Dandi Wang1
Yecheng Liu
Huadong Zhu- 1Emergency Department, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, China
- 2Department of Family Medicine & Division of General Internal Medicine, State Key Laboratory of Complex Severe and Rare Diseases,Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, Beijing, China
- 3Department of Medicine, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, China
- 4Department of Health Care, State Key Laboratory of Complex Severe and Rare Diseases, Peking Union Medical College Hospital, Chinese Academy of Medical Science and Peking Union Medical College, Beijing, China
Background and aims: Current international guidelines suggest that normocapnia should be targeted during the resuscitation phase following cardiac arrest. However, some studies propose that therapeutic hypercapnia might be a potential strategy to enhance cerebral perfusion and improve patient outcomes after cardiac arrest. However, few studies have explored the association between arterial carbon dioxide tension and the prognosis of patients after cardiac arrest. This systematic review and meta-analysis aims to evaluate the influence of arterial carbon dioxide tension on the prognosis of patients after cardiac arrest.
Data sources: We searched MEDLINE, Embase, and Cochrane CENTRAL to identify studies that evaluated the association between the partial pressure of arterial carbon dioxide and outcomes after cardiac arrest. The primary outcome was the neurological status at the end of the follow-up period. The secondary outcomes included short-, mid-, and long-term mortality. The meta-analysis was conducted if statistical heterogeneity was low.
Results: Twelve studies were included. Compared with normocapnia, hypercapnia was associated with lower in-hospital mortality (pooled OR 0.74, 95% CI 0.59–0.92). For neurological outcomes, hypercapnia was associated with less short favorable outcome (pooled OR 0.42, 95% CI 0.22–0.8).
Conclusion: Among cardiac arrest patients, hypercapnia was associated with a reduction in in-hospital mortality and favorable neurological outcome compared to normocapnia.
Systematic review registration: This systematic review was registered with PROSPERO (ID: CRD42025644636).
Introduction
Although cardiopulmonary resuscitation may initially be successful, a large number of patients who survive cardiac arrest do not make it to hospital discharge, and those who do often suffer from severe neurological damage. After cardiac arrest, the subsequent hours and days are characterized by post-cardiac arrest syndrome, which includes a pronounced inflammatory reaction, failure of organ bioenergetics, and an elevated risk of morbidity and mortality.
The partial pressure of arterial carbon dioxide (PaCO2) is crucial for the regulation of cerebral blood flow and has anti-convulsive, anti-inflammatory, and anti-oxidant effects, making therapeutic hypercapnia a potential strategy to improve cerebral perfusion following cardiac arrest. Global guidelines recommend aiming for normocapnia in unconscious adults who have been resuscitated from out-of-hospital cardiac arrest (1). However, many studies have indicated that compared with patients who maintain normocapnia, those exposed to hypercapnia or hypocapnia may have a better neurological prognosis at 12 months (2, 3). However, some studies have presented opposing views. For instance, Ebner et al. (4) indicated that while PaCO2 often fluctuates after cardiac arrest, it is not related to neurological outcomes. In summary, there is currently no consensus on the optimal target for PaCO2 in patients with cardiac arrest.
This meta-analysis aims to compare the effects of different levels of hypercapnia on the prognosis of patients after cardiac arrest, providing new insights into the diagnosis and treatment of this population.
Methods
The systematic review and meta-analysis protocol was prospectively registered in the PROSPERO database [CRD42025644636]. The study was performed in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines.
Eligibility criteria
Ovid MEDLINE (R), Ovid MEDLINE (R) in process and other non-indexed citations, Ovid Embase, CINAHL Plus with Full Text (EBSCO), and the Cochrane Central Register of Controlled Trials (CENTRAL) were searched to identify eligible studies published between database inception and June 2024. The reference lists of review articles, editorials, and citation lists in Scopus were also hand-searched to identify possible additional studies. The end date for the literature search was 1 June 2024. For papers that met the inclusion criteria but that did not report patient outcomes for the PaCO2 group, the authors were contacted to request the original data.
Data search
The key search terms were “Heart Arrest” AND “Carbon Dioxide” AND “Outcomes Assessment.” The detailed search strategy is described in the appendix (Supplementary Table S1). We used the Medical Subject Headings database to identify synonyms, and no language restrictions were applied. For trials that were registered and completed but not yet published, we contacted the authors to obtain the aggregate results.
Study selection and data extraction
Two authors (DDW, ZML) independently screened the articles for inclusion on the basis of their titles and abstracts, and reviewed potentially eligible full-text papers according to predefined criteria. We collected data on trial characteristics, demographics, intervention and control procedures, and review outcomes from the eligible studies. In trials with multiple study arms, only data from the arms relevant to this review were extracted. When crucial information was missing, we contacted the corresponding authors. For transparency, we explicitly categorized the exposure definitions used in the included studies. The majority of studies reported true arterial PaCO2 values obtained from arterial blood gas analysis. However, Inoue et al. (5) used a surrogate measure, obtaining blood samples from a femoral artery or vein, which we have indicated in Table 1.
Table 1. Characteristics of studies included in the meta-analysis.
Risk of bias assessment of individual studies
The 21-item Newcastle–Ottawa scale was used to assess the methodological quality of the studies included in this systematic review (6). The results were collated, and the accuracy was independently checked by two authors. To evaluate the risk of bias in the individual randomized controlled trials (RCTs), we used the revised uniform criteria of the Cochrane risk-of-bias tool for randomized trials, version 2. We assessed the certainty of evidence for each outcome using the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) approach (7). Any disagreements were resolved by discussion and referral to a third author (YCL) if necessary.
Outcomes
The primary outcome was short-term favorable neurological outcomes. The neurological prognosis was described according to the Cerebral Performance Category (CPC), Glasgow Outcome Scale-Extended score, or modified Rankin Scale score. We defined the short-term favorable neurological outcome was defined as CPC 1–2 < 1 month. The secondary outcomes were long-term neurological outcomes, and short-term, mid-term, and long-term mortality. We defined the long-term favorable neurological outcome was defined as CPC 1–2 at 6–12 months, short-, mid-, and long-term mortality were defined as hospital mortality, 1-month mortality, and 6-month mortality, respectively.
Data analysis
We used DerSimonian and Laird random-effects models to conduct the meta-analysis. We generated study weights using the inverse variance weighting method. The results are presented as relative risks and risk differences for dichotomous outcomes and as standardized mean differences for continuous outcomes, all with 95% confidence intervals (CIs) and 95% prediction intervals (RevMan, version 5.3, Cochrane Collaboration). For outcomes with substantial heterogeneity (I2 > 50%), pooled estimates were calculated using a random-effects model and are reported as exploratory with cautious interpretation.
Heterogeneity
Studies were tested for heterogeneity using the I2 statistic. I2 values of <25, 25–50%, and >50% were considered to have low, moderate, and high heterogeneity between studies, respectively. We initially proposed to report a pooled estimate only when the I2 statistic was <50%. However, given that substantial heterogeneity was anticipated across observational studies, we additionally performed random-effects meta-analyses (DerSimonian–Laird and Hartung–Knapp–Sidik–Jonkman [HKSJ] adjustments) even when I2 exceeded 50%, and reported 95% prediction intervals to reflect the uncertainty of effect size distribution. In these cases, the “overall” estimates are presented as exploratory summaries, while interpretation focuses on directionality and robustness rather than precise point estimates. Narrative synthesis was also provided to complement these pooled results. We also considered differences in the follow-up times and variations in the PaCO2 cutoffs used to define normocapnia as potential sources of clinical heterogeneity among the studies. We planned to conduct a sensitivity analysis to determine the effects of these variations on the overall conclusion of the meta-analysis.
Results
Upon database searching, we identified a total of 689 records. Before reading, we removed 45 duplicate records, and based on the titles and abstracts of the articles, 610 were excluded. We carefully reviewed the remaining 34 articles and selected 12that met the inclusion criteria, including 3 RCTs (3, 8, 9) and 9 observational studies (3, 5, 10–16). Figure 1 shows the entire article screening process.
Figure 1. PRISMA flowchart illustrating the identification, screening, and inclusion of relevant literature. PRISMA, Preferred Reporting Items for Systematic reviews and Meta-Analyses.
Study characteristics
All 12 studies were published in the English language between 2013and 2024. The studies originated from Australia and New Zealand (combined), the United States, South Korea, China and the Netherlands, and collectively, they involved a total of 27,243 patients (ranging from 44 to 23,803 patients in each study). Table 1 presents the characteristics of each of the identified articles.
Among the 12 included studies, 11 reported arterial PaCO2 directly from arterial blood gas analysis, whereas 1 studies used surrogate measures. Specifically, 1 reported venous or unspecified PCO2 values on admission.
Association between PaCO2 and neurological outcomes
A total of 8 studies reported neurological outcomes,4 of these studies compared the short neurological outcomes between patients with hypercapnia and those with normocapnia, as well as between those with normocapnia and those with hypocapnia.
Compared with normocapnia, hypercapnia was associated with a significantly reduced odds of short favorable neurological outcome (OR 0.42, 95% CI 0.22–0.80), suggesting a better neurological prognosis (Figure 2). The study results exhibited moderate high heterogeneity (I2 = 52%). We conducted a sensitivity analysis on the aforementioned results (Supplementary Table S2) and found that after removing the study by Lee et al., which kept temperature of 33 °C ± 1 °C for 24 h. The results remained significant, with a substantial reduction in heterogeneity. However, there was no significant difference in the impact of PaCO2 level on the long-term central nervous system prognosis among the three groups of patients. Detailed results are presented in Supplementary Figures S3–S7.
Figure 2. Forest plot of the comparison of short-term neurological outcomes between the hypercapnia group and the normocapnia group.
Association between PaCO2 and mortality
Among the included studies, eight compared the mortality rates between different PaCO2 level. Hypercapnia and hypocapnia vs. normocapnia was associated with reduced in-hospital mortality (random-effects pooled OR 0.74, 95% CI 0.59–0.92; k = 6; I2 = 76%; OR 1.42, 95% CI 1.28–1.58 I2 = 38%; Figure 3). Detailed results are presented in Supplementary Figures S8–S12. Given the substantial heterogeneity in some comparisons (e.g., I2 = 76% for in-hospital mortality) these pooled estimates should be interpreted with caution.
Figure 3. Forest plot of the comparison of mild-term mortality between the normocapnia group and the hypocapnia group.
Discussion
In this meta-analysis, we showed that compared with patients with normocapnia, patients with hypercapnia were significantly associated with bad neurological status after cardiac arrest, especially in the short-term. Additionally, hypercapnia and hypocapnia may be associated with reduced short-term and mid-term mortality compared with normocapnia.
Among individuals who experience cardiac arrest, the main cause of death and long-term disability in survivors during the acute phase is brain injury, commonly referred to as post-cardiac arrest brain injury (or hypoxic–ischemic brain injury [HIBI]). Previous studies have proposed that PaCO2 is a major physiological regulator of cerebral vascular tone, with each 1 mmHg increase in PaCO2 leading to a 2 mL/100 g increase in cerebral blood flow. Furthermore, although it is not sufficient to assess patient outcomes, a multicenter phase 2 randomized trial showed that hypercapnia significantly reduced the release of neuron-specific enolase, a biomarker of brain injury (17–20). However, the aforementioned conclusions are not entirely consistent with the findings of the present study. A primary reason for this discrepancy may be that the included literature did not specify the severity of hypercapnia, and the possibility of extreme or uncontrolled hypercapnia cannot be excluded. Such pronounced hypercapnia could lead to severe respiratory acidosis, causing a significant decrease in intracellular pH, which may in turn inhibit enzymatic function, exacerbate cerebral edema, and potentially increase intracranial pressure, ultimately resulting in deleterious effects.
The present study has some limitations that should be considered. First, a limited number of RCTs were included in the analysis; the analysis primarily included observational studies, which may explain the heterogeneity in our results. Although we used HKSJ adjustments and prediction intervals to mitigate the impact of heterogeneity, these pooled results should be interpreted cautiously and considered exploratory rather than definitive.
Additionally, among the three included RCTs, different carbon dioxide cutoffs were used. Although the sensitivity analysis attempted to remove these articles without significantly affecting the results, we still consider this to be one of the reasons for the high heterogeneity. Second, in the included literature, the timing of ABG sampling varied. Moreover, hypothermia treatment may be used in patients with cardiac arrest, and different temperatures can affect the measured PaCO2. The consistency of results across adjusted and unadjusted estimates suggests that our conclusions are robust, although residual confounding remains possible.
Finally, only a small number of articles graded hypercapnia, so we did not obtain an exact PaCO2 range that was associated with reduced patient mortality and improved neurological outcomes. In light of these limitations, additional studies are needed in the future to validate the conclusions of this article.
Conclusion
In patients with cardiac arrest, hypercapnia was associated with reduced in-hospital mortality, but showed no clear benefit in short-term neurological outcomes and no difference in long-term outcomes. One result suggested a possible increase in 1-month mortality. These findings highlight the complexity of CO2 management after resuscitation and should be validated in future large-scale RCTs.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding authors.
Author contributions
DW: Writing – original draft. ZL: Writing – review & editing. GQ: Writing – review & editing. LC: Writing – review & editing. TZ: Writing – review & editing. HiZ: Writing – review & editing. YL: Writing – review & editing. HaZ: Writing – review & editing.
Funding
The author(s) declare that financial support was received for the research and/or publication of this article. This work was supported by CAMS Innovation Fund for Medical Sciences (CAMS Innovation Fund for Medical Sciences (CIFMS); 2023-I2M-C&T-B-025) and the National Educational Science Planning Project (BLA 200221).
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declare that no Gen AI was used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmed.2025.1687522/full#supplementary-material
Abbreviations
RCT, Randomized controlled trial; CPC, Cerebral Performance Category; NSE, Neuron-specific enolase; RR, Relative risk; CI, Confidence interval; ICU, Intensive care unit; GRADE, Grading of Recommendations Assessment, Development and Evaluation.
References
1. Martin, SS, Aday, AW, and Almarzooq, ZI. 2024 heart disease and stroke statistics: a report of US and global data from the American Heart Association [published correction appears in circulation. 2024 May 7;149(19):e1164]. Circulation. (2024) 149:e347–913. doi: 10.1161/CIR.0000000000001209
2. Eastwood, GM, Schneider, AG, and Suzuki, S. Targeted therapeutic mild hypercapnia after cardiac arrest: a phase II multi-Centre randomised controlled trial (the CCC trial). Resuscitation. (2016) 104:83–90. doi: 10.1016/j.resuscitation.2016.03.023
3. Vaahersalo, J, Bendel, S, Reinikainen, M, Kurola, J, Tiainen, M, Raj, R, et al. Arterial blood gas tensions after resuscitation from out-of-hospital cardiac arrest: associations with long-term neurologic outcome. Crit Care Med. (2014) 42:1463–70. doi: 10.1097/CCM.0000000000000228
4. Ebner, F, Riker, RR, Haxhija, Z, Seder, DB, May, TL, Ullén, S, et al. The association of partial pressures of oxygen and carbon dioxide with neurological outcome after out-of-hospital cardiac arrest: an explorative international cardiac arrest registry 2.0 study. Scand J Trauma Resusc Emerg Med. (2020) 28:67. doi: 10.1186/s13049-020-00760-7
5. Inoue, F, Inoue, A, Nishimura, T, Takahashi, R, Nakatani, Y, Suga, M, et al. PCO2 on arrival as a predictive biomarker in patients with out-of-hospital cardiac arrest. The Am J Emerg Med. (2023) 2023:92–9. doi: 10.1016/j.ajem.2023.04.014
6. Risk of bias tools - RoB 2 tool. (2022). Available online at: www.riskofbias.info/welcome/rob-2-0-tool (Accessed November 1, 2022), (n.d.)
7. Guyatt, GH, Oxman, AD, Vist, GE, Kunz, R, Falck-Ytter, Y, Alonso-Coello, P, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. (2008) 336:924–6. doi: 10.1136/bmj.39489.470347.AD
8. Eastwood, G, Nichol, AD, and Hodgson, C. Mild hypercapnia or normocapnia after out-of-hospital cardiac arrest. N Engl J Med. (2023) 389:45–57. doi: 10.1056/NEJMoa2214552
9. Jakkula, P, Reinikainen, M, Hästbacka, J, Loisa, P, Tiainen, M, Pettilä, V, et al. Targeting two different levels of both arterial carbon dioxide and arterial oxygen after cardiac arrest and resuscitation: a randomised pilot trial. Intensive Care Med. (2018) 44:2112–21. doi: 10.1007/s00134-018-5453-9
10. Roberts, BW, Kilgannon, JH, Chansky, ME, Mittal, N, Wooden, J, and Trzeciak, S. Association between postresuscitation partial pressure of arterial carbon dioxide and neurological outcome in patients with post-cardiac arrest syndrome. Circulation. (2013) 127:210713. doi: 10.1161/CIRCULATIONAHA.112.000168
11. Lee, BK, Jeung, KW, and Lee, HY. Association between mean arterial blood gas tension and outcome in cardiac arrest patients treated with therapeutic hypothermia. Am J Emerg Med. (2014) 32:55–60. doi: 10.1016/j.ajem.2013.09.044
12. Wang, CH, Huang, CH, and Chang, WT. Association between early arterial blood gas tensions and neurological outcome in adult patients following in-hospital cardiac arrest. Resuscitation. (2015) 89:17. doi: 10.1016/j.resuscitation.2015.01.003
13. Okada, N, Matsuyama, T, and Okada, Y. Post-resuscitation partial pressure of arterial carbon dioxide and outcome in patients with out-of-hospital cardiac arrest: a multicenter retrospective cohort study. J Clin Med. (2022) 11. doi: 10.3390/jcm11061523
14. Zhou, D, Li, Z, Zhang, S, Wu, L, Li, Y, and Shi, G. Association between mild hypercapnia and hospital mortality in patients admitted to the intensive care unit after cardiac arrest: a retrospective study. Resuscitation. (2020) 149:30–8. doi: 10.1016/j.resuscitation.2020.01.036
15. Helmerhorst, HJ, Roos-Blom, MJ, van Westerloo, DJ, Abu-Hanna, A, and Jonge, E. Associations of arterial carbon dioxide and arterial oxygen concentrations with hospital mortality after resuscitation from cardiac arrest. Crit Care. (2015) 19:348. doi: 10.1186/s13054-015-1067-6
16. Schneider, AG, Eastwood, GM, and Bellomo, R. Arterial carbon dioxide tension and outcome in patients admitted to the intensive care unit after cardiac arrest. Resuscitation. (2013) 84:927–34. doi: 10.1016/j.resuscitation.2013.02.014
17. Martinell, L, Nielsen, N, and Herlitz, J. Early predictors of poor outcome after out-of hospital cardiac arrest. Crit Care. (2017) 21:96. doi: 10.1186/s13054-017-1677-2
18. Tolins, ML, Henning, DJ, Gaieski, DF, Grossestreuer, AV, Jaworski, A, and Johnson, NJ. Initial arterial carbon dioxide tension is associated with neurological outcome after resuscitation from cardiac arrest. Resuscitation. (2017) 114:53–8. doi: 10.1016/j.resuscitation.2017.03.006
19. Humaloja, J, Litonius, E, and Efendijev, I. Early hyperoxemia is not associated with cardiac arrest outcome. Resuscitation. (2019) 140:185–93. doi: 10.1016/j.resuscitation.2019.04.035
Keywords: cardiac arrest, neurological outcomes, carbon dioxide tension, mortality, meta-analysis
Citation: Wang D, Li Z, Qin G, Cai L, Zhang T, Zhang H, Liu Y and Zhu H (2025) A systematic review and meta-analysis of the association between arterial carbon dioxide tension and patient outcomes after cardiac arrest. Front. Med. 12:1687522. doi: 10.3389/fmed.2025.1687522
Edited by:
Lakshmi Raman, University of Texas Southwestern Medical Center, United StatesReviewed by:
Anne-Marie Guerguerian, University of Toronto, CanadaÖkkeş Zortuk, TC Saglik Bakanligi Bandirma Egitim ve Arastirma Hastanesi, Türkiye
Copyright © 2025 Wang, Li, Qin, Cai, Zhang, Zhang, Liu and Zhu. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Yecheng Liu, cHRjYWxpdUBzaW5hLmNvbQ==; Huadong Zhu, ZHJ6aHVoZEBzaW5hLmNvbQ==