Poor survival with extracorporeal membrane oxygenation in acute respiratory distress syndrome (ARDS) due to coronavirus disease 2019 (COVID-19): Pooled analysis of early reports

Dear Editor,

The novel coronavirus disease 2019 (COVID-19) can induce acute respiratory distress syndrome (ARDS), which can progress to refractory pulmonary failure. In such cases, extracorporeal membrane oxygenation (ECMO) may be considered as a rescue therapy. In a study of ECMO for ARDS in patients with Middle East Respiratory Syndrome Coronavirus (MERS-CoV), a similar coronavirus disease emerged in 2012, a significant decrease of in-hospital mortality rate and length of intensive care unit (ICU) stay was found in patients treated with ECMO compared to those managed with conventional therapy [1]. However, with COVID-19, concerns have been raised about high mortality rate observed in an early report which included data on ECMO in infected patients [2]. It has been suggested that the compounded immunologic insult by both infection and extracorporeal circuit may counterbalance or even offset survival benefits [2]. In this article, we aimed to evaluate ECMO mortality as reported in early COVID-19 epidemiological studies.

An electronic search of Medline (PubMed interface), Scopus and Web of Science, was executed employing the keywords “mortality” OR “death” OR “ECMO” AND “coronavirus 2019” OR “COVID-19” OR “2019-nCoV” OR “SARS-CoV-2”, between 2019 and present time (i.e., March 13, 2020). No language restrictions were applied. The title, abstract and full text of all articles captured with these search criteria were assessed, and those reporting the rate of mortality in COVID-19 patients receiving ECMO were included in pooled analysis. The reference list of all identified studies was also analyzed (forward and backward citation tracking) to detect additional articles.

Data on the number of patients with ARDS and treated with ECMO, and the mortality rate for each was extracted. The obtained data was pooled using a random effects model, with estimation of odds ratio (OR) and its 95% confidence interval (95% CI) for mortality in patients with or without ECMO support. Statistical analysis was performed using MetaXL, software Version 5.3 (EpiGear International Pty Ltd., Sunrise Beach, Australia). The study was carried out in accordance with the declaration of Helsinki and local legislation.

Overall, 87 articles were initially identified based on our electronic and reference search, which after screening by tile, abstract, and full text, 83 were excluded for the following reasons: not related to COVID-19 (n = 27), review articles (n = 9), did not provide relevant data (n = 37), and editorials (n = 10). Thus, a total of 4 studies [[3], [4], [5], [6]] were finally included in our pooled analysis, comprising 562 COVID-19 patients, 234 (41.6%)of which developed ARDS. All studies were from China. Table 1 presents essential study characteristics.

Table 1.

Characteristics of included studies.

Authors	Age (yrs): a	# of patients: n = (# ARDS patients)	Conventional ARDS Therapy: n=	Conventional ARDS Therapy Survivors: n (%)	ECMO: n=	ECMO -Survivors: n (%)
Ruan Q et al. 2020	Survivors: 67 (15–81) Non-Survivors: 50 (44–81)	150 (62)	55	7 (12.7%)	7	0 (0%)
Wu et al. 2020	51 (43–60)	210 (84)	83	40 (48.2)	1	0 (%)
Yang X et al. 2020	59.7 (13.3)	52 (35)	29	9 (31.0%)	6	1 (16.6%)
Zhou F et al. 2020	56.0 (46.0–67.0)	191 (59)	56	9 (16.1%)	3	0 (0%)

^aData presented as mean (SD) or median (IQR). ARDS – Acute Respiratory Distress Syndrome.

Among the 234 ARDS patients, 17 (7.2%) underwent ECMO. The mortality rate was 94.1% in the ECMO patients and 70.9% in conventional therapy patients. The results of the pooled analysis is presented in Fig. 1 . The pooled odds of mortality in ECMO versus conventional therapy was not significantly different (OR: 2.00, 95%CI: 0.49–8.16). There was no observable heterogeneity (I² = 0%, Cochran's Q, p-value = .99).

Fig. 1.

Forest Plots for Odds of Mortality in COVID-19 Patients Receiving of ECMO Therapy versus Conventional Therapy.

The results of this analysis using currently available literature would suggest that ECMO does not seemingly produce neither harm or benefit in COVID-19 patients progressing to ARDS. A few limitations shall be noted, however. Patients may have variable levels of ARDS severity, with those receiving ECMO being potentially more critically ill in some cases, which may have impacted the mortality rates. Nonetheless, the mortality rate in the conventional therapy group was relatively high (70.9%). Data on baseline characteristics and disease courses were not provided on ECMO patients. Lastly, the sample size of ECMO patients was considerably small (n = 17). However, when compared to the largest report of MERS patients receiving ECMO (n = 17), COVID-19 patients seem to have substantially higher mortality to-date (94.1% vs. 65.0%), raising questions about real utility of ECMO in this outbreak [1]. Further research is urgently needed. We encourage authors of future COVID-19 reports to provide more data specifically on the ECMO patients in order to aid in optimal patient selection in a limited resource setting.

Funding

None.

Declaration of Competing Interest

None.