SVRI (r = −0.47; P b .0001), but not MAP, were observed. Linear regression analysis revealed that CI could be calculated using the following equation: CI = PrSO2/24.0. In patients on VA ECMO (n = 12), significant correlations were found between PrSO2 and TCI (r = 0.68; P b .0001), SVRI (r = − 0.47; P b .0001), and MAP (r = 0.27; P = .025). Significant correlations were also found between CrSO2 and TCI (r = 0.68; P b .0001) and SVRI (r = −0.51; P b .0001), but not MAP. Conclusions: Results of the present study suggest that CrSO2 and PrSO2 in particular can be used for noninvasive estimation and monitoring of global circulatory status in patients experiencing cardiogenic shock and individuals undergoing ECMO. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Measurement of hemodynamic variables is often valuable in many critical conditions, especially in patients experiencing cardiogenic shock. However, current methods for cardiac output (CO) measurement are often invasive and limited in their ability to continuously monitor; furthermore, they are not applicable in some cases such as patients undergoing venoarterial extracorporeal membrane oxygenation (VA ECMO). The criterion standard for CO measurement in intensive care—thermodilution using pulmonary artery catheterization (PAC)—does not allow proper measurement of global hemodynamic status in patients on VA ECMO because (i) a substantial portion of blood flow bypasses the lungs and (ii) thermodilution with PAC often losses accuracy in these patients when pulmonary artery flow falls
☆ The present study was supported by a grant from the Czech Ministry of Health, No. 12153. ⁎ Corresponding author. Department of Cardiology, Cardiovascular Center, Na Homolce Hospital, Roentgenova 2, 15000 Prague, Czech Republic. Tel.: +420 257272208; fax: +420 257272974. E-mail address: [email protected] (P. Ostadal).
below 2 L/min. Frequently used minimally invasive or noninvasive methods such as pulse-wave analysis are not suitable in substantial part of these individuals, for example, because of continual blood flow. Furthermore, transpulmonary thermodilution is inappropriate for CO measurements in patients with extracorporeal circulation. Similarly, although echocardiography provides important data regarding cardiac functions, it does not enable continuous monitoring and assessment of global circulatory status in patients undergoing VA ECMO. Near-infrared spectroscopy (NIRS) oximetry has recently emerged as a promising method for the noninvasive measurement of cerebral and tissue oxygenation and perfusion. Several studies have demonstrated its use in the assessment of perioperative brain perfusion in cardiac [1] and noncardiac [2] surgery, in carotid interventions [3], and for monitoring brain and tissue perfusion in several critical conditions [4–7]. To date, however, there are only limited data regarding measurement of cerebral regional oxygen saturation (CrSO2) and peripheral regional oxygen saturation (PrSO2) using NIRS in patients experiencing cardiogenic shock and those undergoing VA ECMO [8,9]. In the present study, we tested the hypothesis that NIRS oximetry can be used to estimate global circulatory status in patients experiencing cardiogenic shock and in patients with cardiogenic shock undergoing
0883-9441/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jcrc.2014.02.003
Please cite this article as: Ostadal P, et al, Noninvasive assessment of hemodynamic variables using near-infrared spectroscopy in patients experiencing cardiogenic shock and in..., J Crit Care (2014), http://dx.doi.org/10.1016/j.jcrc.2014.02.003
2
P. Ostadal et al. / Journal of Critical Care xxx (2014) xxx–xxx
VA ECMO, in whom standard methods of hemodynamic measurement cannot be easily used. 2. Materials and methods 2.1. Study overview The present study was approved by the Na Homolce Hospital Ethics Committee (Prague, Czech Republic). Written informed consent for data collection from patient’s medical records was not required for this retrospective study, and all data were collected and analyzed in a de-identified form. We performed a retrospective analysis of medical records from patients experiencing cardiogenic shock hospitalized at the Cardiology Intensive Care Unit, Na Homolce Hospital in Prague, in whom both CrSO2 and PrSO2 4-channel NIRS oximetry (INVOS, Covidien, Dublin, Ireland) was monitored simultaneously with hemodynamic monitoring using PAC (CCOmbo Pulmonary Artery Catheter and Vigilance II Monitor; Edwards Lifesciences, Irvine, California). The indication for PAC was insufficient response to inotropic and vasopressor therapy (defined as norepinephrine dose N0.2 μg kg −1 min −1 and dobutamin dose N5 μg kg −1 min −1). Since 2010, CrSO2 and PrSO2 NIRS oximetry monitoring has been routinely used at the Na Homolce Hospital in most patients experiencing cardiogenic shock and in almost all patients undergoing VA ECMO. Cerebral oximetry has been performed using two forehead sensors, and peripheral oximetry using 2 sensors placed at the calves. Cerebral oximetry (N 50%) became also one of the major therapeutic goals in these patients, besides mixed venous oxygen saturation and mean arterial pressure (MAP). The medical records of these patients contained oximetry and hemodynamic data collected at 1-hour intervals. In the present analysis, we included values obtained at the beginning of measurement and subsequently at 6-hour intervals for a maximum of 2 days (ie, a maximum of 9 data sets per patient). The reason for selection only data recorded at 6-hour intervals for the analysis was to reflect more different global circulatory status. In patients experiencing cardiogenic shock, values of CO, MAP, central venous pressure, and body surface area were obtained from the medical records; cardiac index (CI) and systemic vascular resistance index (SVRI) were then calculated. In patients undergoing VA ECMO, total circulatory index (TCI) was calculated according to the following formula: TCI ¼ CI þ extracorporeal blood flow index ðEBFIÞ; EBFI ¼ extracorporeal blood flow=BSA In VA ECMO patients, SVRI was then calculated from TCI instead of CI:
recently underwent simultaneous monitoring of NIRS oximetry and hemodynamics were identified. Up to May 2013, 69 nonsurgical patients who experienced cardiogenic shock with persisting or progressing tissue hypoperfusion despite standard therapeutic approaches or refractory cardiac arrest underwent successful VA ECMO (Levitronix Centrimag pump (Thoratec Corporation, Pleasanton, California) using the Quadrox oxygenator (Maquet Cardiopulmonary AG, Hirrlingen, Germany) or Cardiohelp system (Maquet Cardiopulmonary AG, Hirrlingen, Germany) at the Na Homolce Hospital. In 2010, the hospital began using cerebral and peripheral oximetry in these patients, and to date, 44 patients on VA ECMO have been monitored using this noninvasive method. From this group, 12 individuals who underwent simultaneous monitoring of NIRS oximetry and hemodynamics using PAC were identified. In all of these individuals, PAC was placed already before ECMO introduction. In patients with limb ischemia caused by arterial outflow cannula (oximetry b 40% with absolute difference N10% with contralateral limb), distal perfusion bypass was introduced using 5F or 6F arterial sheath. 2.3. Statistical analysis A mean value was calculated for each pair of NIRS oximetry data (left and right brain hemisphere or left and right lower limb); this value was used for further analysis. Spearman correlation and linear regression were performed to analyze the relationship between oximetry and hemodynamic values using GraphPad Prism 6 software (GraphPad Software, Inc, San Diego, California). Stepwise multiple regression analyses of PrSO2 and CrSO2 against the set of hemodynamic parameters were conducted using the MedCalc 12 software (MedCalc Software bvba, Ostend, Belgium). P b .05 was considered statistically significant. 3. Results The baseline characteristics of individual patients are shown in Table 1. In patients experiencing cardiogenic shock, a strong, highly significant correlation was observed between PrSO2 and CI (r = 0.81; 95% confidence interval [95% CI], 0.71-0.87; P b .0001; Fig. 1A), and a weaker yet still highly significant correlation was found between
Table 1 Baseline characteristics of individual patients
Cardiogenic shock
SVRI ¼ ðMAP–CVPÞ 80=TCI In patients on ventilation support, ventilator parameters were set to maintain a target partial pressure of oxygen in arterial blood (paO2) of 10.0 to 13.0 kPa and partial pressure of carbon dioxide (paCO2) of 4.0 to 6.5 kPa. Similarly, in patients undergoing VA ECMO, oxygenator parameters were set to maintain the same target values in the arterial (ie, outflow) line. Blood pressure was monitored invasively from the radial or femoral artery in all patients. 2.2. Study population Most patients with cardiogenic shock experienced acute myocardial infarction or decompensated chronic heart failure. All were treated with inotropes and vasopressors, and the majority (80%) with an intra-aortic balloon pump. All patients who experienced acute myocardial infarction underwent primary percutaneous coronary intervention. For the analysis, 10 consecutive individuals who
VA ECMO
Patient
Cause of circulatory failure
Age (y)
Sex
Alive
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10 11 12
AMI AMI, MR DCMP DCMP AMI ICMP AMI AMI, MR DCMP AMI CA, AMI, MR Myocarditis AMI AMI DCMP DCMP DCMP AMI AMI CA, AMI AMI CA, AMI
53 52 68 60 72 66 67 58 69 71 65 66 66 67 70 67 61 63 65 24 58 54
M M F M M M F M F M M F M F M M M M M M M M
No Yes Yes Yes No Yes No Yes Yes Yes Yes No Yes No Yes Yes Yes Yes No No Yes Yes
AMI indicates acute myocardial infarction; CA, cardiac arrest; DCMP, dilated cardiomyopathy; F, female; ICMP, ischemic cardiomyopathy; M, male; MR, severe mitral regurgitation. Patients marked “Alive: Yes” were either discharged from hospital or alive at 30 days.
Please cite this article as: Ostadal P, et al, Noninvasive assessment of hemodynamic variables using near-infrared spectroscopy in patients experiencing cardiogenic shock and in..., J Crit Care (2014), http://dx.doi.org/10.1016/j.jcrc.2014.02.003
P. Ostadal et al. / Journal of Critical Care xxx (2014) xxx–xxx
3
Fig. 1. Correlation between NIRS oximetry and hemodynamic variables in patients experiencing cardiogenic shock. Dependent values from individual patients are shown in different colors. A, Correlation of peripheral oximetry and CI. B, Correlation of peripheral oximetry and SVRI. C, Correlation of peripheral oximetry and MAP. D, Correlation of cerebral oximetry and CI. E, Correlation of cerebral oximetry and systemic vascular resistance. F, Correlation of cerebral oximetry and MAP.
PrSO2 and SVRI (r = − 0.45; 95% CI, − 0.62 to − 0.23; P b .0001; Fig. 1B) and between PrSO2 and MAP (r = 0.58; 95% CI, 0.38-0.71; P b .0001; Fig. 1C). Moreover, a significant correlation between CrSO2 and CI (r = 0.55; 95% CI, 0.37-0.70; P b .0001; Fig. 1D) and between CrSO2 and SVRI (r = − 0.47; 95% CI, − 0.63 to − 0.27; P b .0001; Fig. 1E) was found, but not between CrSO2 and MAP (r = 0.12; 95% CI, − 0.12 to 0.34; P b .3031; Fig. 1F). An equation relating PrSO2 and CI was determined using linear regression: CI ¼ PrSO2 =24:0ðslope 95%CI 23:0 to 25:0; Pb:0001Þ: Stepwise multiple regression analysis revealed that the PrSO2 values could be predicted by CI (multiple correlation coefficient r =
0.60, P b .0001), MAP (r = 0.43, P = .0002), and SVRI (r = − 0.24, P = .0454), whereas the CrSO2 values were predicted by MAP (r = 0.45, P = .0001) and SVRI (r = − 0.42, P = .0002). In patients undergoing VA ECMO, a significant correlation between PrSO2 and TCI (r = 0.68; 95% CI, 0.53-0.79; P b .0001; Fig. 2A) and between PrSO2 and SVRI (r = − 0.47; 95% CI, − 0.63 to − 0.26; P b .0001; Fig. 2B) was found, and a weak yet still significant correlation between PrSO2 and MAP (r = 0.27; 95% CI, 0.03-0.47; P = .025; Fig. 2C) was found. Furthermore, a significant correlation between CrSO2 and TCI (r = 0.68; 95% CI, 0.52-0.79; P b .0001; Fig. 2D) and between CrSO2 and SVRI (r = − 0.51; 95% CI, − 0.67 to − 0.31; P b .0001; Fig. 2E) was found, but not between CrSO2 and MAP (r = 0.09; 95% CI, − 0.15 to 0.32; P = .46; Fig. 2F). An
Please cite this article as: Ostadal P, et al, Noninvasive assessment of hemodynamic variables using near-infrared spectroscopy in patients experiencing cardiogenic shock and in..., J Crit Care (2014), http://dx.doi.org/10.1016/j.jcrc.2014.02.003
4
P. Ostadal et al. / Journal of Critical Care xxx (2014) xxx–xxx
Fig. 2. Correlation between NIRS oximetry and hemodynamic variables in patients undergoing ECMO. Dependent values from individual patients are shown in different colors. A, Correlation of peripheral oximetry and TCI. B, Correlation of peripheral oximetry and SVRI. C, Correlation of peripheral oximetry and MAP. D, Correlation of cerebral oximetry and TCI. E, Correlation of cerebral oximetry and systemic vascular resistance. F, Correlation of cerebral oximetry and MAP.
equation relating PrSO 2 and TCI was determined using linear regression:
.0014), whereas the CrSO2 values were predicted by TCI only (r = 0.65, P b .0001).
TCI ¼ PrSO2 =21:3ðslope 95%CI 20:0 to 22:5; Pb:0001Þ:
4. Discussion
In patients undergoing VA ECMO, CI can be estimated using the formula: CI ¼ ðPrSO2 =21:3Þ – EBFI: Stepwise multiple regression analysis revealed that in patients on VA ECMO, the PrSO2 values could be predicted by TCI (multiple correlation coefficient r = 0.67, P b .0001) and MAP (r = 0.38, P =
The major observation from the present study was the significant correlation found between NIRS oximetry and hemodynamic variables, primarily CI, in patients experiencing cardiogenic shock and in those undergoing VA ECMO. Moreover, our results indicate that PrSO2 values in particular can be used for the estimation of CI in these patients. These data, therefore, propose a method for rapid and noninvasive assessment of circulatory status in patients experiencing cardiogenic shock and patients undergoing VA ECMO, which also enables continuous monitoring. The advantages of this approach may
Please cite this article as: Ostadal P, et al, Noninvasive assessment of hemodynamic variables using near-infrared spectroscopy in patients experiencing cardiogenic shock and in..., J Crit Care (2014), http://dx.doi.org/10.1016/j.jcrc.2014.02.003
P. Ostadal et al. / Journal of Critical Care xxx (2014) xxx–xxx
be particularly useful in patients undergoing VA ECMO, in whom evaluation of circulatory status is limited, and in individuals with nonpulsatile blood flow. Currently, NIRS oximetry is widely used for the noninvasive measurement of brain oxygen saturation, and the maintenance of CrSO2 values within target limits is a clear therapeutic goal supported by evidence in the literature [7,10–12]. However, we emphasize the advantages of PrSO2 NIRS monitoring, which likely reflect a better estimation of circulatory status than CrSO2. We observed a correlation between both CrSO2 and PrSO2 values and CI (or TCI); however, a stronger relationship was found between PrSO2 and CI in patients experiencing cardiogenic shock. We speculate that the peripheral circulation is less influenced by the various reflexive and adaptive mechanisms than the cerebral circulation. This hypothesis is also supported by the fact that, contrary to PrSO2 and MAP, we did not find any correlation between CrSO2 and MAP in either patient group. The observed correlation between PrSO2 and MAP was also weaker in patients on VA ECMO. These results may reflect the efficacious blood exchange in oxygenator and the perfusion of lower extremities predominantly from extracorporeal blood flow. Nevertheless, based on the stepwise multiple regression, MAP remains the significant predictor for PrSO2, even in patients on VA ECMO. Our observations, however, only reflect hemodynamic alterations in a specific population of patients with cardiogenic shock. In contrast to our results, Lima et al [6] did not find a correlation between tissue oximetry (measured at the thenar eminence) and hemodynamic variables in a mixed population of critically ill patients. Similarly, Uilkema and Groeneveld [13] did not observe a correlation between thenar eminence tissue oximetry and centrally measured circulatory variables in patients after cardiac surgery. Georger et al [14] did not report a correlation between thenar eminence tissue oximetry and CI or MAP in severely hypotensive septic patients. Contrary to these studies, we measured PrSO2 using sensors placed at the calves. This is the standard approach at our institution because it enables rapid detection of lower limb ischemia, which represents one of the most common complications in patients undergoing VA ECMO. However, it is not clear whether different sensor positions (ie, thenar eminence, arms, or body) are equally sensitive to detect changes in circulation status in patients experiencing cardiogenic shock. Our data also contrast with the study by Hershenson et al [15], who reported a significant correlation between CrSO2 and MAP during supraventricular tachycardia. We did not observe a significant relationship between CrSO2 and MAP in either patients experiencing cardiogenic shock or patients undergoing VA ECMO; these differences may be explained by the diverse patient population and the different methods used for oximetry measurements. Our study had several limitations. First, the limited number of patients and the retrospective design may have introduced various biases. However, the present analysis was a pilot study, and we acknowledge that these results must be corroborated in a large, welldesigned trial. Furthermore, our results cannot be generalized to all cases of cardiogenic shock; however, assessment of target organ perfusion with NIRS oximetry provides an opportunity to evaluate global circulatory status, even in patients with complex circulatory disorders in whom other feasible methods for continuous monitoring are not available (ie, cardiogenic shock with left-to-right shunt and nonpulsatile blood flow on VA ECMO). Another major limitation is that the analysis combines within oneplot repeated measures of the same parameter in more than 1 individual. However, this is the pilot study, and the results should be
5
confirmed by a larger clinical trial before applying in clinical practice. The dependent values from individual patients are shown in the graphs by different colors. We cannot also exclude the possibility that the oximetry values were affected by the different paO2 levels; however, we used the same target paO2 values (10.0-13.0 kPa) in all participants. Similarly, paCO2 may significantly influence the cerebral circulation; once again, however, target values for paCO2 (4.0-6.5 kPa) were the same in all patients. Nevertheless, it should be pointed out that the described correlations may only apply within the constrained blood gas parameters.
5. Conclusions Our results suggest that NIRS oximetry is a method that enables rapid, noninvasive assessment and monitoring of circulatory status in patients with cardiogenic shock and those undergoing VA ECMO. Cardiac index can be estimated by dividing the PrSO2 value by 24.0 in patients experiencing cardiogenic shock, and TCI can be assessed by dividing the PrSO2 value by 21.3 in patients undergoing VA ECMO. Nevertheless, a large prospective trial is needed to confirm these observations and to validate the proposed formulas.
References [1] Murkin JM, Adams SJ, Novick RJ, et al. Monitoring brain oxygen saturation during coronary bypass surgery: a randomized, prospective study. Anesth Analg 2007;104:51–8. [2] Casati A, Fanelli G, Pietropaoli P, et al. Continuous monitoring of cerebral oxygen saturation in elderly patients undergoing major abdominal surgery minimizes brain exposure to potential hypoxia. Anesth Analg 2005;101:740–7 [table of contents]. [3] Mille T, Tachimiri ME, Klersy C, et al. Near infrared spectroscopy monitoring during carotid endarterectomy: which threshold value is critical? Eur J Vasc Endovasc Surg 2004;27:646–50. [4] Crookes BA, Cohn SM, Bloch S, et al. Can near-infrared spectroscopy identify the severity of shock in trauma patients? J Trauma 2005;58:806–13 [discussion 813–806]. [5] 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. [6] Lima A, van Bommel J, Sikorska K, et al. The relation of near-infrared spectroscopy with changes in peripheral circulation in critically ill patients. Crit Care Med 2011;39:1649–54. [7] Lipcsey M, Woinarski NC, Bellomo R. Near infrared spectroscopy (NIRS) of the thenar eminence in anesthesia and intensive care. Ann Intensive Care 2012;2:11. [8] Nanas S, Gerovasili V, Dimopoulos S, et al. Inotropic agents improve the peripheral microcirculation of patients with end-stage chronic heart failure. J Card Fail 2008;14:400–6. [9] Wong JK, Smith TN, Pitcher HT, et al. Cerebral and lower limb near-infrared spectroscopy in adults on extracorporeal membrane oxygenation. Artif Organs 2012;36:659–67. [10] Murkin JM, Adams SJ, Pardy E, et al. Monitoring brain oxygen saturation during coronary bypass surgery improves outcomes in diabetic patients: a post hoc analysis. Heart Surg Forum 2011;14:E1–6. [11] Scheeren TW, Schober P, Schwarte LA. Monitoring tissue oxygenation by near infrared spectroscopy (NIRS): background and current applications. J Clin Monit Comput 2012;26:279–87. [12] Ghosh A, Elwell C, Smith M. Review article: cerebral near-infrared spectroscopy in adults: a work in progress. Anesth Analg 2012;115:1373–83. [13] Uilkema RJ, Groeneveld AB. Correlates of thenar near-infrared spectroscopyderived tissue O2 saturation after cardiac surgery. Interact Cardiovasc Thorac Surg 2007;6:265–9. [14] Georger JF, Hamzaoui O, Chaari A, et al. Restoring arterial pressure with norepinephrine improves muscle tissue oxygenation assessed by near-infrared spectroscopy in severely hypotensive septic patients. Intensive Care Med 2010;36:1882–9. [15] Hershenson JA, Ro PS, Miao Y, et al. Changes in hemodynamic parameters and cerebral saturation during supraventricular tachycardia. Pediatr Cardiol 2012;33:286–9.
Please cite this article as: Ostadal P, et al, Noninvasive assessment of hemodynamic variables using near-infrared spectroscopy in patients experiencing cardiogenic shock and in..., J Crit Care (2014), http://dx.doi.org/10.1016/j.jcrc.2014.02.003
