Skip to main content
Download PDF

Role of age as eligibility criterion for ECMO in patients with ARDS: meta-regression analysis

Critical Care volume 28, Article number: 278 (2024) Cite this article

Abstract

Background

Age as an eligibility criterion for V-V ECMO is widely debated and varies among healthcare institutions. We examined how age relates to mortality in patients undergoing V-V ECMO for ARDS.

Methods

Systematic review and meta-regression of clinical studies published between 2015 and June 2024. Studies involving at least 6 ARDS patients treated with V-V ECMO, with specific data on ICU and/or hospital mortality and patient age were included. The search strategy was executed in PubMed, limited to English-language. COVID-19 and non-COVID-19 populations were analyzed separately. Meta-regressions of mortality outcomes on age were performed using gender, BMI, SAPS II, APACHE II, Charlson comorbidity index or SOFA as covariates.

Results

In non-COVID ARDS, the meta-regression of 173 studies with 56,257 participants showed a significant positive association between mean age and ICU/hospital mortality. In COVID-19 ARDS, a significant relationship between mean age and ICU mortality, but not hospital mortality, was found in 103 studies with 21,255 participants. Sensitivity analyses confirmed these findings, highlighting a linear relationship between age and mortality in both groups. For each additional year of mean age, ICU mortality increased by 1.2% in non-COVID ARDS and 1.9% in COVID ARDS.

Conclusions

The relationship between age and ICU mortality is linear and shows no inflection point. Consequently, no age cut-off can be recommended for determining patient eligibility for V-V ECMO.

Introduction

Acute Respiratory Distress Syndrome (ARDS) is a critical condition with a mortality rate exceeding 40% [1]. Despite being the recommended life-saving intervention when conventional treatments fail, standardized selection criteria for veno-venous extracorporeal membrane oxygenation (V-V ECMO) are still lacking. [2]. Although common clinical practice takes age into consideration as selection criteria for ECMO [3], the Extracorporeal Life Support Organization (ELSO) refrains from establishing a specific age threshold due to the variability in mortality rates and age cut-offs across predictive scoring systems and clinical trials [2].

The lack of a specific age limit may be explained by the observation that mortality rates among older patients may range from less than 50% to over 80% [3, 4]. Additionally, a recent randomized clinical trial did not enforce an upper age limit for ECMO eligibility [5]. Furthermore, recent scoring systems incorporate different age thresholds to forecast survival among V-V ECMO patients [6].

Meta-regression analysis is a statistical technique that can be useful to explore sources of heterogeneity and to examine the relationship between study characteristics and the effect sizes reported in the included studies [7]. We conducted a systematic review and meta-regression analysis to describe the relationship between age and mortality in patients with ARDS treated with V-V ECMO.

Methods

This systematic review and meta-analysis of clinical studies was conducted in accordance with recent recommendations [8, 9]. We included studies that met the following criteria: (a) studies (observational, case series, and RCTs) had to include patients in whom ARDS was diagnosed according to the Berlin definition [10] and were treated with V-V ECMO; (b) studies had to enroll ≥ 6 patients; (c) studies had to report age and mortality [Intensive Care Unit (ICU) or hospital]. We conducted a literature search in the PubMed database including studies published between January 1st, 2015, and June 30th, 2024 (last access July 5th, 2024). The following search terms were used: “ECMO” OR “extracorporeal membrane oxygenation” OR “extracorporeal oxygenation”. Articles were identified and assessed for eligibility independently by five of the authors (LB, MB, RdA, DN and ISZ). The database search was restricted to English language. The whole process was supervised by another author (TT), who was consulted also in cases of disagreement about eligibility. The methodological quality of the studies was independently assessed by two of the authors (LB and MB) using previously described methods (https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools). In case of disagreement, a third author (TT) was consulted.

For articles reporting individual patients’ data, mean age and median length of stay were computed. For articles that reported median age and range (min–max) or interquartile range (1st quartile – 3rd quartile), mean age was calculated using appropriate Wan’s formula (see Additional methods, online supplement). Given the specific clinical characteristics and logistic challenges during the COVID-19 pandemic, the "conventional ARDS" and "COVID-19 ARDS" subgroups were examined separately. When papers reported information on subgroups "conventional ARDS" and "COVID-19 ARDS", each subgroup was included in the corresponding analysis.

Meta-regression results for mortality vs. age were reported using the b-regression coefficient. Gender, body mass index (BMI), standardized severity scores at admission [simplified acute physiology score (SAPS II) or acute physiology and chronic health evaluation (APACHE II), sequential organ failure assessment (SOFA) and Charlson comorbidity index] were used as covariates. Severity scores were standardized by dividing the mean by the standard deviation. Sensitivity analyses were performed taking into considerations: (a) the quality of studies; (b) the studies reporting only mean age; (c) the possible non-linear relationship between age and outcomes.

The Stata command “metaprop” was used to perform meta-analyses of proportions and to estimate statistical heterogeneity between studies. A random-effects model using the Der Simonian and Laird method (14) was used to estimate mortality using proportions and 95% confidence intervals (95% CI) and to produce forest plots. Meta-regression was used to investigate whether heterogeneity between studies could be explained by one or more study characteristics [11]. The significance level was set at p < 0.05. All statistical analyses were performed using Stata version 15 (StataCorp, 2017. Stata statistical software: Release 15. College Station, TX). Additional methods are presented in the online supplement.

Results

Figure E1(online supplement) presents the results of the search and reasons for exclusion. The 261 eligible studies included 77,512 participants. Individual studies with the variables of interest are presented in Table E1 (non-COVID) and Table E2 (COVID). Fifteen studies were included in both tables because they reported information on subgroups of non-COVID and COVID patients.

Fig. 1
figure 1

Meta-regression of ICU mortality on mean age reported by authors or estimated using Wan’s formula when only median and IQR was reported in studies including non-COVID-19 patients (panel A) and COVID-19 patients (panel B). Each study is represented by a circle sized according to the study weight. The red line represents the linear regression estimate

Full size image

Non-COVID ARDS

We included 173 studies with a total of 56,257 participants (2 RCTs, 2 case series and 169 observational). ICU mortality was reported in 51 studies (29%); in-hospital mortality was reported in 137 studies (79%). Mean age was available for 82 studies, while 91 studies reported the median and inter-quartile range (Table E1, online supplement). In studies reporting ICU mortality, mean age ranged between 29 and 73 years, and rate of death between 16 and 100%; heterogeneity (I2) amounted to 93%. A positive significant association was found between mean age and ICU mortality (b = 0.012; SE(b): 0.003; p = 0.001) (Fig. 1, panel A and Table E3, online supplement). Variability in mean age explained part of the existing heterogeneity (residual heterogeneity I2 = 58% and R2 = 27%). Multiple meta-regressions showed that the association between age and outcome remained significant after adjusting for all other covariates, except for models adjusted for BMI that included only 12 studies and 2252 participants (Table E3).

In studies reporting in-hospital mortality, mean age ranged between 29 and 73 years, and rate of death between 0 and 100%; heterogeneity amounted to 92%. Meta-regression showed a positive significant association between patients’ mean age and in-hospital mortality (b = 0.008; SE(b): 0.002; p < 0.001) (Fig. 2, panel A and Table E3, online supplement). Variability in age explained part of the existing heterogeneity (residual heterogeneity I2 = 60%). In this case, the association between age and outcome remained significant after adjusting for all the other covariates (Table E3).

Fig. 2
figure 2

Meta-regression of in-hospital mortality on mean age reported by authors or estimated using Wan’s formula when only median and IQR was reported in studies including non-COVID-19 patients (panel A) and COVID-19 patients (panel B). Each study is represented by a circle sized according to the study weight. The red line represents the linear regression estimate

Full size image

COVID-19 ARDS

We included 103 studies conducted between 2020 and 2024 with a total of 21,255 participants (1 RCT, 5 case series and 97 observational). Mean age was available for 45 studies, while 58 studies reported the median and inter-quartile range (Table E2, online supplement). ICU mortality was reported in 26 studies (25%), while in-hospital mortality was reported in 89 studies (86%) (Table E2, online supplement); mean age ranged between 34 and 67 years. In studies reporting ICU mortality, rate of death ranged between 5 and 80% with I2 equal to 91%. In studies reporting in-hospital mortality, mortality ranged between 17 and 75% and I2 was 82.8%. No significant association was found between mean patients age and in-hospital mortality (Fig. 2, panel B and Table E4, online supplement). Vice versa, meta-regression showed a significant relationship between patients’ mean age and ICU mortality (b = 0.018; SE(b): 0.008; p = 0.040) (Fig. 1, panel B and Table E4, online supplement). The association between age and outcome remained significant, after adjusting for selected covariates, except when adjusting for severity. This latter model included only 4 studies and 138 participants (Table E4).

Sensitivity analyses

Results of the sensitivity analyses based on good and fair-quality studies (quality assessment process is summarized in Tables E5a-b-c and E6a-b-c, online supplement) confirmed the findings in non-COVID and COVID patients (Tables E7 and E8 and Figure E2, online supplement); sensitivity analyses based on studies reporting only mean age showed that only the relationship between in-hospital mortality and age in non-COVID population remained significant (Tables E7 and E8 and Figure E3, panel C, online supplement); lastly, analyses exploring the possible non-linear relationship between age and outcomes confirmed that the relationships between ICU mortality and in-hospital mortality with age in the two groups were not accounted for by higher-order (quadratic or cubic) functions (TableS E7 and E8).

Discussion

Results of the present study indicate that in ARDS patients treated with V-V ECMO (both COVID and non-COVID), ICU mortality increases linearly with age without a clear age cut-off that would contraindicate V-V ECMO.

Age is a critical factor in determining eligibility for ECMO in severe ARDS due to its resource-intensive nature and associated risks. Nevertheless, while age is often factored into clinical decision-making as a criterion for ECMO selection [6], the existing data supporting a precise age threshold for V-V ECMO indication remain sparse and contradictory, precluding the establishment of a specific age cutoff [2]. In fact, although a recent study reported a non-linear age-mortality relationship in patients with ARDS [12], retrospective analyses have shown varying survival rates across different age groups [3], with some studies suggesting “acceptable” outcomes even in patients over 65 years of age [4]. Moreover, (a) a recent RCT has included older patients without observing significant differences in outcomes compared to younger cohorts [5]; (b) although age is included in mortality prediction scores for ECMO, variations exist in the age classes used [6, 13] and even very young patients receiving V-V ECMO exhibit age-related outcome differences [14]. Our meta-regression analysis showed a linear relationship between age and ICU/in-hospital mortality across COVID and non-COVID studies. These data confirm a very recent registry-based cohort study that demonstrated a linear relationship between age and in-hospital mortality and an increase in post-ECMO complications starting as early as 30 years of age [15].

Strengths of this study include its comprehensive search, robust methodology, and application of meta-regression to generate evidence from pooled data, that may represent a reasonable option to generate robust evidence from pooled data when the inherent sample heterogeneity may limit the design and conduct of clinical trials in the critical care settings [16]. Limitations include the predominance of observational studies and the limited number of studies including elderly patients, posing challenges in defining a clear age threshold. Moreover, although the utmost care has been taken in excluding papers presenting data from the same databases (i.e., duplications or overlaps), we cannot totally exclude the possibility that a small number of patients may be present as duplicates in more than one included study. Another limitation of this study is the lack of the a priori registration of the protocol in PROSPERO.

In 2022, more than 20% of the EU population aged 65 and over. In the United States, approximately 17% of the nation's population (more than 56 million) is aged 65 and older. Consequently, the number of elderly patients admitted to intensive care units (ICUs) is expected to increase [17]. Under these circumstances, data regarding the outcome of older patients treated with advanced forms of artificial support such as ECMO may optimize the care of these patients in terms of targeting resources to those patients who are most likely to benefit [18]. Severity of the acute condition and age-related decline of resilience to stress are the most relevant determinants of critically ill geriatric patients [19]. However, the aging process is characterized by substantial inter-individual heterogeneity of such decline. Some studies suggest that age alone should not be a criterion for refusing prolonged mechanical ventilation in patients over 70 [20].

Under these circumstances, the observed linear relationship between age and mortality in ARDS patients (both COVID and non-COVID) receiving V-V ECMO treatment confirms that, although age plays a significant role in determining outcomes, individual variations in resilience or frailty may complicate simple age-based criteria. Furthermore, short-term outcomes like ICU or hospital mortality do not provide a comprehensive picture. Future research should focus on the impact of age and frailty on long-term outcomes, including functional status and quality of life in ICU survivors.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

References

  1. Bellani G, Laffey JG, Pham T, Fan E, Brochard L, Esteban A, Gattinoni L, van Haren F, Larsson A, McAuley DF, et al. Epidemiology, patterns of care, and mortality for patients with acute respiratory distress syndrome in intensive care units in 50 Countries. JAMA. 2016;315(8):788–800.

    Article  PubMed  CAS  Google Scholar 

  2. Tonna JE, Abrams D, Brodie D, Greenwood JC, Rubio Mateo-Sidron JA, Usman A, Fan E. Management of adult patients supported with venovenous extracorporeal membrane oxygenation (VV ECMO): guideline from the extracorporeal life support organization (ELSO). ASAIO J. 2021;67(6):601–10.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  3. Deatrick KB, Mazzeffi MA, Galvagno SM Jr, Tesoriero RB, Kaczoroswki DJ, Herr DL, Dolly K, Rabinowitz RP, Scalea TM, Menaker J. Outcomes of venovenous extracorporeal membrane oxygenation when stratified by age: How old is too old? ASAIO J. 2020;66(8):946–51.

    Article  PubMed  CAS  Google Scholar 

  4. Giani M, Forlini C, Fumagalli B, Rona R, Pesenti A, Foti G. Indication for venovenous extracorporeal membrane oxygenation: is 65 years old, too old? ASAIO J. 2021;67(1): e55.

    Article  PubMed  Google Scholar 

  5. Combes A, Hajage D, Capellier G, Demoule A, Lavoue S, Guervilly C, Da Silva D, Zafrani L, Tirot P, Veber B, et al. Extracorporeal membrane oxygenation for severe acute respiratory distress syndrome. N Engl J Med. 2018;378(21):1965–75.

    Article  PubMed  Google Scholar 

  6. Schmidt M, Zogheib E, Roze H, Repesse X, Lebreton G, Luyt CE, Trouillet JL, Brechot N, Nieszkowska A, Dupont H, et al. The PRESERVE mortality risk score and analysis of long-term outcomes after extracorporeal membrane oxygenation for severe acute respiratory distress syndrome. Intensive Care Med. 2013;39(10):1704–13.

    Article  PubMed  PubMed Central  CAS  Google Scholar 

  7. DerSimonian R, Laird N. Meta-analysis in clinical trials revisited. Contemp Clin Trials. 2015;45(Pt A):139–45.

    Article  PubMed  PubMed Central  Google Scholar 

  8. Higgins J, Eldridge S, Li Te: Chapter 23: Including variants on randomized trials. In: Cochrane Handbook for Systematic Reviews of Interventions version 63 (updated February 2022). Edited by Higgins J, Thomas J, Chandler J, Cumpston M, Li T, Page M, Welch V: Cochrane; 2022.

  9. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, Shamseer L, Tetzlaff JM, Akl EA, Brennan SE, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372: n71.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Ranieri VM, Rubenfeld GD, Thompson BT, Ferguson ND, Caldwell E, Fan E, Camporota L, Slutsky AS. Acute respiratory distress syndrome: the Berlin Definition. JAMA. 2012;307(23):2526–33.

    PubMed  Google Scholar 

  11. Higgins J, Thompson S, Deeks J, Altman D. Statistical heterogeneity in systematic reviews of clinical trials: a critical appraisal of guidelines and practice. J Health Serv Res Policy. 2002;7(1):51–61.

    Article  PubMed  Google Scholar 

  12. Patel BM, Reilly JP, Bhalla AK, Smith LS, Khemani RG, Jones TK, Meyer NJ, Harhay MO, Yehya N. Association between age and mortality in pediatric and adult acute respiratory distress syndrome. Am J Respir Crit Care Med. 2024;209(7):871–8.

    Article  PubMed  Google Scholar 

  13. Schmidt M, Bailey M, Sheldrake J, Hodgson C, Aubron C, Rycus PT, Scheinkestel C, Cooper DJ, Brodie D, Pellegrino V, et al. Predicting survival after extracorporeal membrane oxygenation for severe acute respiratory failure the respiratory extracorporeal membrane oxygenation survival prediction (RESP) score. Am J Respir Crit Care Med. 2014;189(11):1374–82.

    Article  PubMed  Google Scholar 

  14. Brown SR, Assy J, Anderson MP, Thiagarajan RR, Brogan TV. Outcomes after respiratory extracorporeal life support in teens and young adults: an extracorporeal life support organization registry analysis. Crit Care Med. 2024;52(1):11–9.

    Article  PubMed  Google Scholar 

  15. Fernando SM, Brodie D, Barbaro RP, Agerstrand C, Badulak J, Bush EL, Mueller T, Munshi L, Fan E, MacLaren G, et al. Age and associated outcomes among patients receiving venovenous extracorporeal membrane oxygenation for acute respiratory failure: analysis of the extracorporeal life support organization registry. Intensive Care Med. 2024;50(3):395–405.

    Article  PubMed  CAS  Google Scholar 

  16. Berlin JA, Golub RM. Meta-analysis as evidence: building a better pyramid. JAMA. 2014;312(6):603–5.

    Article  PubMed  CAS  Google Scholar 

  17. Chelluri L, Pinsky MR, Donahoe MP, Grenvik A. Long-term outcome of critically ill elderly patients requiring intensive care. JAMA. 1993;269(24):3119–23.

    Article  PubMed  CAS  Google Scholar 

  18. de Rooij SE, Govers A, Korevaar JC, Abu-Hanna A, Levi M, de Jonge E. Short-term and long-term mortality in very elderly patients admitted to an intensive care unit. Intensive Care Med. 2006;32(7):1039–44.

    Article  PubMed  Google Scholar 

  19. Mousai O, Tafoureau L, Yovell T, Flaatten H, Guidet B, Jung C, de Lange D, Leaver S, Szczeklik W, Fjolner J, et al. Clustering analysis of geriatric and acute characteristics in a cohort of very old patients on admission to ICU. Intensive Care Med. 2022;48(12):1726–35.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Dardaine V, Dequin PF, Ripault H, Constans T, Ginies G. Outcome of older patients requiring ventilatory support in intensive care: impact of nutritional status. J Am Geriatr Soc. 2001;49(5):564–70.

    Article  PubMed  CAS  Google Scholar 

Download references

Acknowledgements

The authors wish to dedicate this work to the memory of Elisa Maietti, bright woman and excellent statistician.

Funding

Italian Ministry of University and Research – Progetti di ricerca di Rilevante Interesse Nazionale (PRIN) 2017 – Project no. 2017J4BE7A.

Author information

Authors and Affiliations

  1. Department of Medical and Surgical Sciences (DIMEC), Alma Mater Studiorum – University of Bologna, Via Massarenti, 9, 40138, Bologna, Italy

    Tommaso Tonetti, Rossana Di Staso, Laura Bambini, Martina Bordini, Rosanna D’Albo, Domenico Nocera & Irene Sbaraini Zernini

  2. Anesthesiology and General Intensive Care Unit, IRCCS Azienda Ospedaliero-Universitaria di Bologna, Bologna, Italy

    Tommaso Tonetti

  3. Dipartimento di Medicina Sperimentale, Campus Ecotekne University of Salento, Lecce, Italy

    Luciana Mascia

  4. Cardio-Thoracic and Vascular Anesthesia and Intensive Care Unit, IRCCS Azienda Ospedaliero-Universitaria di Bologna, Bologna, Italy

    Ilaria Turriziani

  5. Department of Biomedical and Neuromotor Sciences (DIBINEM), Alma Mater Studiorum – University of Bologna, Bologna, Italy

    Paola Rucci

  6. Department of Anesthesia and Intensive Care Medicine, Università degli studi di Bari “Aldo Moro”, Ospedale Policlinico, Bari, Italy

    V. Marco Ranieri

Authors
  1. Tommaso Tonetti
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Rossana Di Staso
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Laura Bambini
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Martina Bordini
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Rosanna D’Albo
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Domenico Nocera
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Irene Sbaraini Zernini
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Ilaria Turriziani
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Luciana Mascia
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Paola Rucci
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. V. Marco Ranieri
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Concept and design: VMR, TT, PR. Data acquisition: RDS, MB, LB, RDA, DN, ISZ, IT, TT. Statistical analysis: RDS, PR. Data analysis and interpretation: RDS, PR, TT, VMR. Writing the first draft: TT, VMR, RDS, PR. Reviewing the manuscript for important intellectual content: LM, IT. Approval of manuscript final version: all authors.

Corresponding author

Correspondence to Tommaso Tonetti.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher's Note

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

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, 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 you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. 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-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Tonetti, T., Di Staso, R., Bambini, L. et al. Role of age as eligibility criterion for ECMO in patients with ARDS: meta-regression analysis. Crit Care 28, 278 (2024). https://doi.org/10.1186/s13054-024-05074-z

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13054-024-05074-z

Keywords

Critical Care

ISSN: 1364-8535