- Email: [email protected]
Femoral venous oxygen saturation obtained during CPR predicts successful resuscitation in a pig model
American Journal of Emergency Medicine 33 (2015) 941–945
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Original Contribution
Femoral venous oxygen saturation obtained during CPR predicts successful resuscitation in a pig model☆ Mu Jin Kim, MD a, Kyung Woon Jeung, MD, PhD b, Byung Kook Lee, MD, PhD b, Sung Soo Choi, EMTP c, Sang Wook Park, EMTP b, Kyung Hwan Song, MD b, Sung Min Lee, MD b, Yong Il Min, MD, PhD b,⁎ a b c
Department of Emergency Medicine, Songjeong Sarang Hospital, Gwangju, Republic of Korea Department of Emergency Medicine, Chonnam National University Hospital, Gwangju, Republic of Korea Department of Emergency Medical Service, Howon University, Gunsan, Republic of Korea
a r t i c l e
i n f o
Article history: Received 15 January 2015 Received in revised form 16 March 2015 Accepted 6 April 2015
a b s t r a c t Purpose: Central venous oxygen saturation has been shown to reflect the adequacy of tissue oxygenation during cardiopulmonary resuscitation (CPR), thereby enabling the assessment of CPR quality and the prediction of restoration of spontaneous circulation (ROSC). The femoral vein can be easily accessed during CPR. We determined if femoral venous oxygen saturation (SFVO2) values obtained during CPR could reliably predict ROSC in a pig model. Methods: After 15 minutes of untreated ventricular fibrillation followed by 8 minutes of basic life support, 19 pigs underwent advanced cardiovascular life support. During advanced cardiovascular life support, femoral venous blood samples were obtained at 4-minute intervals. The abilities of SFVO2 and coronary perfusion pressure (CPP) to predict ROSC were evaluated by calculating the areas under receiver operating characteristic curves (AUCs). Results: Eight pigs (42.1%) achieved ROSC. The resuscitated animals had significantly higher CPP (P b .001) and SFVO2 (P b .001) values than the nonresuscitated animals, and there was a significant correlation between SFVO2 and CPP values (r = 0.684; P b .001). The CPPs of the resuscitated and nonresuscitated animals overlapped considerably; however, there was minimal overlap between the 2 groups for SFVO2. Femoral venous oxygen saturation significantly predicted ROSC with an AUC of 0.997 (95% confidence interval, 0.911-1.000; P b .001), and it had a larger AUC than CPP (AUC, 0.964; 95% confidence interval, 0.855-0.997; P b .001). The AUC difference, however, was not statistically significant (P = .157). Conclusion: In this study, SFVO2 values obtained during CPR exhibited a significant correlation with CPP and reliably predicted ROSC. © 2015 Elsevier Inc. All rights reserved.
1. Introduction In recent years, there has been an increased emphasis on physiologic monitoring during cardiopulmonary resuscitation (CPR) [1]. Several studies have examined coronary perfusion pressure (CPP), end-tidal carbon dioxide (ETCO2), and central venous oxygen saturation (SCVO2) as means for monitoring the effectiveness of CPR, and recent resuscitation guidelines recommended the use of these parameters, when feasible, to optimize chest compressions and guide vasopressor therapy during cardiac arrest [1-8]. Moreover, recent studies have suggested that titrating resuscitative efforts according to these parameters improves resuscitation outcomes [9-11].
☆ Funding sources/disclosures: This research was supported by Basic Science Research Program through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology (NRF-2012R1A1A2039466). The funder had no role in the study design, data collection, analysis, decision to publish, or the preparation of the manuscript. ⁎ Corresponding author. Tel.: +82 62 220 6809; fax: +82 62 228 7417. E-mail address: [email protected] (Y.I. Min). http://dx.doi.org/10.1016/j.ajem.2015.04.004 0735-6757/© 2015 Elsevier Inc. All rights reserved.
The key to successful resuscitation is tissue oxygen delivery sufficient to meet the cellular demands of ischemic vital organs. Coronary perfusion pressure and ETCO2 correlate with blood flow generated during CPR [12,13]. However, these variables do not provide information about the relationship between systemic oxygen delivery and demand. SCVO2, which is used as a surrogate for mixed venous oxygen saturation, is different from ETCO2 and CPP in that it reflects the balance between systemic oxygen delivery and demand during CPR. Central venous oxygen saturation has been shown to reflect the adequacy of tissue oxygenation during CPR, thus enabling the assessment of CPR quality and the prediction of restoration of spontaneous circulation (ROSC) [8,14]. In a study that included 100 out-of-hospital cardiac arrest (OHCA) patients, patients who achieved ROSC were found to have higher SCVO2 values than those who could not be resuscitated [8]. No patient in that study experienced ROSC without reaching a SCVO2 of at least 30%. The measurement of SCVO2 requires a catheter to be placed in the vena cava near the right atrium (RA), which is usually inserted through the internal jugular or subclavian vein. However, placement of a central venous catheter through the internal jugular or subclavian vein is usually impractical during CPR because it interferes with resuscitation interventions, including chest compressions and airway
942
M.J. Kim et al. / American Journal of Emergency Medicine 33 (2015) 941–945
management [15]. Currently, the use of SCVO2 during CPR is limited to patients who have a central venous catheter in place before cardiac arrest. The femoral vein can be easily accessed during CPR, and establishing femoral venous access does not interfere with resuscitation interventions. Previous studies have shown that femoral venous cannulation can be performed quickly and safely with a high rate of success, even during CPR [16,17]. Therefore, if oxygen saturation of venous blood obtained from the femoral vein can be used with confidence as a substitute for SCVO2 during CPR, then the use of venous oxygen saturation to assess CPR quality and predict ROSC may be able to be extended to cardiac arrest patients who do not have a central venous catheter in place before cardiac arrest (which is the case for most OHCA patients). However, to our knowledge, the suitability of femoral venous oxygen saturation (SFV O 2) as a substitute for SCVO2 has not yet been investigated. In this study, we investigated the reliability of SFVO2 in predicting ROSC using a realistic model of OHCA. We hypothesized that SFVO2 values obtained during CPR can reliably predict ROSC. 2. Methods The Animal Care and Use Committee of Chonnam National University approved the protocol of this study. All animal care and experiments were conducted according to the Institutional Animal Care and Use Committee guidelines. 2.1. Animal preparation Nineteen male domestic pigs weighing 25.0 kg (22.5-29.7) were used. The animal preparation was performed as previously described [18-20]. After premedication (ketamine 20 mg/kg, xylazine 2.2 mg/kg, atropine 0.04 mg/kg, administered intramuscularly), anesthesia was induced with 50%:50% N2O:O2 and 2% to 5% sevoflurane via a mask. After tracheal intubation, the pigs were ventilated at a tidal volume of 15 mL/kg. Anesthesia was continued with 70%:30% N2O:O2 and 0.5% to 2% sevoflurane, which was titrated to prevent signs of pain (reactive wide pupils, tachycardia, and hypertension). In our experience [18-20], the experimental preparation was completed in a very short time (approximately 40 minutes), and neither hemodynamic instabilities nor signs of pain were observed with this anesthesia protocol. The ventilatory rates were adjusted to achieve normocapnia. A catheter was inserted into an ear vein, and saline was administered to maintain an RA pressure of 6 to 12 mm Hg. A double-lumen catheter was advanced from the right femoral artery to the thoracic aorta to monitor aortic pressure and to obtain blood samples. A 6F introducer sheath was inserted into the right femoral vein and advanced 3 cm under direct visualization to obtain femoral venous blood samples. The right external jugular vein was cannulated with an introducer sheath to monitor the RA pressure and to insert a right ventricle pacing catheter. Electrocardiogram leads were placed on the limbs to monitor the heart rhythm, and the rectal temperature was maintained between 37°C and 38°C.
8/min with high-flow oxygen (10 L/min) through the use of a volumemarked bag devised by Cho et al [21]. At the commencement of ACLS, all animals received 0.5 U/kg of vasopressin intravenously. After 4 minutes of ACLS, 0.02 mg/kg of epinephrine was administered every 3 minutes as needed. During ACLS, defibrillation was attempted using a single biphasic 150-J electric shock at 2-minute intervals. Sustained ROSC was defined as maintenance of a systolic aortic pressure of at least 60 mm Hg for at least 5 consecutive minutes [22,23]. If ROSC was not achieved within 12 minutes of ACLS, the resuscitation efforts were discontinued. The successfully resuscitated animals were euthanized using potassium chloride. Autopsies were performed to document any injuries to the thoracic and abdominal cavities caused by CPR. 2.3. Measurements An investigator obtained 2 mL of blood from the introducer sheath inserted into the femoral vein using a BD Preset syringe (Becton, Dickinson and Company, Plymouth, UK) immediately before the induction of VF, 3 minutes after the start of ACLS, and at 4-minute intervals thereafter until ROSC was achieved or until the end of the 12-minute ACLS period. All blood samples were immediately analyzed using a blood gas/electrolyte analyzer (GEM Premier 3000; Instrumentation Laboratory, Lexington, KY). The aortic pressure, RA pressure, and standard lead II electrocardiogram were continuously monitored (CS/3 CCM; Datex-Ohmeda, Helsinki, Finland) and transferred to a personal computer using S/5 Collect software (Datex-Ohmeda). The CPP at the time of femoral venous blood sampling was calculated by subtracting the RA end-diastolic pressure from the simultaneous aortic enddiastolic pressure; it was averaged over 5 compression cycles. 2.4. Statistical analysis The sample size calculation was based on a previous study [8] and resulted in a minimum requirement of 7 animals in each group (ROSC vs no ROSC), with α set at .05 and a power of 0.90. Assuming an ROSC rate of 40% based on our previous studies [18-20], 19 animals were used for this study. Continuous variables were tested for normality using the Shapiro-Wilk test. Normally distributed variables were analyzed using independent t tests and summarized as means ± SDs, whereas nonnormally distributed variables were analyzed using Mann-Whitney U tests and summarized as medians with interquartile ranges. Comparisons of categorical variables were performed using χ2 or Fisher exact test, as indicated. Correlations between nonnormally distributed continuous variables were assessed using Spearman rank correlation coefficients. Receiver operating characteristic (ROC) analyses were performed to examine the abilities of SFVO2 and CPP to predict ROSC. Comparisons of the areas under the ROC curves (AUCs) were performed as recommended by DeLong et al [24]. Data analyses were performed using PASW/SPSS software, version 18 (IBM, Inc, Chicago, IL). The ROC curves were calculated and compared using MedCalc version 14.10.2 (free trial; MedCalc Software, Mariakerke, Belgium). Statistical significance was set at P b .05.
2.2. Cardiac arrest and CPR 3. Results Ventricular fibrillation (VF) was induced by applying an electrical current (60 Hz, 30 mA alternating current) via a right ventricle pacing catheter. After 15 minutes of untreated VF, basic life support using cycles of 30 chest compressions followed by 2 ventilations with ambient air was started. Closed chest compressions were administered by 2 investigators (SSC and SML) for all of the animals. A HeartStart MRx Monitor/ Defibrillator (Philips Medical Systems, Seattle, WA) with Q-CPR assisted in maintaining a chest compression rate of 100/min and a compression depth of 25% of the anterior-posterior diameter of the chest wall. After 8 minutes of basic life support, advanced cardiovascular life support (ACLS) was initiated according to current resuscitation guidelines [1]. Asynchronous positive pressure ventilations were provided at a rate of
3.1. Baseline characteristics Precardiac arrest baseline characteristics are displayed in Table 1. There were no significant differences between the groups in terms of the baseline measurements. 3.2. Coronary perfusion pressure and blood gases obtained from the femoral vein during CPR Of the 19 animals studied, 8 (42.1%) achieved ROSC after 6.0 minutes (6.0-7.5 minutes) of ACLS. Twelve femoral venous blood samples (8 at
M.J. Kim et al. / American Journal of Emergency Medicine 33 (2015) 941–945
943
Table 1 Prearrest baseline characteristics
Weight (kg) Male sex, n (%) Systolic aortic pressure (mm Hg) Diastolic aortic pressure (mm Hg) Mean aortic pressure (mm Hg) Systolic RA pressure (mm Hg) Diastolic RA pressure (mm Hg) Mean RA pressure (mm Hg) Arterial pH Arterial PCO2 (mm Hg) Arterial PO2 (mm Hg) Arterial HCO3− (mmol/L) Arterial base excess (mmol/L) Arterial SO2 (%) Venous pH Venous PCO2 (mm Hg) Venous PO2 (mm Hg) Venous HCO3− (mmol/L) Venous base excess (mmol/L) Venous SO2 (%)
Total (n = 19)
No ROSC (n = 11)
ROSC (n = 8)
P
25.0 (22.5-29.7) 13 (68.4) 117 (111-125) 77 (71-87) 93 (88-102) 10.0 (9.0-12.5) 6.0 (5.0-7.5) 8.0 (7.0-9.5) 7.500 (7.460-7.525) 37.0 (36.0-40.0) 140.0 (99.5-148.5) 30.3 (28.1-31.2) 7.3 (4.4-8.6) 99.0 (98.0-99.0) 7.430 (7.395-7.460) 49.0 (46.5-51.0) 34.0 (30.5-39.0) 30.0 (27.9-31.2) 7.7 (5.1-9.8) 67.0 (60.0-78.0)
25.0 (22.7-31.6) 8 (72.7) 117 (111-129) 79 (73-86) 93 (88-102) 11.0 (9.5-12.5) 6.0 (5.5-7.5) 8.0 (7.5-9.5) 7.500 (7.465-7.515) 37.0 (36.0-39.5) 140.0 (111.5-147.5) 30.3 (28.1-30.8) 7.3 (4.4-8.2) 99.0 (98.5-99.0) 7.420 (7.400-7.455) 49.0 (46.5-51.0) 33.0 (29.0-39.5) 29.7 (27.9-31.2) 7.3 (5.1-9.8) 64.0 (53.5-75.0)
24.8 (22.6-26.8) 5 (62.5) 117 (111-123) 75 (67-89) 96 (86-101) 9.5 (8.8-12.0) 5.5 (5.0-6.8) 7.5 (7.0-9.0) 7.510 (7.445-7.538) 39.5 (36.3-40.3) 123.0 (96.8-152.8) 30.3 (27.7-31.9) 7.1 (4.0-9.4) 98.5 (97.8-99.3) 7.450 (7.388-7.473) 47.5 (46.8-50.3) 36.5 (32.3-39.0) 30.4 (27.5-31.6) 8.3 (5.2-9.9) 71.5 (62.0-78.0)
.836 1.000 .836 .563 .869 .505 .767 .674 .804 .588 .836 .679 .741 .692 .454 .618 .535 .836 .804 .619
3 minutes, 2 at 7 minutes, and 2 at 11 minutes after the start of ACLS) were obtained during ACLS in the successfully resuscitated animals. In the 11 animals that did not achieve sustained ROSC, 3 femoral venous blood samples were obtained from each (at 3, 7, and 11 minutes after the start of ACLS, respectively), except for 1 animal in which a blood sample was only obtained at 3 minutes after the start of ACLS. This animal achieved ROSC for 2 minutes at 4 minutes after the start of ACLS and thereafter could not achieve sustained ROSC. Thus, a total of 31 femoral venous blood samples were obtained during ACLS in the nonresuscitated animals. The results for CPP (at the time of femoral venous blood sampling) and blood gases from the femoral venous blood samples obtained during ACLS are displayed in Table 2. The resuscitated animals had significantly higher CPP (P b .001) and SFVO2 (P b .001) values than the nonresuscitated animals. There was a significant positive correlation between SFVO2 and CPP (r = 0.684; P b .001) (Fig. 1).
3.3. Reliability of SFVO2 in predicting ROSC As shown in Fig. 2, SFVO2 results for the resuscitated and nonresuscitated animals barely overlapped. No animal achieved sustained ROSC with a SFVO2 of less than 25%. Only 1 of the nonresuscitated animals had an SFVO2 of 26% at a single time point (7 minutes after the start of ACLS). Meanwhile, the CPP results overlapped considerably between the resuscitated and nonresuscitated animals. Four of the nonresuscitated animals had a CPP value higher than the commonly quoted necessary threshold CPP of 15 mm Hg for ROSC on 7 occasions [4]. One animal could not achieve ROSC, although its CPP was greater than 15 mm Hg at all of the sampling time points (32.8, 35.0, and 38.9 mm Hg at 3, 7, and 11 minutes after the start of ACLS, respectively).
The performances of SFVO2 and CPP in predicting ROSC are displayed in Fig. 3. Femoral venous oxygen saturation significantly predicted ROSC with an AUC of 0.997 (95% confidence interval, 0.911-1.000; P b .001). The AUC for SFVO2 was larger than the AUC for CPP (AUC, 0.964; 95% confidence interval, 0.855-0.997; P b .001), although the difference in the AUC results was not statistically significant (P = .157). 4. Discussion In the present study, SFVO2 during ACLS showed a strong significant correlation with CPP and was a significant predictor of ROSC. Moreover, the AUC for SFVO2, although not statistically significant, was higher than the AUC for CPP. Based on these results, SFVO2 seems to be no less accurate or reliable than CPP during ACLS. Coronary perfusion pressure, ETCO2, and SCVO2 have been used as physiologic parameters for assessing the efficacy of CPR [1-8]. Coronary perfusion pressure monitoring, although highly reliable in determining the effectiveness of CPR, is rarely available in the clinical setting because CPP measurement requires placements of both aortic and central venous catheters. End-tidal carbon dioxide is an easily applicable and noninvasive parameter that has been shown to correlate well with cardiac output during CPR [13,25]. However, ETCO2 readings do not reflect tissue oxygenation, which is an important factor in successful resuscitation. Furthermore, medications commonly used during CPR can influence ETCO2 readings and, thus, the reliability of ETCO2 for assessing CPR efficacy [26-28]. Central venous oxygen saturation has advantages over ETCO2 in that it can reflect tissue oxygenation during CPR and is minimally affected by medication use [8,29]. However, it
Table 2 Coronary perfusion pressures and blood gases of the femoral venous blood samples obtained during ACLS
Coronary perfusion pressure (mm Hg) pH Venous PCO2 (mm Hg) Venous PO2 (mm Hg) Venous HCO3− (mmol/L) Venous base excess (mmol/L) Venous SO2 (%)
No ROSC (n = 31)
ROSC (n = 12)
P
8.7 (2.1-15.0) 6.979 ± 0.078 90.6 ± 11.4 16.0 (10.0-18.0) 15.5 ± 3.0 −10.1 ± 4.4 6.0 (4.0-12.3)
43.0 (38.0-53.6) 7.064 ± 0.046 73.3 ± 11.5 41.5 (30.5-52.0) 16.4 ± 1.8 −9.4 ± 3.2 52.0 (33.5-65.8)
b.001 b.001 b.001 b.001 .345 .609 b.001
Fig. 1. Scatterplot demonstrating the relationship between SFVO2 and CPP.
944
M.J. Kim et al. / American Journal of Emergency Medicine 33 (2015) 941–945
Fig. 2. Scatterplot of SFVO2 values (A) and CPPs (B) in the successfully resuscitated and nonresuscitated animals.
requires central venous access that is not easily accomplished without interfering with resuscitation interventions. Femoral catheter insertion is a quick, safe, and reliable method of obtaining venous access, even in patients undergoing CPR [16,17]. In a previous study that included 20 patients undergoing CPR, femoral venous catheters were successfully placed under ultrasound guidance in 90% of patients within approximately 2 minutes [16]. Given the results of our study and the relative ease of femoral venous catheter placement, we believe that SFVO2 can be a practical method for monitoring tissue oxygenation during CPR for OHCA patients after further confirmation through clinical studies. Venous oxygen saturation, as a parameter reflecting the balance between systemic oxygen delivery and consumption, has long been used to determine the adequacy of tissue oxygenation [30,31]. Ideally, venous oxygen saturation should be measured from mixed venous blood samples obtained from a pulmonary artery catheter. Because the placement of a pulmonary artery catheter is difficult and time consuming, SCVO2 has largely replaced mixed venous oxygen saturation. Several researchers have suggested that continuous SCVO2 monitoring is a reliable method for evaluating tissue oxygenation and predicting ROSC in the course of CPR [8,29]. On the other hand, there has been controversy over whether SFVO2 can be used as a surrogate for SCVO2[32-34]. Several studies on adult patients not in cardiac arrest
Fig. 3. Receiver operating characteristic curves for SFVO2 and CPP.
indicated that SFVO2 should not be used as a surrogate for SCVO2 due to the lack of agreement between the 2 measures [32,33]. In contrast, Emerman et al [34] compared blood gas values of pulmonary arterial, central venous, and femoral venous blood samples obtained simultaneously in dogs undergoing CPR and found no significant differences in oxygen tension among the 3 sites. The present study did not examine the relationship between SCVO2 and SFVO2 because we did not obtain central venous blood samples. Doing so would have carried the possibility of a falsely high CPP at the time of blood sampling because withdrawing blood from a central venous catheter lowers the RA pressure. Although our study cannot provide information on the interchangeability of SCVO2 and SFVO2 during CPR, the excellent predictive performance of SFVO2 observed in the present study suggests that SFVO2 can be used as a substitute for SCVO2 during CPR. In this study, CPP overlapped considerably between the resuscitated and nonresuscitated animals. In particular, 4 (36.4%) of the nonresuscitated animals could not be resuscitated, although they achieved CPP values higher than the commonly quoted threshold level [4]. The reasons why these animals could not achieve ROSC despite their seemingly adequate CPP values are not readily apparent. One possible explanation is that a selected CPP value at any single time point may not reflect the total CPP level and, hence, the perfusion to which the myocardium was exposed before that time point, which may be inadequate for achieving ROSC. Alternatively, it may also be attributed to the prolonged untreated cardiac arrest duration of the present study. A recent study suggested that higher CPP values may be required to achieve ROSC after a prolonged duration of untreated VF compared to shorter untreated arrest intervals [35]. On the other hand, there was minimal overlap in SFVO2 results between the resuscitated and nonresuscitated animals in this study. Accordingly, we postulate that SFVO2 reflects the total amount of perfusion over time better than a CPP value at a specific time point. A previous experimental study showed that venous blood gases exhibited gradual changes during the CPR period [36]. Although the AUC for predicting ROSC was comparable between CPP and SFVO2 in our study, we believe that SFVO2 may be more clinically useful because neither the CPP provided over time during CPR nor the required CPP threshold is available for CPR in the clinical setting. This study has several limitations. First, it was conducted on young and healthy pigs, and, therefore, the results should be verified in humans. Second, we used a pig model of prolonged VF. Ischemic contracture, which frequently occurs in prolonged cardiac arrest, compromises the hemodynamic effectiveness of CPR [37,38]. Previous
M.J. Kim et al. / American Journal of Emergency Medicine 33 (2015) 941–945
studies have indicated that the severity of ischemic contracture increases as the duration of untreated VF increases [37,38]. Thus, the duration of prolonged untreated VF in our study might have adversely affected CPP and SFVO2 values during CPR. Third, our study did not evaluate the effects of chest compression quality on SFVO2. Further studies are required to determine whether SFVO2 can reliably reflect chest compression quality over the course of CPR. Fourth, SFVO2 measurements were performed intermittently and not continuously, and further studies are required to evaluate the value of continuous SFVO2 monitoring. Fifth, our study did not examine ETCO2 levels during CPR, and, thus, it is unclear if SFVO2 can predict ROSC more reliably than ETCO2. Sixth, the animals in this study underwent mechanical ventilation with a relatively high tidal volume (15 mL/kg) during preparation of the experiment. Although several experimental studies in pigs have used this tidal volume [37,39], recent studies in pigs suggest that even a short-term high tidal volume during the course of mechanical ventilation can adversely affect outcomes by damaging lung parenchyma, reducing cardiac output, and inducing a systemic inflammatory reaction [40,41]. Seventh, RA pressure was maintained at 6 to 12 mm Hg before the induction of VF. However, RA pressures may not be normal or constant before the onset of cardiac arrest in the clinical setting. Thus, this intervention might have compromised the results of this study. 5. Conclusions This study aimed to evaluate the reliability of SFVO2 values obtained during CPR for predicting successful resuscitation. In a realistic pig model of OHCA, SFVO2 values obtained during CPR displayed a significant correlation with CPP and reliably predicted ROSC. Our study suggests that further studies in human cardiac arrest patients are warranted to evaluate the usefulness of SFVO2 monitoring during CPR in the clinical setting. References [1] Newmar RW, Otto CW, Link MS, Kronick SL, Shuster M, Callaway CW, et al. Part 8: adult advanced cardiovascular life support: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation 2010; 122:S729–67. [2] Niemann JT, Criley JM, Rosborough JP, Niskanen RA, Alferness C. Predictive indices of successful cardiac resuscitation after prolonged arrest and experimental cardiopulmonary resuscitation. Ann Emerg Med 1985;14:521–8. [3] Kern KB, Ewy GA, Voorhees WD, Babbs CF, Tacker WA. Myocardial perfusion pressure: a predictor of 24-hour survival during prolonged cardiac arrest in dogs. Resuscitation 1988;16:241–50. [4] Paradis NA, Martin GB, Rivers EP, Goetting MG, Appleton TJ, Feingold M, et al. Coronary perfusion pressure and the return of spontaneous circulation in human cardiopulmonary resuscitation. JAMA 1990;263:1106–13. [5] Sanders A, Atlas M, Ewy GA, Kern KB, Bragg S. Expired PCO2 as an index of coronary perfusion pressure. Am J Emerg Med 1985;3:147–9. [6] Falk JL, Rackow EC, Weil MH. End-tidal carbon dioxide concentration during cardiopulmonary resuscitation. N Engl J Med 1988;318:607–11. [7] Grmec S, Klemen P. Does the end-tidal carbon dioxide (EtCO2) concentration have prognostic value during out-of-hospital cardiac arrest? Eur J Emerg Med 2001;8: 263–9. [8] Rivers EP, Martin GB, Smithline H, Rady MY, Schultz CH, Goetting MG, et al. The clinical implications of continuous central venous oxygen saturation during human CPR. Ann Emerg Med 1992;21:1094–101. [9] Friess SH, Sutton RM, Bhalala U, Maltese MR, Naim MY, Bratinov G, et al. Hemodynamic directed cardiopulmonary resuscitation improves short-term survival from ventricular fibrillation cardiac arrest. Crit Care Med 2013;41:2698–704. [10] Friess SH, Sutton RM, French B, Bhalala U, Maltese MR, Naim MY, et al. Hemodynamic directed CPR improves cerebral perfusion pressure and brain tissue oxygenation. Resuscitation 2014;85:1298–303. [11] Hamrick JL, Hamrick JT, Lee JK, Lee BH, Koehler RC, Shaffner DH. Efficacy of chest compressions directed by end-tidal CO2 feedback in a pediatric resuscitation model of basic life support. J Am Heart Assoc 2014;3:e000450. [12] Lindner KH, Prengel AW, Pfenninger EG, Lindner IM, Strohmenger HU, Georgieff M, et al. Vasopressin improves vital organ blood flow during closed-chest cardiopulmonary resuscitation in pigs. Circulation 1995;91:215–21. [13] Gudipati CV, Weil MH, Bisera J, Deshmukh HG, Rackow EC. Expired carbon dioxide: a noninvasive monitor of cardiopulmonary resuscitation. Circulation 1988;77:234–9.
945
[14] Snyder AB, Salloum LJ, Barone JE, Conley M, Todd M, DiGiacomo JC. Predicting shortterm outcome of cardiopulmonary resuscitation using central venous oxygen tension measurements. Crit Care Med 1991;19:111–3. [15] Kim Y, Park H. Interruption of chest compression for central venous catheterization during cardiopulmonary resuscitation. Korean J Crit Care Med 2014;29:172–6. [16] Hilty WM, Hudson PA, Levitt MA, Hall JB. Real-time ultrasound-guided femoral vein catheterization during cardiopulmonary resuscitation. Ann Emerg Med 1997;29: 331–6. [17] Swanson RS, Uhlig PN, Gross PL, McCabe CJ. Emergency intravenous access through the femoral vein. Ann Emerg Med 1984;13:244–7. [18] Jeung KW, Ryu HH, Song KH, Lee BK, Lee HY, Heo T, et al. Variable effects of highdose adrenaline relative to standard-dose adrenaline on resuscitation outcomes according to cardiac arrest duration. Resuscitation 2011;82:932–6. [19] Lee HY, Lee BK, Jeung KW, Lee SM, Jung YH, Lee GS, et al. Potassium induced cardiac stand still during conventional cardiopulmonary resuscitation in a pig model of prolonged ventricular fibrillation cardiac arrest: a feasibility study. Resuscitation 2013;84:378–83. [20] Lee BK, Jeung KW, Choi SS, Park SW, Yun SW, Lee SM, et al. Effects of the administration of 2,3-butanedione monoxime during conventional cardiopulmonary resuscitation on ischaemic contracture and resuscitability in a pig model of out-of-hospital cardiac arrest. Resuscitation 2015;87:26–32. [21] Cho YC, Cho SW, Chung SP, Yu K, Kwon OY, Kim SW. How can a single rescuer adequately deliver tidal volume with a manual resuscitator? An improved device for delivering regular tidal volume. Emerg Med J 2011;28:40–3. [22] Kralovich KA, Morris DC, Dereczyk BE, Simonetti V, Williams M, Rivers EP, et al. Hemodynamic effects of aortic occlusion during hemorrhagic shock and cardiac arrest. J Trauma 1997;42:1023–8. [23] Huo TT, Zeng Y, Liu XN, Sun L, Han HZ, Chen HG, et al. Hydrogen-rich saline improves survival and neurological outcome after cardiac arrest and cardiopulmonary resuscitation in rats. Anesth Analg 2014;119:368–80. [24] DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988;44:837–45. [25] Weil MH, Bisera J, Trevino RP, Rackow EC. Cardiac output and end-tidal carbon dioxide. Crit Care Med 1985;13:907–9. [26] Cantineau JP, Merckx P, Lambert Y, Sorkine M, Bertrand C, Duvaldestin P. Effect of epinephrine on end-tidal carbon dioxide pressure during prehospital cardiopulmonary resuscitation. Am J Emerg Med 1994;12:267–70. [27] Lindberg L, Liao Q, Steen S. The effects of epinephrine/norepinephrine on end-tidal carbon dioxide concentration, coronary perfusion pressure and pulmonary arterial blood flow during cardiopulmonary resuscitation. Resuscitation 2000;43:129–40. [28] Okamoto H, Hoka S, Kawasaki T, Okuyama T, Takahashi S. Changes in end-tidal carbon dioxide tension following sodium bicarbonate administration: correlation with cardiac output and haemoglobin concentration. Acta Anaesthesiol Scand 1995;39:79–84. [29] Nakazawa K, Hikawa Y, Saitoh Y, Tanaka N, Yasuda K, Amaha K. Usefulness of central venous oxygen saturation monitoring during cardiopulmonary resuscitation. Intensive Care Med 1994;20:450–1. [30] Baele PL, McMichan JC, Marsh HM, Sill JC, Southorn PA. Continuous monitoring of mixed venous oxygen saturation in critically ill patients. Anesth Analg 1982;61: 513–7. [31] Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, et al. Early goaldirected therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345:1368–77. [32] Davison DL, Chawla LS, Selassie L, Jones EM, McHone KC, Vota AR, et al. Femoral-based central venous oxygen saturation is not a reliable substitute for subclavian/internal jugular-based central venous oxygen saturation in patients who are critically ill. Chest 2010;138:76–83. [33] van Beest PA, van der Schors A, Liefers H, Coenen LG, Braam RL, Habib N, et al. Femoral venous oxygen saturation is no surrogate for central venous oxygen saturation. Crit Care Med 2012;40:3196–201. [34] Emerman CL, Pinchak AC, Hagen JF, Hancock D. A comparison of venous blood gases during cardiac arrest. Am J Emerg Med 1988;6:580–3. [35] Reynolds JC, Salcido DD, Menegazzi JJ. Coronary perfusion pressure and return of spontaneous circulation after prolonged cardiac arrest. Prehosp Emerg Care 2010; 14:78–84. [36] Ralston SH, Voorhees WD, Showen L, Schmitz P, Kougias C, Tacker WA. Venous and arterial blood gases during and after cardiopulmonary resuscitation in dogs. Am J Emerg Med 1985;3:132–6. [37] Klouche K, Weil MH, Sun S, Tang W, Povoas HP, Kamohara T, et al. Evolution of the stone heart after prolonged cardiac arrest. Chest 2002;122:1006–11. [38] Klouche K, Weil MH, Sun S, Tang W, Kamohara T. Echo-Doppler observations during cardiac arrest and cardiopulmonary resuscitation. Crit Care Med 2000; 28:N212–3. [39] Gong P, Hua R, Zhang Y, Zhao H, Tang Z, Mei X, et al. Hypothermia-induced neuroprotection is associated with reduced mitochondrial membrane permeability in a swine model of cardiac arrest. J Cereb Blood Flow Metab 2013;33:928–34. [40] Kobr J, Kuntscher V, Treska V, Molacek J, Vobruba V, Fremuth J, et al. Adverse effects of the high tidal volume during mechanical ventilation of normal lung in pigs. Bratisl Lek Listy 2008;109:45–51. [41] Vobruba V, Klimenko OV, Kobr J, Cerna O, Pokorna P, Mikula I, et al. Effects of high tidal volume mechanical ventilation on production of cytokines, iNOS, and MIP-1β proteins in pigs. Exp Lung Res 2013;39:1–8.
Contents lists available at ScienceDirect
American Journal of Emergency Medicine journal homepage: www.elsevier.com/locate/ajem
Original Contribution
Femoral venous oxygen saturation obtained during CPR predicts successful resuscitation in a pig model☆ Mu Jin Kim, MD a, Kyung Woon Jeung, MD, PhD b, Byung Kook Lee, MD, PhD b, Sung Soo Choi, EMTP c, Sang Wook Park, EMTP b, Kyung Hwan Song, MD b, Sung Min Lee, MD b, Yong Il Min, MD, PhD b,⁎ a b c
Department of Emergency Medicine, Songjeong Sarang Hospital, Gwangju, Republic of Korea Department of Emergency Medicine, Chonnam National University Hospital, Gwangju, Republic of Korea Department of Emergency Medical Service, Howon University, Gunsan, Republic of Korea
a r t i c l e
i n f o
Article history: Received 15 January 2015 Received in revised form 16 March 2015 Accepted 6 April 2015
a b s t r a c t Purpose: Central venous oxygen saturation has been shown to reflect the adequacy of tissue oxygenation during cardiopulmonary resuscitation (CPR), thereby enabling the assessment of CPR quality and the prediction of restoration of spontaneous circulation (ROSC). The femoral vein can be easily accessed during CPR. We determined if femoral venous oxygen saturation (SFVO2) values obtained during CPR could reliably predict ROSC in a pig model. Methods: After 15 minutes of untreated ventricular fibrillation followed by 8 minutes of basic life support, 19 pigs underwent advanced cardiovascular life support. During advanced cardiovascular life support, femoral venous blood samples were obtained at 4-minute intervals. The abilities of SFVO2 and coronary perfusion pressure (CPP) to predict ROSC were evaluated by calculating the areas under receiver operating characteristic curves (AUCs). Results: Eight pigs (42.1%) achieved ROSC. The resuscitated animals had significantly higher CPP (P b .001) and SFVO2 (P b .001) values than the nonresuscitated animals, and there was a significant correlation between SFVO2 and CPP values (r = 0.684; P b .001). The CPPs of the resuscitated and nonresuscitated animals overlapped considerably; however, there was minimal overlap between the 2 groups for SFVO2. Femoral venous oxygen saturation significantly predicted ROSC with an AUC of 0.997 (95% confidence interval, 0.911-1.000; P b .001), and it had a larger AUC than CPP (AUC, 0.964; 95% confidence interval, 0.855-0.997; P b .001). The AUC difference, however, was not statistically significant (P = .157). Conclusion: In this study, SFVO2 values obtained during CPR exhibited a significant correlation with CPP and reliably predicted ROSC. © 2015 Elsevier Inc. All rights reserved.
1. Introduction In recent years, there has been an increased emphasis on physiologic monitoring during cardiopulmonary resuscitation (CPR) [1]. Several studies have examined coronary perfusion pressure (CPP), end-tidal carbon dioxide (ETCO2), and central venous oxygen saturation (SCVO2) as means for monitoring the effectiveness of CPR, and recent resuscitation guidelines recommended the use of these parameters, when feasible, to optimize chest compressions and guide vasopressor therapy during cardiac arrest [1-8]. Moreover, recent studies have suggested that titrating resuscitative efforts according to these parameters improves resuscitation outcomes [9-11].
☆ Funding sources/disclosures: This research was supported by Basic Science Research Program through the National Research Foundation of Korea funded by the Ministry of Education, Science and Technology (NRF-2012R1A1A2039466). The funder had no role in the study design, data collection, analysis, decision to publish, or the preparation of the manuscript. ⁎ Corresponding author. Tel.: +82 62 220 6809; fax: +82 62 228 7417. E-mail address: [email protected] (Y.I. Min). http://dx.doi.org/10.1016/j.ajem.2015.04.004 0735-6757/© 2015 Elsevier Inc. All rights reserved.
The key to successful resuscitation is tissue oxygen delivery sufficient to meet the cellular demands of ischemic vital organs. Coronary perfusion pressure and ETCO2 correlate with blood flow generated during CPR [12,13]. However, these variables do not provide information about the relationship between systemic oxygen delivery and demand. SCVO2, which is used as a surrogate for mixed venous oxygen saturation, is different from ETCO2 and CPP in that it reflects the balance between systemic oxygen delivery and demand during CPR. Central venous oxygen saturation has been shown to reflect the adequacy of tissue oxygenation during CPR, thus enabling the assessment of CPR quality and the prediction of restoration of spontaneous circulation (ROSC) [8,14]. In a study that included 100 out-of-hospital cardiac arrest (OHCA) patients, patients who achieved ROSC were found to have higher SCVO2 values than those who could not be resuscitated [8]. No patient in that study experienced ROSC without reaching a SCVO2 of at least 30%. The measurement of SCVO2 requires a catheter to be placed in the vena cava near the right atrium (RA), which is usually inserted through the internal jugular or subclavian vein. However, placement of a central venous catheter through the internal jugular or subclavian vein is usually impractical during CPR because it interferes with resuscitation interventions, including chest compressions and airway
942
M.J. Kim et al. / American Journal of Emergency Medicine 33 (2015) 941–945
management [15]. Currently, the use of SCVO2 during CPR is limited to patients who have a central venous catheter in place before cardiac arrest. The femoral vein can be easily accessed during CPR, and establishing femoral venous access does not interfere with resuscitation interventions. Previous studies have shown that femoral venous cannulation can be performed quickly and safely with a high rate of success, even during CPR [16,17]. Therefore, if oxygen saturation of venous blood obtained from the femoral vein can be used with confidence as a substitute for SCVO2 during CPR, then the use of venous oxygen saturation to assess CPR quality and predict ROSC may be able to be extended to cardiac arrest patients who do not have a central venous catheter in place before cardiac arrest (which is the case for most OHCA patients). However, to our knowledge, the suitability of femoral venous oxygen saturation (SFV O 2) as a substitute for SCVO2 has not yet been investigated. In this study, we investigated the reliability of SFVO2 in predicting ROSC using a realistic model of OHCA. We hypothesized that SFVO2 values obtained during CPR can reliably predict ROSC. 2. Methods The Animal Care and Use Committee of Chonnam National University approved the protocol of this study. All animal care and experiments were conducted according to the Institutional Animal Care and Use Committee guidelines. 2.1. Animal preparation Nineteen male domestic pigs weighing 25.0 kg (22.5-29.7) were used. The animal preparation was performed as previously described [18-20]. After premedication (ketamine 20 mg/kg, xylazine 2.2 mg/kg, atropine 0.04 mg/kg, administered intramuscularly), anesthesia was induced with 50%:50% N2O:O2 and 2% to 5% sevoflurane via a mask. After tracheal intubation, the pigs were ventilated at a tidal volume of 15 mL/kg. Anesthesia was continued with 70%:30% N2O:O2 and 0.5% to 2% sevoflurane, which was titrated to prevent signs of pain (reactive wide pupils, tachycardia, and hypertension). In our experience [18-20], the experimental preparation was completed in a very short time (approximately 40 minutes), and neither hemodynamic instabilities nor signs of pain were observed with this anesthesia protocol. The ventilatory rates were adjusted to achieve normocapnia. A catheter was inserted into an ear vein, and saline was administered to maintain an RA pressure of 6 to 12 mm Hg. A double-lumen catheter was advanced from the right femoral artery to the thoracic aorta to monitor aortic pressure and to obtain blood samples. A 6F introducer sheath was inserted into the right femoral vein and advanced 3 cm under direct visualization to obtain femoral venous blood samples. The right external jugular vein was cannulated with an introducer sheath to monitor the RA pressure and to insert a right ventricle pacing catheter. Electrocardiogram leads were placed on the limbs to monitor the heart rhythm, and the rectal temperature was maintained between 37°C and 38°C.
8/min with high-flow oxygen (10 L/min) through the use of a volumemarked bag devised by Cho et al [21]. At the commencement of ACLS, all animals received 0.5 U/kg of vasopressin intravenously. After 4 minutes of ACLS, 0.02 mg/kg of epinephrine was administered every 3 minutes as needed. During ACLS, defibrillation was attempted using a single biphasic 150-J electric shock at 2-minute intervals. Sustained ROSC was defined as maintenance of a systolic aortic pressure of at least 60 mm Hg for at least 5 consecutive minutes [22,23]. If ROSC was not achieved within 12 minutes of ACLS, the resuscitation efforts were discontinued. The successfully resuscitated animals were euthanized using potassium chloride. Autopsies were performed to document any injuries to the thoracic and abdominal cavities caused by CPR. 2.3. Measurements An investigator obtained 2 mL of blood from the introducer sheath inserted into the femoral vein using a BD Preset syringe (Becton, Dickinson and Company, Plymouth, UK) immediately before the induction of VF, 3 minutes after the start of ACLS, and at 4-minute intervals thereafter until ROSC was achieved or until the end of the 12-minute ACLS period. All blood samples were immediately analyzed using a blood gas/electrolyte analyzer (GEM Premier 3000; Instrumentation Laboratory, Lexington, KY). The aortic pressure, RA pressure, and standard lead II electrocardiogram were continuously monitored (CS/3 CCM; Datex-Ohmeda, Helsinki, Finland) and transferred to a personal computer using S/5 Collect software (Datex-Ohmeda). The CPP at the time of femoral venous blood sampling was calculated by subtracting the RA end-diastolic pressure from the simultaneous aortic enddiastolic pressure; it was averaged over 5 compression cycles. 2.4. Statistical analysis The sample size calculation was based on a previous study [8] and resulted in a minimum requirement of 7 animals in each group (ROSC vs no ROSC), with α set at .05 and a power of 0.90. Assuming an ROSC rate of 40% based on our previous studies [18-20], 19 animals were used for this study. Continuous variables were tested for normality using the Shapiro-Wilk test. Normally distributed variables were analyzed using independent t tests and summarized as means ± SDs, whereas nonnormally distributed variables were analyzed using Mann-Whitney U tests and summarized as medians with interquartile ranges. Comparisons of categorical variables were performed using χ2 or Fisher exact test, as indicated. Correlations between nonnormally distributed continuous variables were assessed using Spearman rank correlation coefficients. Receiver operating characteristic (ROC) analyses were performed to examine the abilities of SFVO2 and CPP to predict ROSC. Comparisons of the areas under the ROC curves (AUCs) were performed as recommended by DeLong et al [24]. Data analyses were performed using PASW/SPSS software, version 18 (IBM, Inc, Chicago, IL). The ROC curves were calculated and compared using MedCalc version 14.10.2 (free trial; MedCalc Software, Mariakerke, Belgium). Statistical significance was set at P b .05.
2.2. Cardiac arrest and CPR 3. Results Ventricular fibrillation (VF) was induced by applying an electrical current (60 Hz, 30 mA alternating current) via a right ventricle pacing catheter. After 15 minutes of untreated VF, basic life support using cycles of 30 chest compressions followed by 2 ventilations with ambient air was started. Closed chest compressions were administered by 2 investigators (SSC and SML) for all of the animals. A HeartStart MRx Monitor/ Defibrillator (Philips Medical Systems, Seattle, WA) with Q-CPR assisted in maintaining a chest compression rate of 100/min and a compression depth of 25% of the anterior-posterior diameter of the chest wall. After 8 minutes of basic life support, advanced cardiovascular life support (ACLS) was initiated according to current resuscitation guidelines [1]. Asynchronous positive pressure ventilations were provided at a rate of
3.1. Baseline characteristics Precardiac arrest baseline characteristics are displayed in Table 1. There were no significant differences between the groups in terms of the baseline measurements. 3.2. Coronary perfusion pressure and blood gases obtained from the femoral vein during CPR Of the 19 animals studied, 8 (42.1%) achieved ROSC after 6.0 minutes (6.0-7.5 minutes) of ACLS. Twelve femoral venous blood samples (8 at
M.J. Kim et al. / American Journal of Emergency Medicine 33 (2015) 941–945
943
Table 1 Prearrest baseline characteristics
Weight (kg) Male sex, n (%) Systolic aortic pressure (mm Hg) Diastolic aortic pressure (mm Hg) Mean aortic pressure (mm Hg) Systolic RA pressure (mm Hg) Diastolic RA pressure (mm Hg) Mean RA pressure (mm Hg) Arterial pH Arterial PCO2 (mm Hg) Arterial PO2 (mm Hg) Arterial HCO3− (mmol/L) Arterial base excess (mmol/L) Arterial SO2 (%) Venous pH Venous PCO2 (mm Hg) Venous PO2 (mm Hg) Venous HCO3− (mmol/L) Venous base excess (mmol/L) Venous SO2 (%)
Total (n = 19)
No ROSC (n = 11)
ROSC (n = 8)
P
25.0 (22.5-29.7) 13 (68.4) 117 (111-125) 77 (71-87) 93 (88-102) 10.0 (9.0-12.5) 6.0 (5.0-7.5) 8.0 (7.0-9.5) 7.500 (7.460-7.525) 37.0 (36.0-40.0) 140.0 (99.5-148.5) 30.3 (28.1-31.2) 7.3 (4.4-8.6) 99.0 (98.0-99.0) 7.430 (7.395-7.460) 49.0 (46.5-51.0) 34.0 (30.5-39.0) 30.0 (27.9-31.2) 7.7 (5.1-9.8) 67.0 (60.0-78.0)
25.0 (22.7-31.6) 8 (72.7) 117 (111-129) 79 (73-86) 93 (88-102) 11.0 (9.5-12.5) 6.0 (5.5-7.5) 8.0 (7.5-9.5) 7.500 (7.465-7.515) 37.0 (36.0-39.5) 140.0 (111.5-147.5) 30.3 (28.1-30.8) 7.3 (4.4-8.2) 99.0 (98.5-99.0) 7.420 (7.400-7.455) 49.0 (46.5-51.0) 33.0 (29.0-39.5) 29.7 (27.9-31.2) 7.3 (5.1-9.8) 64.0 (53.5-75.0)
24.8 (22.6-26.8) 5 (62.5) 117 (111-123) 75 (67-89) 96 (86-101) 9.5 (8.8-12.0) 5.5 (5.0-6.8) 7.5 (7.0-9.0) 7.510 (7.445-7.538) 39.5 (36.3-40.3) 123.0 (96.8-152.8) 30.3 (27.7-31.9) 7.1 (4.0-9.4) 98.5 (97.8-99.3) 7.450 (7.388-7.473) 47.5 (46.8-50.3) 36.5 (32.3-39.0) 30.4 (27.5-31.6) 8.3 (5.2-9.9) 71.5 (62.0-78.0)
.836 1.000 .836 .563 .869 .505 .767 .674 .804 .588 .836 .679 .741 .692 .454 .618 .535 .836 .804 .619
3 minutes, 2 at 7 minutes, and 2 at 11 minutes after the start of ACLS) were obtained during ACLS in the successfully resuscitated animals. In the 11 animals that did not achieve sustained ROSC, 3 femoral venous blood samples were obtained from each (at 3, 7, and 11 minutes after the start of ACLS, respectively), except for 1 animal in which a blood sample was only obtained at 3 minutes after the start of ACLS. This animal achieved ROSC for 2 minutes at 4 minutes after the start of ACLS and thereafter could not achieve sustained ROSC. Thus, a total of 31 femoral venous blood samples were obtained during ACLS in the nonresuscitated animals. The results for CPP (at the time of femoral venous blood sampling) and blood gases from the femoral venous blood samples obtained during ACLS are displayed in Table 2. The resuscitated animals had significantly higher CPP (P b .001) and SFVO2 (P b .001) values than the nonresuscitated animals. There was a significant positive correlation between SFVO2 and CPP (r = 0.684; P b .001) (Fig. 1).
3.3. Reliability of SFVO2 in predicting ROSC As shown in Fig. 2, SFVO2 results for the resuscitated and nonresuscitated animals barely overlapped. No animal achieved sustained ROSC with a SFVO2 of less than 25%. Only 1 of the nonresuscitated animals had an SFVO2 of 26% at a single time point (7 minutes after the start of ACLS). Meanwhile, the CPP results overlapped considerably between the resuscitated and nonresuscitated animals. Four of the nonresuscitated animals had a CPP value higher than the commonly quoted necessary threshold CPP of 15 mm Hg for ROSC on 7 occasions [4]. One animal could not achieve ROSC, although its CPP was greater than 15 mm Hg at all of the sampling time points (32.8, 35.0, and 38.9 mm Hg at 3, 7, and 11 minutes after the start of ACLS, respectively).
The performances of SFVO2 and CPP in predicting ROSC are displayed in Fig. 3. Femoral venous oxygen saturation significantly predicted ROSC with an AUC of 0.997 (95% confidence interval, 0.911-1.000; P b .001). The AUC for SFVO2 was larger than the AUC for CPP (AUC, 0.964; 95% confidence interval, 0.855-0.997; P b .001), although the difference in the AUC results was not statistically significant (P = .157). 4. Discussion In the present study, SFVO2 during ACLS showed a strong significant correlation with CPP and was a significant predictor of ROSC. Moreover, the AUC for SFVO2, although not statistically significant, was higher than the AUC for CPP. Based on these results, SFVO2 seems to be no less accurate or reliable than CPP during ACLS. Coronary perfusion pressure, ETCO2, and SCVO2 have been used as physiologic parameters for assessing the efficacy of CPR [1-8]. Coronary perfusion pressure monitoring, although highly reliable in determining the effectiveness of CPR, is rarely available in the clinical setting because CPP measurement requires placements of both aortic and central venous catheters. End-tidal carbon dioxide is an easily applicable and noninvasive parameter that has been shown to correlate well with cardiac output during CPR [13,25]. However, ETCO2 readings do not reflect tissue oxygenation, which is an important factor in successful resuscitation. Furthermore, medications commonly used during CPR can influence ETCO2 readings and, thus, the reliability of ETCO2 for assessing CPR efficacy [26-28]. Central venous oxygen saturation has advantages over ETCO2 in that it can reflect tissue oxygenation during CPR and is minimally affected by medication use [8,29]. However, it
Table 2 Coronary perfusion pressures and blood gases of the femoral venous blood samples obtained during ACLS
Coronary perfusion pressure (mm Hg) pH Venous PCO2 (mm Hg) Venous PO2 (mm Hg) Venous HCO3− (mmol/L) Venous base excess (mmol/L) Venous SO2 (%)
No ROSC (n = 31)
ROSC (n = 12)
P
8.7 (2.1-15.0) 6.979 ± 0.078 90.6 ± 11.4 16.0 (10.0-18.0) 15.5 ± 3.0 −10.1 ± 4.4 6.0 (4.0-12.3)
43.0 (38.0-53.6) 7.064 ± 0.046 73.3 ± 11.5 41.5 (30.5-52.0) 16.4 ± 1.8 −9.4 ± 3.2 52.0 (33.5-65.8)
b.001 b.001 b.001 b.001 .345 .609 b.001
Fig. 1. Scatterplot demonstrating the relationship between SFVO2 and CPP.
944
M.J. Kim et al. / American Journal of Emergency Medicine 33 (2015) 941–945
Fig. 2. Scatterplot of SFVO2 values (A) and CPPs (B) in the successfully resuscitated and nonresuscitated animals.
requires central venous access that is not easily accomplished without interfering with resuscitation interventions. Femoral catheter insertion is a quick, safe, and reliable method of obtaining venous access, even in patients undergoing CPR [16,17]. In a previous study that included 20 patients undergoing CPR, femoral venous catheters were successfully placed under ultrasound guidance in 90% of patients within approximately 2 minutes [16]. Given the results of our study and the relative ease of femoral venous catheter placement, we believe that SFVO2 can be a practical method for monitoring tissue oxygenation during CPR for OHCA patients after further confirmation through clinical studies. Venous oxygen saturation, as a parameter reflecting the balance between systemic oxygen delivery and consumption, has long been used to determine the adequacy of tissue oxygenation [30,31]. Ideally, venous oxygen saturation should be measured from mixed venous blood samples obtained from a pulmonary artery catheter. Because the placement of a pulmonary artery catheter is difficult and time consuming, SCVO2 has largely replaced mixed venous oxygen saturation. Several researchers have suggested that continuous SCVO2 monitoring is a reliable method for evaluating tissue oxygenation and predicting ROSC in the course of CPR [8,29]. On the other hand, there has been controversy over whether SFVO2 can be used as a surrogate for SCVO2[32-34]. Several studies on adult patients not in cardiac arrest
Fig. 3. Receiver operating characteristic curves for SFVO2 and CPP.
indicated that SFVO2 should not be used as a surrogate for SCVO2 due to the lack of agreement between the 2 measures [32,33]. In contrast, Emerman et al [34] compared blood gas values of pulmonary arterial, central venous, and femoral venous blood samples obtained simultaneously in dogs undergoing CPR and found no significant differences in oxygen tension among the 3 sites. The present study did not examine the relationship between SCVO2 and SFVO2 because we did not obtain central venous blood samples. Doing so would have carried the possibility of a falsely high CPP at the time of blood sampling because withdrawing blood from a central venous catheter lowers the RA pressure. Although our study cannot provide information on the interchangeability of SCVO2 and SFVO2 during CPR, the excellent predictive performance of SFVO2 observed in the present study suggests that SFVO2 can be used as a substitute for SCVO2 during CPR. In this study, CPP overlapped considerably between the resuscitated and nonresuscitated animals. In particular, 4 (36.4%) of the nonresuscitated animals could not be resuscitated, although they achieved CPP values higher than the commonly quoted threshold level [4]. The reasons why these animals could not achieve ROSC despite their seemingly adequate CPP values are not readily apparent. One possible explanation is that a selected CPP value at any single time point may not reflect the total CPP level and, hence, the perfusion to which the myocardium was exposed before that time point, which may be inadequate for achieving ROSC. Alternatively, it may also be attributed to the prolonged untreated cardiac arrest duration of the present study. A recent study suggested that higher CPP values may be required to achieve ROSC after a prolonged duration of untreated VF compared to shorter untreated arrest intervals [35]. On the other hand, there was minimal overlap in SFVO2 results between the resuscitated and nonresuscitated animals in this study. Accordingly, we postulate that SFVO2 reflects the total amount of perfusion over time better than a CPP value at a specific time point. A previous experimental study showed that venous blood gases exhibited gradual changes during the CPR period [36]. Although the AUC for predicting ROSC was comparable between CPP and SFVO2 in our study, we believe that SFVO2 may be more clinically useful because neither the CPP provided over time during CPR nor the required CPP threshold is available for CPR in the clinical setting. This study has several limitations. First, it was conducted on young and healthy pigs, and, therefore, the results should be verified in humans. Second, we used a pig model of prolonged VF. Ischemic contracture, which frequently occurs in prolonged cardiac arrest, compromises the hemodynamic effectiveness of CPR [37,38]. Previous
M.J. Kim et al. / American Journal of Emergency Medicine 33 (2015) 941–945
studies have indicated that the severity of ischemic contracture increases as the duration of untreated VF increases [37,38]. Thus, the duration of prolonged untreated VF in our study might have adversely affected CPP and SFVO2 values during CPR. Third, our study did not evaluate the effects of chest compression quality on SFVO2. Further studies are required to determine whether SFVO2 can reliably reflect chest compression quality over the course of CPR. Fourth, SFVO2 measurements were performed intermittently and not continuously, and further studies are required to evaluate the value of continuous SFVO2 monitoring. Fifth, our study did not examine ETCO2 levels during CPR, and, thus, it is unclear if SFVO2 can predict ROSC more reliably than ETCO2. Sixth, the animals in this study underwent mechanical ventilation with a relatively high tidal volume (15 mL/kg) during preparation of the experiment. Although several experimental studies in pigs have used this tidal volume [37,39], recent studies in pigs suggest that even a short-term high tidal volume during the course of mechanical ventilation can adversely affect outcomes by damaging lung parenchyma, reducing cardiac output, and inducing a systemic inflammatory reaction [40,41]. Seventh, RA pressure was maintained at 6 to 12 mm Hg before the induction of VF. However, RA pressures may not be normal or constant before the onset of cardiac arrest in the clinical setting. Thus, this intervention might have compromised the results of this study. 5. Conclusions This study aimed to evaluate the reliability of SFVO2 values obtained during CPR for predicting successful resuscitation. In a realistic pig model of OHCA, SFVO2 values obtained during CPR displayed a significant correlation with CPP and reliably predicted ROSC. Our study suggests that further studies in human cardiac arrest patients are warranted to evaluate the usefulness of SFVO2 monitoring during CPR in the clinical setting. References [1] Newmar RW, Otto CW, Link MS, Kronick SL, Shuster M, Callaway CW, et al. Part 8: adult advanced cardiovascular life support: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation 2010; 122:S729–67. [2] Niemann JT, Criley JM, Rosborough JP, Niskanen RA, Alferness C. Predictive indices of successful cardiac resuscitation after prolonged arrest and experimental cardiopulmonary resuscitation. Ann Emerg Med 1985;14:521–8. [3] Kern KB, Ewy GA, Voorhees WD, Babbs CF, Tacker WA. Myocardial perfusion pressure: a predictor of 24-hour survival during prolonged cardiac arrest in dogs. Resuscitation 1988;16:241–50. [4] Paradis NA, Martin GB, Rivers EP, Goetting MG, Appleton TJ, Feingold M, et al. Coronary perfusion pressure and the return of spontaneous circulation in human cardiopulmonary resuscitation. JAMA 1990;263:1106–13. [5] Sanders A, Atlas M, Ewy GA, Kern KB, Bragg S. Expired PCO2 as an index of coronary perfusion pressure. Am J Emerg Med 1985;3:147–9. [6] Falk JL, Rackow EC, Weil MH. End-tidal carbon dioxide concentration during cardiopulmonary resuscitation. N Engl J Med 1988;318:607–11. [7] Grmec S, Klemen P. Does the end-tidal carbon dioxide (EtCO2) concentration have prognostic value during out-of-hospital cardiac arrest? Eur J Emerg Med 2001;8: 263–9. [8] Rivers EP, Martin GB, Smithline H, Rady MY, Schultz CH, Goetting MG, et al. The clinical implications of continuous central venous oxygen saturation during human CPR. Ann Emerg Med 1992;21:1094–101. [9] Friess SH, Sutton RM, Bhalala U, Maltese MR, Naim MY, Bratinov G, et al. Hemodynamic directed cardiopulmonary resuscitation improves short-term survival from ventricular fibrillation cardiac arrest. Crit Care Med 2013;41:2698–704. [10] Friess SH, Sutton RM, French B, Bhalala U, Maltese MR, Naim MY, et al. Hemodynamic directed CPR improves cerebral perfusion pressure and brain tissue oxygenation. Resuscitation 2014;85:1298–303. [11] Hamrick JL, Hamrick JT, Lee JK, Lee BH, Koehler RC, Shaffner DH. Efficacy of chest compressions directed by end-tidal CO2 feedback in a pediatric resuscitation model of basic life support. J Am Heart Assoc 2014;3:e000450. [12] Lindner KH, Prengel AW, Pfenninger EG, Lindner IM, Strohmenger HU, Georgieff M, et al. Vasopressin improves vital organ blood flow during closed-chest cardiopulmonary resuscitation in pigs. Circulation 1995;91:215–21. [13] Gudipati CV, Weil MH, Bisera J, Deshmukh HG, Rackow EC. Expired carbon dioxide: a noninvasive monitor of cardiopulmonary resuscitation. Circulation 1988;77:234–9.
945
[14] Snyder AB, Salloum LJ, Barone JE, Conley M, Todd M, DiGiacomo JC. Predicting shortterm outcome of cardiopulmonary resuscitation using central venous oxygen tension measurements. Crit Care Med 1991;19:111–3. [15] Kim Y, Park H. Interruption of chest compression for central venous catheterization during cardiopulmonary resuscitation. Korean J Crit Care Med 2014;29:172–6. [16] Hilty WM, Hudson PA, Levitt MA, Hall JB. Real-time ultrasound-guided femoral vein catheterization during cardiopulmonary resuscitation. Ann Emerg Med 1997;29: 331–6. [17] Swanson RS, Uhlig PN, Gross PL, McCabe CJ. Emergency intravenous access through the femoral vein. Ann Emerg Med 1984;13:244–7. [18] Jeung KW, Ryu HH, Song KH, Lee BK, Lee HY, Heo T, et al. Variable effects of highdose adrenaline relative to standard-dose adrenaline on resuscitation outcomes according to cardiac arrest duration. Resuscitation 2011;82:932–6. [19] Lee HY, Lee BK, Jeung KW, Lee SM, Jung YH, Lee GS, et al. Potassium induced cardiac stand still during conventional cardiopulmonary resuscitation in a pig model of prolonged ventricular fibrillation cardiac arrest: a feasibility study. Resuscitation 2013;84:378–83. [20] Lee BK, Jeung KW, Choi SS, Park SW, Yun SW, Lee SM, et al. Effects of the administration of 2,3-butanedione monoxime during conventional cardiopulmonary resuscitation on ischaemic contracture and resuscitability in a pig model of out-of-hospital cardiac arrest. Resuscitation 2015;87:26–32. [21] Cho YC, Cho SW, Chung SP, Yu K, Kwon OY, Kim SW. How can a single rescuer adequately deliver tidal volume with a manual resuscitator? An improved device for delivering regular tidal volume. Emerg Med J 2011;28:40–3. [22] Kralovich KA, Morris DC, Dereczyk BE, Simonetti V, Williams M, Rivers EP, et al. Hemodynamic effects of aortic occlusion during hemorrhagic shock and cardiac arrest. J Trauma 1997;42:1023–8. [23] Huo TT, Zeng Y, Liu XN, Sun L, Han HZ, Chen HG, et al. Hydrogen-rich saline improves survival and neurological outcome after cardiac arrest and cardiopulmonary resuscitation in rats. Anesth Analg 2014;119:368–80. [24] DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988;44:837–45. [25] Weil MH, Bisera J, Trevino RP, Rackow EC. Cardiac output and end-tidal carbon dioxide. Crit Care Med 1985;13:907–9. [26] Cantineau JP, Merckx P, Lambert Y, Sorkine M, Bertrand C, Duvaldestin P. Effect of epinephrine on end-tidal carbon dioxide pressure during prehospital cardiopulmonary resuscitation. Am J Emerg Med 1994;12:267–70. [27] Lindberg L, Liao Q, Steen S. The effects of epinephrine/norepinephrine on end-tidal carbon dioxide concentration, coronary perfusion pressure and pulmonary arterial blood flow during cardiopulmonary resuscitation. Resuscitation 2000;43:129–40. [28] Okamoto H, Hoka S, Kawasaki T, Okuyama T, Takahashi S. Changes in end-tidal carbon dioxide tension following sodium bicarbonate administration: correlation with cardiac output and haemoglobin concentration. Acta Anaesthesiol Scand 1995;39:79–84. [29] Nakazawa K, Hikawa Y, Saitoh Y, Tanaka N, Yasuda K, Amaha K. Usefulness of central venous oxygen saturation monitoring during cardiopulmonary resuscitation. Intensive Care Med 1994;20:450–1. [30] Baele PL, McMichan JC, Marsh HM, Sill JC, Southorn PA. Continuous monitoring of mixed venous oxygen saturation in critically ill patients. Anesth Analg 1982;61: 513–7. [31] Rivers E, Nguyen B, Havstad S, Ressler J, Muzzin A, Knoblich B, et al. Early goaldirected therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345:1368–77. [32] Davison DL, Chawla LS, Selassie L, Jones EM, McHone KC, Vota AR, et al. Femoral-based central venous oxygen saturation is not a reliable substitute for subclavian/internal jugular-based central venous oxygen saturation in patients who are critically ill. Chest 2010;138:76–83. [33] van Beest PA, van der Schors A, Liefers H, Coenen LG, Braam RL, Habib N, et al. Femoral venous oxygen saturation is no surrogate for central venous oxygen saturation. Crit Care Med 2012;40:3196–201. [34] Emerman CL, Pinchak AC, Hagen JF, Hancock D. A comparison of venous blood gases during cardiac arrest. Am J Emerg Med 1988;6:580–3. [35] Reynolds JC, Salcido DD, Menegazzi JJ. Coronary perfusion pressure and return of spontaneous circulation after prolonged cardiac arrest. Prehosp Emerg Care 2010; 14:78–84. [36] Ralston SH, Voorhees WD, Showen L, Schmitz P, Kougias C, Tacker WA. Venous and arterial blood gases during and after cardiopulmonary resuscitation in dogs. Am J Emerg Med 1985;3:132–6. [37] Klouche K, Weil MH, Sun S, Tang W, Povoas HP, Kamohara T, et al. Evolution of the stone heart after prolonged cardiac arrest. Chest 2002;122:1006–11. [38] Klouche K, Weil MH, Sun S, Tang W, Kamohara T. Echo-Doppler observations during cardiac arrest and cardiopulmonary resuscitation. Crit Care Med 2000; 28:N212–3. [39] Gong P, Hua R, Zhang Y, Zhao H, Tang Z, Mei X, et al. Hypothermia-induced neuroprotection is associated with reduced mitochondrial membrane permeability in a swine model of cardiac arrest. J Cereb Blood Flow Metab 2013;33:928–34. [40] Kobr J, Kuntscher V, Treska V, Molacek J, Vobruba V, Fremuth J, et al. Adverse effects of the high tidal volume during mechanical ventilation of normal lung in pigs. Bratisl Lek Listy 2008;109:45–51. [41] Vobruba V, Klimenko OV, Kobr J, Cerna O, Pokorna P, Mikula I, et al. Effects of high tidal volume mechanical ventilation on production of cytokines, iNOS, and MIP-1β proteins in pigs. Exp Lung Res 2013;39:1–8.