Cardiac arrest (CA), including in-hospital cardiac arrest (IHCA) and out-of-hospital cardiac arrest (OHCA), is a significant public health issue. In the United States, more than 600,000 people suffer from cardiac arrest each year^{1, 2}, and the incidence of cardiac arrest has increased over recent years³. Patients who have suffered from cardiac arrest have high mortality attributed to ischemia hypoxia and reperfusion injury, and those who survive often present with long-term neurological dysfunction and significantly poor prognosis⁴. Therefore, early screening for neurological symptoms should be carried out promptly in patients with cardiac arrest.

Relevant guidelines indicate a multimodal neuro-prognostication strategy to predict neurological outcomes, including measuring blood levels of neuron-specific enolase⁵. NSE is a biomarker that is part of the multimodal evaluation of the post-resuscitation neurological prognosis, recognized in the resuscitation guidelines, but not routinely tested on admission. Hemoglobin, which primarily consists of erythrocytes circulating in the blood, is a significant oxygen transport protein⁶. It has been reported that serum level of hemoglobin correlated with poor neurologic outcomes⁷. Given that serum level of hemoglobin testing is simple and inexpensive, we conducted a meta-analysis to determine the impact of serum level of hemoglobin as a prognostic blood biomarker for neurological outcomes in post-cardiac arrest patients.

This study was carried out according to the principles outlined in the A 24-step guide on how to design, conduct, and successfully publish a systematic review and meta-analysis in medical research⁸ and the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines⁹. Our research question was designed considering study population, intervention, comparison, and outcome (PICO) as follows: population (P) = post-cardiac arrest adult patients; intervention (I) = serum Hb level; comparator (C) = none; outcome (O) = neurological outcomes.

Search strategy

All comprehensive articles published before April 2023 that estimated the neurological prognostic effect of Serum level of hemoglobin in adult patients with cardiac arrest were searched in PubMed, Web of Science, Cochrane Library, and Embase databases by two experienced reviewers (Hong-xiang Hou and Liang Zhao). The reference lists of eligible studies were also searched to identify any studies that were not identified in the initial search.

The following search terms were used: “arrest, heart” or “cardiac arrest” or “arrest, cardiac” or “asystole” or “asystoles” or “cardiopulmonary arrest” or “arrest, cardiopulmonary” or “Ventricular Fibrillation” or “Fibrillation, Ventricular” or “Fibrillations, Ventricular” or “Ventricular Fibrillations” or “Tachycardia” or “Tachycardias or “Tachyarrhythmia” or “Tachyarrhythmias” or “Death, Sudden, Cardiac” or “Sudden Cardiac Death” or “Cardiac Death, Sudden” or “Death, Sudden Cardiac” or “Cardiac Sudden Death” or “Death, Cardiac Sudden” or “Sudden Death, Cardiac” or “Sudden Cardiac Arrest” or “Arrest, Sudden Cardiac” or “Cardiac Arrests, Sudden” or “Cardiac Arrest, Sudden” and “hemoglobin” or “Eryhem” or “ferrous, hemoglobin” or “hemoglobin, ferrous” and “prognoses” or “prognostic factors” or “factor, prognostic”. Study retrieval is shown in Supplementary Table 1.

Two authors carefully reviewed the title and abstract of all articles and independently scanned all articles based on predefined inclusion and exclusion criteria.

Inclusion criteria: (1) the included population was over 18 years of age; (2) non-traumatic cardiac arrest; (3) serum level of hemoglobin were detected; and (4) articles using the cerebral performance categories (CPC) scale to evaluate neurological outcomes.

Exclusion criteria: (1) age < 18 years; (2) patients with traumatic cardiac arrest; (3) intracranial bleeding; and (4) irrelevant topic, letters, duplicate data or publications, reviews, meta-analyses, case reports, comments, editorials, or animal experiments.

Quality assessment

Two reviewers independently used the Newcastle–Ottawa Scale (NOS) for non-randomized studies to assess the quality of the included studies¹⁰.The NOS consists of eight items that were divided into three domains: cohort selection, comparability, and outcome assessment. Grading the quality of evidence and strength of recommendations using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) method based on risk of bias, inconsistency, indirectness, imprecision, and publication bias^{11, 12}. At the same time, we use the Quality in Prognostic Studies (QUIPS) tool¹³ to assess the risk of bias. The tool consists of six domains: study participation, study attrition, prognostic factor measurement, outcome measurement, study confounding, statistical analysis and reporting. A low, moderate or high risk of bias was assessed for each domain. The original study was reevaluated by a third author when differences arose between the primary reviewers.

Data extraction

Two independent researchers extracted the relevant data for patients from all eligible studies. Any discordant assessments were resolved by a third investigator. The extracted variables were as follows: first author^’s name, year of publication, country in which the study was conducted, geographic location, inclusion period, study type, sample size, cardiac arrest type (OHCA vs. IHCA), poor neurologic outcome (PNO), serum hemoglobin sampling time, assessment tool of outcome measurement, point of outcome measurement, mean ± standard deviation (± SD), Hb level, age, male, and CPR duration. If the mean ± SD was not available, interquartile range and median^,,16 were converted into mean ± SD, using the method established by Wan et al.¹⁷. Based on the CPC score, neurological outcomes were defined as good or poor.

Statistical analysis

The association between serum level of hemoglobin and neurological outcome was estimated for every study using the standardized mean difference (SMD) and 95% confidence interval (CI). Serum hemoglobin unit measurements differed between studies. We tried to convert it into uniform units, but we could not achieve it because only the mean value was provided in the article, so we chose to use SMD. The Cochrane Q test (p < 0.10) and the I² statistic were used to assess heterogeneity among the included studies. The star chart also was used to test heterogeneity. The heterogeneity results of PQ < 0.1 and/or I² > 50% were considered high heterogeneity using the random effects models, otherwise PQ ≥ 0.1 and/or I² ≤ 50% were considered as statistically significant heterogeneity using the fixed-effects model^{18, 19}. A funnel plot and Egger^’s linear regression test were used to assess publication bias. Sensitivity analysis was used to judge robustness. Stata version 16.0 was used to perform all analyses.

Study selection

Using the established study retrieval method, 1257 studies were identified in the database search: 151 studies were duplicates and 1073 irrelevant publications without cardiac arrest or/and neurological outcomes were excluded after reviewing titles and abstracts. After full-text articles were assessed for eligibility, 11 publications were included in the meta-analysis^{7, ,,16, ,,,,,,26}. The screening process is shown in Fig. 1.

Flow diagram of studies included in the meta-analysis.

Full size image

Document quality assessment and data extraction

Two studies were multi-center prospective observational studies (mPOS), and one study was a single-center prospective observational study (sPOS). The remaining eight studies were single-center retrospective observational studies (sROS). Seven studies only included OHCA patients, one study only included IHCA patients, and two studies included both OHCA and IHCA patients. One study did not clarify the CA type. All articles used CPC scores for neurologic outcomes. The baseline information of the included studies are presented in Table 1 and Supplementary Table 2. According to NOS analysis, quality scores ranged from 4 to 9, and five studies were rated as high quality. The GRADE protocol was used to assess the certainty of the evidence. The evidence for hemoglobin as a prognostic marker was rated as very low due to risk of bias or evidence of publication bias (Supplementary Fig. 11). QUIPS has been used for the evaluation of prognostic studies. The overall results of the quality assessment are shown in Supplementary Fig. 12.

Meta-analysis

The meta-analysis included 11 studies with a total of 2519 cardiac arrest patients including 647 patients with a good neurological prognosis and 1872 patients with a poor neurological prognosis. Articles in this study had heterogeneity values of I² = 59.7% and Q test P < 0.1, suggesting that there was heterogeneity among the literatures selected for this study. The random effect model was applied to our meta-analysis (Fig. 2). This analysis showed that serum level of hemoglobin in the good prognosis group was 0.60 higher than that in the poor prognosis group, and this difference was statistically significant (P < 0.05). A star chart was used to further investigate heterogeneity (Fig. 3). This analysis indicated the studies by Kei Hayashida et al. and Chih-Hung Wang et al. affected the heterogeneity; removing these two articles from the analysis using the random effect model improved the heterogeneity (Fig. 4). Nine studies with a total of 1598 cardiac arrest patients including a total of 518 patients with good neurological prognosis and 1080 patients with poor neurological prognosis. The SMD value of the 1589 patients was 0.6, with a 95%CI of 0.49–0.71 (Z = 10.85, P < 0.05), suggesting that higher serum level of hemoglobin may be a relevant factor for better neurological prognosis. Sensitivity analysis was conducted on the remaining nine articles (Supplementary Fig. 1). Figure 5 shows that the funnel plot analysis was symmetrical with a P-value > 5 after testing for publication bias. The Begg’s and Egger’s tests (Supplementary Fig. 2) further suggest there was no publication bias in the nine studies. In the regression analysis (Supplementary Fig. 3), Cardiac arrest type was considered the source of heterogeneity. Therefore, subgroup analysis was conducted according to cardiac arrest type (Fig. 6). We found that cardiac arrest type affected the results of the meta-analysis. The results of other subgroup analysis regarding race, location, research type, serum hemoglobin sampling time, and outcome measurement time point are provided in Supplementary Figs. 4–7. The results indicate that serum hemoglobin sampling time was a source of heterogeneity. We also analysed the relationship between bystander CPR, initial shockable rhythm, time to ROSC and neurological prognosis for which data were available(Supplementary Figs. 8–10). Between Initial shockable rhythm, time to ROSC and neurological prognosis had heterogeneity. Bystander CPR may improve neurological prognosis in patients survived after cardiac arrest.

Forest plot of 11 studies. favorable: good neurological prognosis group; unfavorable: poor neurological prognosis group.

Full size image

Star chart.

Full size image

Relationship between serum level of hemoglobin and neurological outcomes. favorable: good neurological prognosis group; unfavorable: poor neurological prognosis group.

Full size image

Funnel plot.

Full size image

Subgroup analysis. favorable: good neurological prognosis group; unfavorable: poor neurological prognosis group OHCA: out-of-hospital cardiac arrest; IHCA: in-hospital cardiac arrest.

Full size image

Hemoglobin is key factor of blood oxygen-carrying capacity, and as such serum hemoglobin correlates with the prognosis of various diseases. There is accumulating evidence demonstrating that serum hemoglobin concentration is associated with poor neurological outcome in a variety of brain injuries²⁷, including traumatic brain injury (TBI), subarachnoid hemorrhage (SAH), stroke, and hemorrhage^,,,31. A meta-analysis demonstrated that serum hemoglobin concentration may be associated with mortality after transcatheter aortic valve implantation³². A retrospective study demonstrated that serum albumin and serum level of hemoglobin at admission predict mortality in children with TBI³³. A previous study has shown that higher serum level of hemoglobin at admission were involved in better outcomes in patients with spontaneous, nontraumatic intracerebral hemorrhage³⁴ The literature indicates that risk of progression of IgA nephropathy decreases with increases in serum level of hemoglobin³⁵. Further evidence suggests that a decline in serum level of hemoglobin by ≥ 3 g/dl is related to an increased risk of mortality in patients with acute coronary syndromes³⁶. An earlier study demonstrated that elevating serum hemoglobin concentration at admission may reduce the risk of death after discharge in patients with acute exacerbation of chronic obstructive pulmonary disease³⁷. It has been shown that anemic mice have a worse prognosis after traumatic brain injury compared to non-anemic mice, the exact mechanism of which is unclear³⁸. A meta-analysis by Mori and colleagues suggested that serum level of hemoglobin were associated with overall mortality and disease progression in patients with metastatic hormone-sensitive prostate cancer³⁹. Another meta-analysis revealed that stroke patients who present with anemia from the onset have a higher risk of mortality⁴⁰. A recent study has shown that lower serum level of hemoglobin at admission may predict extent of kidney damage in patients with type 2 diabetes⁴¹. To the best of our knowledge, our meta-analysis is the first to investigate the association of serum level of hemoglobin with neurological outcomes in patients following cardiac arrest. This meta-analysis showed that higher serum hemoglobin levels may improve neurological prognosis in patients who have survived after cardiac arrest. Our findings are consistent with those of Cavicchi et al.¹⁵. It was also reported that patients following OHCA who had higher serum level of hemoglobin achieved full neurological recovery⁴². However, because the data was incomplete in that study it was not selected for inclusion in the current meta-analysis⁴² Another study reported that anemia is a risk factor for cardiac arrest, which was also a general conclusion in this meta-analysis²⁴. However, Tran et al. published in 2020 that there was no significant correlation between serum level of hemoglobin and neurological prognosis of IHCA patients⁴³. Our heterogeneity and subgroup analysis contradict the findings of Tran et al., in that we found that high serum level of hemoglobin can improve neurological prognosis in OHCA, IHCA and OHCA + IHCA patients. Therefore, serum hemoglobin testing may help physicians and patients choose more appropriate treatments.

This meta-analysis has several limitations that should be noted. First, the number of included is small and subgroup analysis was prone to error. Second, some studies had high heterogeneity, and subgroup analysis indicated that cardiac arrest type and serum hemoglobin sampling time might be the sources of heterogeneity. All included studies were cohort studies (only two studies were multicenter studies). Thus, we need large multicentre prospective cohort type of study to confirm the results of this meta-analysis. To reduce heterogeneity, more data from other ethnic groups or countries should be included. Thirdly, the results may vary depending on the time point of CPC measurement⁴⁴.

1. Andersen, L. W., Holmberg, M. J., Berg, K. M., Donnino, M. W. & Granfeldt, A. In-hospital cardiac arrest: A review. Jama 321(12), 1200–1210 (2019).

   Article  PubMed  PubMed Central  Google Scholar 

2. Benjamin, E. J. et al. Heart disease and stroke statistics-2019 update: A report from the american heart association. Circulation 139(10), e56–e528 (2019).

   Article  PubMed  Google Scholar 

3. Holmberg, M. J. et al. Annual incidence of adult and pediatric in-hospital cardiac arrest in the United States. Circ. Cardiovasc. Qual Outcomes 12(7), e005580 (2019).

   Article  PubMed  PubMed Central  Google Scholar 

4. Dyson, K. et al. International variation in survival after out-of-hospital cardiac arrest: A validation study of the Utstein template. Resuscitation 138, 168–181 (2019).

   Article  PubMed  Google Scholar 

5. Nolan, J. P. et al. European resuscitation council and european society of intensive care medicine guidelines 2021: Post-resuscitation care. Intensive Care Med. 47(4), 369–421 (2021).

   Article  PubMed  PubMed Central  Google Scholar 

6. Li, C., Sun, T. & Jiang, C. Recent advances in nanomedicines for the treatment of ischemic stroke. Acta Pharm. Sin. B 11(7), 1767–1788 (2021).

   Article  CAS  PubMed  Google Scholar 

7. Bouglé, A. et al. Determinants and significance of cerebral oximetry after cardiac arrest: A prospective cohort study. Resuscitation 99, 1–6 (2016).

   Article  PubMed  Google Scholar 

8. Muka, T. et al. A 24-step guide on how to design, conduct, and successfully publish a systematic review and meta-analysis in medical research. Eur. J. Epidemiol. 35(1), 49–60 (2020).

   Article  PubMed  Google Scholar 

9. Liberati, A. et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: Explanation and elaboration. Bmj 339, b2700 (2009).

   Article  PubMed  PubMed Central  Google Scholar 

10. Lepard, J. R. et al. Differences in outcomes of mandatory motorcycle helmet legislation by country income level: A systematic review and meta-analysis. PLoS Med. 18(9), e1003795 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

11. Guyatt, G. H., Oxman, A. D., Schünemann, H. J., Tugwell, P. & Knottnerus, A. GRADE guidelines: A new series of articles in the journal of clinical epidemiology. J. Clin. Epidemiol. 64(4), 380–382 (2011).

    Article  PubMed  Google Scholar 

12. Atkins, D. et al. Grading quality of evidence and strength of recommendations. Bmj 328(7454), 1490 (2004).

    Article  PubMed  Google Scholar 

13. Hayden, J. A., van der Windt, D. A., Cartwright, J. L., Côté, P. & Bombardier, C. Assessing bias in studies of prognostic factors. Ann. Intern. Med. 158(4), 280–286 (2013).

    Article  PubMed  Google Scholar 

14. Albaeni, A. et al. The association between post resuscitation hemoglobin level and survival with good neurological outcome following out of hospital cardiac arrest. Resuscitation 99, 7–12 (2016).

    Article  PubMed  Google Scholar 

15. Zama Cavicchi, F. et al. Low hemoglobin and venous saturation levels are associated with poor neurological outcomes after cardiac arrest. Resuscitation 153, 202–208 (2020).

    Article  PubMed  Google Scholar 

16. Cho, Y. M. et al. Blood ammonia is a predictive biomarker of neurologic outcome in cardiac arrest patients treated with therapeutic hypothermia. Am. J. Emerg. Med. 30(8), 1395–1401 (2012).

    Article  PubMed  Google Scholar 

17. Wan, X., Wang, W., Liu, J. & Tong, T. Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med. Res. Methodol. 14, 135 (2014).

    Article  PubMed  PubMed Central  Google Scholar 

18. Higgins, B. T., Barlow, D. R., Heagerty, N. E. & Lin, T. J. Anterior vs. posterior approach for total hip arthroplasty, a systematic review and meta-analysis. J. Arthroplasty 30(3), 419–434 (2015).

    Article  PubMed  Google Scholar 

19. Huedo-Medina, T. B., Sanchez-Meca, J., Marin-Martinez, F. & Botella, J. Assessing heterogeneity in meta-analysis: Q statistic or I-2 index?. Psychol. Methods 11(2), 193–206 (2006).

    Article  PubMed  Google Scholar 

20. SOS-Kanto Study Group. Relationship between the hemoglobin level at hospital arrival and post-cardiac arrest neurologic outcome. Am. J. Emerg. Med. 30(5), 770–774 (2012).

    Article  Google Scholar 

21. Ameloot, K. et al. Low hemoglobin levels are associated with lower cerebral saturations and poor outcome after cardiac arrest. Resuscitation 96, 280–286 (2015).

    Article  CAS  PubMed  Google Scholar 

22. Wormsbecker, A., Sekhon, M. S., Griesdale, D. E., Wiskar, K. & Rush, B. The association between anemia and neurological outcome in hypoxic ischemic brain injury after cardiac arrest. Resuscitation 112, 11–16 (2017).

    Article  PubMed  Google Scholar 

23. Wang, C. H. et al. Association between hemoglobin levels and clinical outcomes in adult patients after in-hospital cardiac arrest: A retrospective cohort study. Intern. Emerg. Med. 11(5), 727–736 (2016).

    Article  PubMed  Google Scholar 

24. Oh, S. J. et al. Age is related to neurological outcome in patients with out-of-hospital cardiac arrest (OHCA) receiving therapeutic hypothermia (TH). Am. J. Emerg. Med. 36(2), 243–247 (2018).

    Article  PubMed  Google Scholar 

25. Kim, D. et al. Hemoglobin concentration is associated with neurologic outcome after cardiac arrest in patients treated with targeted temperature management. Clin. Exp. Emerg. Med. 5(3), 150–155 (2018).

    Article  PubMed  PubMed Central  Google Scholar 

26. Hayashida, K. et al. Estimated cerebral oxyhemoglobin as a useful indicator of neuroprotection in patients with post-cardiac arrest syndrome: A prospective, multicenter observational study. Crit. Care 18(4), 500 (2014).

    Article  PubMed  PubMed Central  Google Scholar 

27. Lelubre, C., Bouzat, P., Crippa, I. A. & Taccone, F. S. Anemia management after acute brain injury. Crit. Care 20(1), 152 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

28. Oddo, M. et al. Anemia and brain oxygen after severe traumatic brain injury. Intensive Care Med. 38(9), 1497–1504 (2012).

    Article  CAS  PubMed  Google Scholar 

29. Ayling, O. G. S., Ibrahim, G. M., Alotaibi, N. M., Gooderham, P. A. & Macdonald, R. L. Anemia after aneurysmal subarachnoid hemorrhage is associated with poor outcome and death. Stroke 49(8), 1859–1865 (2018).

    Article  PubMed  Google Scholar 

30. Tang, X. et al. Iron-deficiency and estrogen are associated with ischemic stroke by up-regulating transferrin to induce hypercoagulability. Circ. Res. 127(5), 651–663 (2020).

    Article  CAS  PubMed  Google Scholar 

31. Zeng, Y. J. et al. Anemia on admission increases the risk of mortality at 6 months and 1 year in hemorrhagic stroke patients in China. J. Stroke Cerebrovasc. Dis. 23(6), 1500–1505 (2014).

    Article  PubMed  Google Scholar 

32. Takagi, H., Hari, Y., Kawai, N. & Ando, T. Meta-analysis of impact of anemia and hemoglobin level on survival after transcatheter aortic valve implantation. Am. J. Cardiol. 123(2), 306–314 (2019).

    Article  PubMed  Google Scholar 

33. Luo, H. C., Fu, Y. Q., You, C. Y., Liu, C. J. & Xu, F. Comparison of admission serum albumin and hemoglobin as predictors of outcome in children with moderate to severe traumatic brain injury: A retrospective study. Medicine 98(44), e17806 (2019).

    Article  PubMed  PubMed Central  Google Scholar 

34. Acosta, J. N. et al. Admission hemoglobin levels are associated with functional outcome in spontaneous intracerebral hemorrhage. Crit. Care Med. 49(5), 828–837 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

35. Oh, T. R., Song, S. H., Choi, H. S., Kim, C. S., Han, S. H., Kang, K. P., Kwon, Y. J., Kim, S. W., Ma, S. K., Bae, E. H., The association between serum hemoglobin and renal prognosis of IgA nephropathy. J. Clin. Med., 10(2), (2021).

36. Leonardi, S. et al. Prognostic implications of declining hemoglobin content in patients hospitalized with acute coronary syndromes. J. Am. Coll. Cardiol. 77(4), 375–388 (2021).

    Article  CAS  PubMed  Google Scholar 

37. Toft-Petersen, A. P., Torp-Pedersen, C., Weinreich, U. M. & Rasmussen, B. S. Association between hemoglobin and prognosis in patients admitted to hospital for COPD. Int. J. Chron. Obstruct. Pulmon Dis. 11, 2813–2820 (2016).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

38. Litofsky, N. S. et al. Anaemia worsens early functional outcome after traumatic brain injury: A preliminary study. Brain Inj. 32(3), 342–349 (2018).

    Article  PubMed  Google Scholar 

39. Mori, K. et al. Prognostic value of hemoglobin in metastatic hormone-sensitive prostate cancer: A systematic review and meta-analysis. Clin. Genitourin. Cancer 18(4), e402–e409 (2020).

    Article  PubMed  Google Scholar 

40. Barlas, R. S. et al. Impact of hemoglobin levels and anemia on mortality in acute stroke: Analysis of UK regional registry data, systematic review, and meta-analysis. J. Am. Heart Assoc. 5(8), e003019 (2018).

    Article  Google Scholar 

41. Song, H. Y., Wei, C. M., Zhou, W. X., Hu, H. F. & Wan, Q. J. Association between admission hemoglobin level and prognosis in patients with type 2 diabetes mellitus. World J. Diabetes 12(11), 1917–1927 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

42. Iiya, M. et al. Combination of hemoglobin and low-flow duration can predict neurological outcome in the initial phase of out-of-hospital cardiac arrest. J. Crit. Care 47, 269–273 (2018).

    Article  CAS  PubMed  Google Scholar 

43. Tran, L. N. et al. The association between post-cardiac arrest cerebral oxygenation and survival with favorable neurological outcomes: A multicenter study. Resuscitation 154, 85–92 (2020).

    Article  PubMed  Google Scholar 

44. Scheel, M. et al. The prognostic value of gray-white-matter ratio in cardiac arrest patients treated with hypothermia. Scand. J. Trauma Resusc. Emerg. Med. 21, 23 (2013).

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Authors and Affiliations

1. Department of Emergency, the First Hospital of Jilin University, Xinmin Street 1, Chaoyang District, Changchun, China

   Hongxiang Hou, Li Pang, Zuolong Liu & Ji-Hong Xing

2. Rehabilitation Department, the First Hospital of Jilin University, Changchun, China

   Liang Zhao

Contributions

H.-X.H. designed the study, and conceived the review. H.-X.H. and L.Z. performed the searches and screened studies for eligibility. H.-X.H. and L.Z. assessed the quality of the papers, H.-X.H. and Z.-L.L. executed statistical analysis. H.-X.H. drafted the manuscript, and L.P. contributed substantially to its revision. J.H.X. and H.H. takes responsibility for the paper as a whole.

Corresponding author

Correspondence to Ji-Hong Xing.

Competing interests

The authors declare no competing interests.

Publisher's note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions