key: cord-1030056-g09ayl1c
authors: Riera, Jordi; Alcántara, Sara; Bonilla, Camilo; Fortuna, Phillip; Blandino Ortiz, Aaron; Vaz, Ana; Albacete, Carlos; Millán, Pablo; Ricart, Pilar; Victoria Boado, María; Ruiz de Gopegui, Pablo; Santa Teresa, Patricia; Sandoval, Elena; Pérez-Chomón, Helena; González-Pérez, Alfredo; Duerto, Jorge; Gimeno, Ricardo; Colomina, Joaquín; Gómez, Vanesa; Renedo, Gloria; Naranjo, José; Alfonso García, Manuel; Rodríguez-Ruiz, Emilio; Eduardo Silva, Pedro; Pérez, Daniel; Veganzones, Javier; Voces, Roberto; Martínez, Sergi; Blanco-Schweizer, Pablo; García, Marta; Villanueva-Fernández, Héctor; Paz Fuset, María; María Luna, Stephani; Martínez-Martínez, María; Argudo, Eduard; Chiscano, Luis; Roncon-Albuquerque, Roberto
title: Risk factors for mortality in patients with COVID-19 needing extracorporeal respiratory support
date: 2021-11-25
journal: Eur Respir J
DOI: 10.1183/13993003.02463-2021
sha: 890440de277aa0d8016185df90fc621f8fb11c9d
doc_id: 1030056
cord_uid: g09ayl1c

Series describing the evolution of patients with severe acute respiratory distress syndrome (ARDS) secondary to COVID-19 and supported with extracorporeal membrane oxygenation (ECMO) during the first wave of the pandemic have reported mortalities ranging from 30 to 60% [1, 2]. More recent publications have demonstrated a trend towards a higher mortality in COVID-19 patients receiving support in later periods of the pandemic, even though the overall mortality of the disease seems lower [3, 4]. The reasons for this difference are not clear.

Passeig Vall d'Hebron, 119. 08035, Barcelona, Spain. Tel No: +34658884293. https://orcid.org/0000-0002-1738-4448 jorriera@vhebron.net / @jrdelbrio. Dear Editor, Series describing the evolution of patients with severe acute respiratory distress syndrome (ARDS) secondary to COVID-19 and supported with extracorporeal membrane oxygenation (ECMO) during the first wave of the pandemic have reported mortalities ranging from 30 to 60% [1, 2] . More recent publications have demonstrated a trend towards a higher mortality in COVID-19 patients receiving support in later periods of the pandemic, even though the overall mortality of the disease seems lower [3, 4] . The reasons for this difference are not clear.

The ECMOVIBER study (The use of ECMO during the coVid-19 pandemic in the IBERian peninsula) is a retrospective-prospective observational cohort study which included consecutive adult patients with SARS-CoV-2 infection admitted to the ICU with severe ARDS and rescued with extracorporeal respiratory support from March 1 st to December 1 st 2020 across 24 ECMO centers (22 in Spain and two in Portugal). In general, inclusion and exclusion criteria for ECMO were those applied in the EOLIA trial [5] . Patients were followed for six months after ECMO commencement.

The aim of the study was to identify factors associated with hospital mortality. The study protocol was approved by the local ethics committees at all the participating centers. The need for informed consent was waived in view of the retrospective nature of the analysis, and because only data available in the medical records were collected. Qualitative variables were described as numbers and percentages.

Quantitative variables were described as means and standard deviations or medians and interquartile ranges (IQR). Confidence intervals for all analyses were set at 95%. The Kaplan-Meier method was A total of 338 patients at the 24 centers received ECMO support during the study period. In 319 (94.4%) cases ECMO was started as a supportive measure for ARDS. Overall mean age was 53±10 years, 258 patients (80.9%) were male, and hypertension was the most common comorbidity [121 (37.9%)]. Patients were cannulated after a median of 5 (3) (4) (5) (6) (7) (8) (9) days from the initiation of mechanical ventilation (MV), 7 (4-13) days after ICU admission and 17 (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22) days after symptom presentation.

Coinfection at ECMO initiation was recorded in 95 (29.8%) cases. Ninety-six (30.1%) patients were supported at high-volume centers (>30 ECMO cases/year) and 129 (40.4%) needed ECMO retrieval and transport.

One hundred and eighty (56.4%) patients were successfully decannulated. The median duration of ECMO support was 17 (9-32) days, with 84 (26.3%) runs lasting more than 30 days. One hundred and fifty-six patients (48.9%) were discharged alive, 156 (48.9%) died in the hospital and seven patients were still hospitalized at six-month follow-up (one of whom remained on ECMO).

Associations of variables with hospital mortality are detailed in figure 1 (upper panel). Age, ischemic cardiomyopathy, center case volume, ECMO retrieval, days from symptoms to cannulation, driving pressure prior to ECMO and drainage cannula size were associated with hospital mortality. In contrast, pre-ECMO MV days were not associated with survival (figure 1a). In the multivariate analysis including the wave in which ECMO support was received, the hazard ratio for hospital mortality was four times higher in patients over 65, and the survival rate of patients supported at centers with a volume of 30 cases or above was significantly higher (figures 1b and 1c). Longer time from symptom onset and higher driving pressure before cannulation were associated with a higher risk of hospital mortality, while higher levels of PEEP at day three of ECMO support were associated with a lower risk of hospital mortality. Larger drainage cannula diameter was also associated with a reduced risk of death on ECMO [HR 0.88 (0.80-0.96), p=0.005] but not when it was included in the multivariate model for hospital mortality.

In this large multicenter series of COVID-19 patients receiving ECMO support, we detected several factors associated with hospital mortality. The study of these factors might help in the identification of COVID-19 cases who might benefit the most from ECMO and provide valuable information for improving the management of these patients.

Regarding the criteria for indicating the technique, a period of MV longer than 7-10 days prior to ECMO commencement has traditionally been considered as a contraindication. Recently, Díaz et al indicated that length of MV had no impact on survival [6] ; however, the study included only 11 patients with prolonged duration of MV prior to ECMO. Our analysis, which includes 72 patients with more than 10 days of pre-ECMO MV, confirmed the absence of this association; in fact, we found an association between longer time between symptom onset and ECMO initiation with a higher risk for hospital mortality. This time frame includes the period of non-invasive oxygen therapy prior to intubation, in which the lung could be further damaged leading to a potential decrease in lung resilience. Over the course of the pandemic the time of invasive MV initiation changed, with a general tendency for an early start during the first wave and delayed intubation in later periods. In these later phases there was also a dispersion of ECMO cases (4) . In this regard, and in agreement with previous results [7] , we found that patients supported at centers with a high case volume had a lower mortality.

The recommendation to concentrate cases at high-volume centers should include the capability to retrieve the sickest patients admitted in hospitals that do not have access to this technique [8, 9] . Not surprisingly, in our series, patients who were retrieved had lower hospital mortality, a finding that reinforces the idea that attention should be centralized. The age cut-off point for ruling out ECMO support has also been a matter of debate [10] . In our series, in agreement with previous reports [11] , survival rates were significantly lower in patients aged over 65, and so it seems reasonable to establish age above 65 as a relative contraindication for COVID-19 ECMO support.

Adequate management of the extracorporeal support, together with the pathological condition itself, also have a direct impact on outcome. Lung parenchyma of COVID-19 patients needing ECMO is severely affected, with notably high lung elastance and very low residual function [12, 13] . High ECMO flows are needed in this situation in order to ensure adequate oxygen delivery and allow ultraprotective MV. For this reason, the diameter of the drainage cannula has to be large enough to achieve these high ECMO flows. Our analysis found information suggesting that a larger diameter of the drainage cannula was a protective factor. However, this association should be confirmed in further investigations. MV management of patients on ECMO has also been considered [14] . Our multivariate analysis indicated a relation between higher PEEP levels on the third day of ECMO support and a lower risk of hospital mortality. This result is in agreement with those of previous reports including patients with ARDS [15] , but no information has been published to date about PEEP titration in COVID-19 ECMO patients.

In conclusion, when indicating ECMO support for ARDS in COVID-19 patients, center case volume, age, driving pressure and the duration of symptoms should be taken into account. Length of MV prior to ECMO per se should not be included in this decision. In the management of these patients, PEEP levels should be kept high during the first days. 

Elisabet Gallart

Iago Sousa (Department of Cardiology)

Hospital Universitari Son Espases: Marta Ocón, María Riera (Department of Intensive Care)

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

Extracorporeal membrane oxygenation support in COVID-19: an international cohort study of the Extracorporeal Life Support Organization registry

Extracorporeal membrane oxygenation for COVID-19 during first and second waves

Increased mortality in patients with COVID-19 receiving extracorporeal respiratory support during the second wave of the pandemic

Extracorporeal Membrane Oxygenation for Severe Acute Respiratory Distress Syndrome

ECMO for COVID-19-Associated Severe ARDS in Chile: A Nationwide Incidence and Cohort Study

For the Paris ECMO-COVID-19 investigators. Extracorporeal membrane oxygenation network organization and clinical outcomes during the COVID-19 pandemic in Greater Paris, France: a multicenter cohort study

Extracorporeal membrane oxygenation retrieval in coronavirus disease 2019: a case-series of 19 patients supported at a high-volume extracorporeal membrane oxygenation center

Planning and provision of ECMO services for severe ARDS during the COVID-19 pandemic and other outbreaks of emerging infectious diseases

Extracorporeal membrane oxygenation: evolving epidemiology and mortality

The PRESERVE mortality risk score and analysis of long-term outcomes after extracorporeal membrane oxygenation for severe acute respiratory distress syndrome

Management of COVID-19 Respiratory Distress

Histopathological features in fatal COVID-19 acute respiratory distress syndrome

Associations between ventilator settings during extracorporeal membrane oxygenation for refractory hypoxemia and outcome in patients with acute respiratory distress syndrome: a pooled individual patient data analysis: mechanical ventilation during ECMO

Mechanical ventilation management during extracorporeal membrane oxygenation for acute respiratory distress syndrome: a retrospective international multicenter study

The authors declare that they have no conflict of interest.

Hospital Research Institute (VHIR).The following practitioners provided care for the study patients: