a b s t r a c t

Purpose: We investigated the predictive value of the gradient between arterial carbon dioxide (PaCO₂) and end- tidal carbon dioxide (ETCO₂) (Pa-ETCO₂) in post-cardiac arrest patients for in-hospital mortality.

Methods: This retrospective observational study evaluated cardiac arrest patients admitted to the emergency de- partment of a Tertiary university hospital. The PaCO₂ and ETCO₂ values at 6, 12, and 24 h after return of sponta- neous circulation (ROSC) were obtained from medical records and Pa-ETCO₂ gap was calculated as the difference between PaCO₂ and ETCO₂ at each time point. Multivariate logistic regression analysis was performed to verify the relationship between Pa-ETCO₂ gap and clinical variables. Receiver operating characteristic (ROC) curve anal- ysis was performed to determine the cutoff value of Pa-ETCO₂ for predicting in-hospital mortality.

Results: The final analysis included 58 patients. In univariate analysis, Pa-ETCO₂ gaps were significantly lower in survivors than in non-survivors at 12 h [12.2 (6.5-14.8) vs. 13.9 (12.1-19.6) mmHg, p = 0.040] and 24 h [9.1 (6.3-10.5) vs. 17.1 (13.1-23.2) mmHg, p b 0.001)] after ROSC. In multivariate analysis, Pa-ETCO₂ gap at 24 h after ROSC was related to in-hospital mortality [odds ratio (95% confidence interval): 1.30 (1.07-1.59), p = 0.0101]. In ROC curve analysis, the optimal cut-off value of Pa-ETCO₂ gap at 24 h after ROSC was 10.6 mmHg (area under the curve, 0.843), with 77.8% sensitivity and 85.7% specificity.

Conclusion: The Pa-ETCO₂ gap at 24 h after ROSC was associated with in-hospital mortality in post-cardiac arrest patients.

(C) 2018

Introduction

End-tidal carbon dioxide (ETCO₂) is useful for monitoring the Quality of cardiopulmonary resuscitation (CPR) during resuscitation and venti- lation status in post-cardiac arrest patients [1-4]. ETCO₂ is also a valu- able predictor for in-hospital mortality in post-cardiac arrest patients [5-8]. In general, ETCO₂ correlates with arterial partial pressure of car- bon dioxide (PaCO₂) and the gradient between the two variables should be 2-5 mmHg [9-11]. However, the gradient may be increased by respi- ratory dead space or low pulmonary circulation and can present as a ventilation/perfusion (V/Q) mismatch [12-17]. Patients with a V/Q mis- match or increased gradient between PaCO₂ and ETCO₂ (Pa-ETCO₂ gap) have a high probability of in-hospital mortality [18,19]. This phenome- non may occur in post-cardiac arrest patients because of traumatic lung injury secondary to vigorous chest compression, early-onset pneu- monia due to aspiration, pulmonary interstitial edema secondary to is- chemia-reperfusion injury, or Myocardial stunning, which can lead to deterioration of pulmonary function [20-25]. We investigated the

* Corresponding author at: Department of Emergency Medicine, Yonsei University Wonju College of Medicine, 20 Ilsan-ro, Wonju 26426, Republic of Korea.

E-mail address: [email protected]. (K.-C. Cha).

predictive value of the Pa-ETCO₂ gap for in-hospital mortality in post- cardiac arrest patients.

Material and methods

Study design

This retrospective observational study evaluated cardiac arrest pa- tients admitted to the emergency department of a tertiary university hospital between March 2011 and February 2017. The study protocol was approved by Institutional Review Board of Wonju Severance Chris- tian Hospital (YWMR-CR317049).

In the Wonju region, patients with out-of-hospital cardiac arrest (OHCA) are managed by emergency medical technicians dispatched from a fire department. EMTs provide both basic and ad- vanced life support, including defibrillation and advanced airway man- agement for a minimum of 5 min at the scene. If they cannot achieve return of spontaneous circulation (ROSC), the patient is transported to the nearest emergency department (ED) while the EMTs continue to perform cardiopulmonary resuscitation (CPR) in the ambulance. Once ROSC is achieved, the patient is referred to our hospital. In the hospital, the patient received comprehensive post-cardiac arrest care including therapeutic hypothermia at 32-34 ?C or targeted temperature

https://doi.org/10.1016/j.ajem.2018.04.025

0735-6757/(C) 2018

2 Y.W. Kim et al. / American Journal of Emergency Medicine 37 (2019) 1-4

management (TTM) at 33 ?C or 36 ?C. ventilator support to achieve normoxia and normocarbia is provided during ED and intensive care unit stays. The tidal volume, respiratory rate, and positive end-ex- piratory pressure are controlled to achieve a target arterial oxygen sat- uration (SaO₂) of 94-98%, ETCO₂ of 30-40 mmHg and PaCO₂ of 35-45 mmHg. Pulse oximetry (Tram-rac 4A, GE Medical Systems, WI, USA) and ETCO₂ (CAPNOSTAT mainstream CO₂ module, GE Medical Systems, WI, USA) were monitored continuously, and sequential arterial blood gas analysis (ABGA) was performed at 6, 12, and 24 h after ROSC, followed by at every 24 h. Patient care and other ancillary tests were de- cided by the intensivist on duty.

We included patients older than 18 years who had survived for N24 h after successful resuscitation from out-of-hospital non-traumatic car- diac arrest. Patients transferred from other hospitals or without matched data for ETCO₂ and ABGA were excluded.

Study variables

Clinical data obtained from medical records included age, sex, his- tory of previous pulmonary disease, witnessed cardiac arrest, bystander CPR, initial presenting rhythm, etiology of arrest, estimated total col- lapse time, total duration of CPR, cumulative number of defibrillation at- tempts, temperature of therapeutic hypothermia or Targeted temperature management (TTM), and in-hospital mortality. Previous pulmonary disease was confirmed by medical records and chest X-ray or computed tomography (CT) reviewed by an independent radiologist to our study. Pre-arrest pulmonary dysfunction was defined as history of previous pulmonary disease or cardiac arrest caused by acute respira- tory failure due to lung parenchymal or pulmonary vascular disease. The values of PaCO₂ and ETCO₂ at 6, 12, and 24 h after ROSC were also ob- tained from medical records, and Pa-ETCO₂ gap was calculated as the difference between PaCO₂ and ETCO₂ at each time point.

Data analysis

Continuous variables were reported as median values (interquartile range: IQR) and were compared with the Mann-Whitney U test. Nomi- nal data were calculated as percentage of frequency of occurrence and compared using chi-square or Fisher’s exact test, as appropriate. Linear mixed model analysis was used to compare PaCO₂, ETCO₂, and Pa-ETCO₂ gap at 6, 12, and 24 h after ROSC between survivors and non-survivors. Multivariate logistic regression analysis was performed to determine whether age, sex, witnessed cardiac arrest, bystander CPR, initial pre- senting rhythm, estimated total collapse time, total duration of CPR, TTM, or pre-arrest pulmonary dysfunction affected in-hospital mortal- ity. The resulting odds ratios (ORs) are presented with 95% confidence intervals (95% CIs). A two-sided p-value b0.05 was considered

statistically significant. Statistical analysis was performed using R ver- sion 3.4.0 (The R Foundation for Statistical Computing, Vienna, Austria).

Results

General characteristics

During the study period, 590 non-traumatic OHCA patients over 18 years old adult patients were visited to our ED from scene. Among them, 227 (38.5%) patients archived ROSC and 66 patients (11.2%) of them survived N24 h after ROSC. 8 patients were excluded due to mismatched sampling times for ETCO₂ and PaCO₂. Finally, 58 patients were analyzed. There were no differences in general characteristics, except age, be- tween survivors and non-survivors. Survivors were younger than non- survivors (p = 0.010) (Table 1).

Comparison of PaCO₂, ETCO₂, and Pa-ETCO₂ gap in survivor and non- survivor groups

There were no differences in PaCO₂ at 6, 12, and 24 h after ROSC be- tween survivors and non-survivors. ETCO₂ at 6 h [32.0 (27.0-37.0) mmHg vs. 27.0 (20.0-33.0) mmHg, p = 0.023], 12 h [33.0 (30.0-35.0)

mmHg vs. 26.0 (23.0-30.0) mmHg, p = 0.001], and 24 h (37.0 (32.0-

40.0) mmHg vs. 28.0 (21.0-33.0) mmHg, p b 0.001] was significantly higher in survivors than in non-survivors. Pa-ETCO₂ gaps at 12 h [12.2 (6.5-14.8) mmHg vs. 13.9 (12.1-19.6) mmHg, p = 0.040] and 24 h

[9.1 (6.3-10.5) mmHg vs. 17.1 (13.1-23.2) mmHg, p b 0.001] after

ROSC were significantly lower in survivors than in non-survivors. There were no group-time interactions for PaCO₂, ETCO₂, and Pa- ETCO₂ (p = 0.237, 283, and 0.207 respectively) (Fig. 1).

Predictive value of Pa-ETCO2 gap for in-hospital mortality

Multivariate logistic regression analysis revealed that the Pa-ETCO₂ gap [OR (95% CI): 1.30 (1.07-1.59), p = 0.010] at 24 h after ROSC was

related to in-hospital mortality (Table 2). The Pa-ETCO2 gap had an area under the receiver operating characteristic curve of 0.842 at 24 h, and the optimal cut-off value of Pa-ETCO₂ gap at 24 h was 10.6 mmHg, with 77.8% sensitivity and 85.7% specificity (Fig. 2).

Discussion

The Pa-ETCO₂ gap at 24 h after ROSC was associated with in-hospital mortality in post-cardiac arrest patients in this study. Previous studies observed that the Pa-ETCO₂ gap increased when dead space ventilation increased or cardiac output decreased, demonstrating its value as a pre- dictor of disease severity or outcomes in critically ill patients [15-18]. In

Table 1

General characteristics.

	Total	Survivors	Non-survivors	p-value
	(n = 58)	(n = 15)	(n = 43)
Age (y)	60 (53-75)	54 (51-60)	71 (55-76)	0.010
Male gender, no. (%)	41 (70.7)	13 (86.7)	28 (65.1)	0.188
Witnessed, no. (%)	44 (75.9)	12 (80.0)	32 (74.4)	0.742
Bystander CPR, no. (%)	32 (55.2)	10 (66.7)	22 (51.2)	0.374
Shockable rhythm, no. (%)	7 (12.1)	3 (20.0)	4 (9.3)	0.360
Cumulative number of defibrillation attempts	1 (1-4)	3 (2-6)	1 (1-3)	0.714
cardiac etiology, no. (%)	19 (32.8)	7 (46.7)	12 (27.9)	0.213
Estimated total collapse time (min)	31 (22-44)	29 (20-39)	34 (23-44)	0.494
Total CPR duration (min)	28 (19-35)	27 (20-31)	28 (19-35)	0.644
Therapeutic hypothermia or TTM, no. (%)	53 (91.4)	14 (93.3)	39 (90.7)	1.000
33 ?C	50 (94.3)	13 (92.9)	37 (94.9)	1.000
36 ?C	3 (5.7)	1 (7.1)	2 (5.1)	1.000
Pre-arrest pulmonary dysfunction, no. (%)	18 (31.0)	3 (20.0)	15(34.9)	0.348

?Variables are presented as median (interquartile range) or frequency (%).

^?CPR: cardiopulmonary resuscitation, TTM: targeted temperature management.

Y.W. Kim et al. / American Journal of Emergency Medicine 37 (2019) 1-4 3

Fig. 1. Trends of PaCO₂ (A) ETCO₂ (B) and Pa-ETCO₂ gap (C) of survivors and non-survivors during 24 h after ROSC. There was no group-time interaction in all variables. (p = 0.237, 283 and 0.207 respectively). ^*ROSC: return of spontaneous circulation.

Table 2

Multivariate logistic regression analysis of Pa-ETCO₂ gap at each time point for prediction of in-hospital mortality.

Variables (mmHg)	Odds ratio (95% CI)	p-value
6h Pa-ETCO2 gap	1.11(0.99-1.26)	0.073
12h Pa-ETCO2 gap	1.08(0.97-1.20)	0.186
24h Pa-ETCO2 gap	1.30(1.07-1.60)	0.010

post-cardiac arrest patients, global ischemia and reperfusion injury can deteriorate pulmonary and Cardiovascular function, which might be presented as increased Pa-ETCO₂ gap. [26]. Patients who do not over- come these critically adverse events tend to deteriorate over time, as demonstrated by the increased Pa-ETCO₂ gap in this study.

Post-cardiac arrest patients with Poor neurologic outcomes can be a burden to their families and society responsible for their care [27]. Various Prognostic tools have been used to predict patient outcomes,

Fig. 2. Analysis by receiver operating characteristic curve for 24 h Pa-ETCO₂ gap for predicting in-hospital mortality.

including bedside neurologic examination, somatosensory evoked po- tentials, electroencephalography (EEG), brain CT, magnetic resonance imaging (MRI), and biochemical markers, but they have limitations [28-33]. Sedatives or Neuromuscular blocking agents might mask a patient’s response, and there is risk in moving critically ill patients from the ICU to a CT or an MRI facility for neuroimaging. Several studies indicated that continuous EEG monitoring is the most reliable prognos- tic tool, but EEG requires specialist review [34]. The Pa-ETCO₂ gap can be easily obtained at any time, and any medical personnel can objectively review the results, making it a useful to identify patients at higher like- lihood of in-hospital mortality and in need of additional therapeutic intervention.

In our study, increased Pa-ETCO₂ gap was mainly caused by reduc- tion of ETCO₂. Guidelines for post-cardiac arrest patient care recom- mend continuous monitoring of ETCO₂ to promote cerebral perfusion by maintaining an optimal PaCO₂ level [34]. Thus, an increased Pa- ETCO₂ gap would be a clue indicating uncorrected adverse event occur- rence. Therefore, to clarify whether the Pa-ETCO₂ gap is increasing, PaCO₂ should be determined if the ETCO₂ remains decreased in a post- cardiac arrest patient dependent on a mechanical ventilator. It is insuf- ficient to monitor the ETCO₂ alone to optimize the PaCO₂ level, and fac- tors causing deterioration in a patient’s condition should be identified and corrected.

This study had several limitations. First, it is possible that mis- matched PaCO₂ and ETCO₂ values were included in analysis, even though data were prospectively collected using a clinical practice proto- col. Second, even though we excluded patients with pulmonary dys- function to minimize bias due to a mismatch between PaCO₂ and ETCO₂ values, it is possible that patients with undetected pulmonary dysfunction were included. Third, characteristics of enrolled patients might affect the Pa-ETCO₂ gap or in-hospital mortality because the ma- jorities were male, Non-shockable rhythm or non-cardiac etiology even though these variables did not related with in-hospital mortality in mul- tivariate logistic regression analysis. Fourth, this study might not be generalized for all patients with cardiac arrest because we included pa- tients who survived N24 h after ROSC only. Finally, this was a single-cen- ter, observational study with a small sample size. A larger, population- based, multi-center study is needed to generalize our results.

Conclusions

The gradient between PaCO₂ and ETCO₂ at 24 h after ROSC is associ- ated with in-hospital mortality in post-cardiac arrest patients. It might be useful predictor of in-hospital mortality in post-cardiac arrest pa- tients because it can be obtained easily and evaluated objectively.

4 Y.W. Kim et al. / American Journal of Emergency Medicine 37 (2019) 1-4

Sources of support

There was no source of support.

Conflict of interest

All authors have nothing to declare on conflict of interest.

References

1. Link MS, Berkow LC, Kudenchuk PJ, Halperin HR, Hess EP, Moitra VK, et al. Part 7: adult advanced cardiovascular life support: 2015 American Heart Association guide- lines update for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2015;132(18 Suppl 2):S444-64.
2. Lee MJ, Rho TH, Kim H, Kang GH, Kim JS, Rho SG, et al. Part 3. Advanced cardiac life support: 2015 Korean guidelines for cardiopulmonary resuscitation. Clin Exp Emerg Med 2016;3(Suppl):S17-26.
3. Soar J, Nolan JP, Bottiger BW, Perkins GD, Lott C, Carli P, et al. European resuscitation council guidelines for resuscitation 2015: section 3. Adult advanced life support. Re- suscitation 2015;95:100-47.
4. Falk JL, Rackow EC, Weil MH. End-tidal carbon dioxide concentration during cardio- pulmonary resuscitation. N Engl J Med 1988;318(10):607-11.
5. Grmec S, Kupnik D. Does the Mainz Emergency Evaluation Scoring (MEES) in com- bination with capnometry (MEESc) help in the prognosis of outcome from cardio- pulmonary resuscitation in a prehospital setting? Resuscitation 2003;58(1):89-96.
6. Grmec S, Klemen P. Does the end-tidal carbon dioxide concentration have prognostic value during out-of-hospital cardiac arrest? Eur J Emerg Med 2001;8 (4):263-9.
7. Sanders AB, Kern KB, Otto CW, Milander MM, Ewy GA. End-tidal carbon dioxide monitoring during cardiopulmonary resuscitation: a prognostic indicator for sur- vival. JAMA 1989;262(10):1347-51.
8. Grmec S, Krizmaric M, Mally S, Kozelj A, Spindler M, Lesnik B. Utstein style analysis of out-of-hospital cardiac arrest-bystander CPR and end expired carbon dioxide. Resuscitation 2007;72(3):404-14.
9. Stock MC. Capnography for adults. Crit Care Clin 1995;11(1):219-32.
10. Nunn J, Hill D. Respiratory dead space and arterial to end-tidal CO2 tension differ- ence in anesthetized man. J Appl Physiol 1960;15(3):383-9.
11. Fletcher R, Jonson B. Deadspace and the single breath test for carbon dioxide during anaesthesia and artificial ventilation. Effects of tidal volume and frequency of respi- ration. Br J Anaesth 1984;56(2):109-19.
12. Hatle L, Rokseth R. The arterial to end-expiratory carbon dioxide tension gradient in acute pulmonary embolism and other cardiopulmonary diseases. Chest 1974;66(4): 352-7.
13. Hardman JG, Aitkenhead AR. Estimating alveolar dead space from the arterial to end-tidal CO(2) gradient: a modeling analysis. Anesth Analg 2003;97(6):1846-51.
14. Fletcher R. Invasive and Noninvasive measurement of the respiratory deadspace in anesthetized children with cardiac disease. Anesth Analg 1988;67(5):442-7.
15. Yousuf T, Brinton T, Murtaza G, Wozniczka D, Ahmad K, Iskandar J, et al. Establishing a gradient between partial pressure of arterial carbon dioxide and end-tidal carbon

    dioxide in patients with acute respiratory distress syndrome. J Invest Med 2017;65 (2):338-41.

    Ornato JP, Garnett AR, Glauser FL. Relationship between cardiac output and the end- trial carbon dioxide tension. Ann Emerg Med 1990;19(10):1104-6.

16. Yamanaka MK, Sue DY. Comparison of arterial-end-tidal PCO2 difference and dead space/tidal volume ratio in respiratory failure. Chest 1987;92(5):832-5.
17. Shetty AL, Lai KH, Byth K. The CO2 GAP project-CO2 GAP as a prognostic tool in emergency departments. Emerg Med Australas 2010;22(6):524-31.
18. Tyburski JG, Carlin AM, Harvey EH, Steffes C, Wilson RF. End-tidal CO2-arterial CO2 differences: a useful intraoperative mortality marker in trauma surgery. J Trauma 2003;55(5):892-7.
19. Cha KC, Kim YW, Kim HI, Kim OH, Cha YS, Kim H, et al. Parenchymal lung injuries related to standard cardiopulmonary resuscitation. Am J Emerg Med 2017;35(1): 117-21.
20. Adrie C, Adib-Conquy M, Laurent I, Monchi M, Vinsonneau C, Fitting C, et al. Success- ful cardiopulmonary resuscitation after cardiac arrest as a “sepsis-like” syndrome. Circulation 2002;106(5):562-8.
21. Samborska-Sablik A, Sablik Z, Gaszynski W. The role of the immuno-inflammatory response in patients after cardiac arrest. Arch Med Sci 2011;7(4):619-26.
22. Kang DH, Kim J, Rhee JE, Kim T, Kim K, Jo YH, et al. The risk factors and prognostic implication of acute pulmonary edema in resuscitated cardiac arrest patients. Clin Exp Emerg Med 2015;2(2):110-6.
23. Kakavas S, Mongardon N, Cariou A, Gulati A, Xanthos T. Early-onset pneumonia after out-of-hospital cardiac arrest. J Infect 2015;70(6):553-62.
24. Kern KB, Hilwig RW, Rhee KH, Berg RA. Myocardial dysfunction after resuscitation from cardiac arrest: an example of global myocardial stunning. J Am Coll Cardiol 1996;28(1):232-40.
25. Neumar RW, Nolan JP, Adrie C, Aibiki M, Berg RA, Bottiger BW, et al. Post-cardiac ar- rest syndrome. Circulation 2008;118(23):2452-83.
26. Hamel MB, Phillips R, Teno J, Davis RB, Goldman L, Lynn J, et al. cost effectiveness of aggressive care for patients with nontraumatic coma. Crit Care Med 2002;30(6): 1191-6.
27. Booth CM, Boone RH, Tomlinson G, Detsky AS. Is this patient dead, vegetative, or se- verely neurologically impaired?: assessing outcome for comatose survivors of car- diac arrest. JAMA 2004;291(7):870-9.
28. Wijdicks EF, Bamlet WR, Maramattom BV, Manno EM, McClelland RL. Validation of a new coma scale: the FOUR score. Ann Neurol 2005;58(4):585-93.
29. Levy D, Caronna J, Singer B, Lapinski R, Frysman H, Plum F. Predicting outcome from hypoxic-ischemic coma. JAMA 1985;253(10):1420-6.
30. Martens P. serum neuron-specific enolase as a prognostic marker for irreversible brain damage in comatose cardiac arrest survivors. Acad Emerg Med 1996;3(2): 126-31.
31. Hachimi-Idrissi S, Van der Auwera M, Schiettecatte J, Ebinger G, Michotte Y, Huyghens L. S-100 protein as early predictor of regaining consciousness after out of hospital cardiac arrest. Resuscitation 2002;53(3):251-7.
32. Zandbergen EG, de Haan RJ, Stoutenbeek CP, Koelman JH, Hijdra A. Systematic re- view of Early prediction of poor outcome in anoxicischaemic coma. Lancet 1998; 352(9143):1808-12.
33. Callaway CW, Donnino MW, Fink EL, Geocadin RG, Golan E, Kern KB, et al. Part 8: Post-cardiac arrest care. Circulation 2015;132(18 suppl 2):S465-82.