A highly pathogenic human coronavirus causing respiratory disease emerged in Saudi Arabia in 2012. This viral infection termed Middle East respiratory syndrome coronavirus (MERS-CoV) is associated with high mortality rate in approximately 36% of reported patients.

The World Health Organization (WHO) reported 1,374 laboratory-confirmed worldwide infections, including at least 490 related deaths, from September, 2012, to July 24, 2015.2 The higher incidence of MERS-CoV infections in Saudi Arabia may be related to multiple factors, including seasonality, increased proactive screening, poor infection control measures, low relative humidity, and high temperature.

Infected patients with MERS-CoV usually have abnormal findings on chest radiography, ranging from subtle to extensive unilateral and bilateral abnormalities. MERS progresses rapidly to respiratory failure, in approximately 2/3 of infected patients, which has a high mortality rate, particularly in immunocompromised patients.

Extracorporeal membrane oxygenation (ECMO) has emerged as a rescue therapy in patients with refractory hypoxemia during the H1N1 epidemic.The use of veno-venous (VV)-ECMO provides respiratory support for patients with respiratory failure, whereas the use of veno-arterial (VA)-ECMO could be helpful in those with cardiorespiratory failure.10 However, the survival rate of the infected patients with H1N1 who required the use of ECMO varies considerably among the Caucasian and Asian countries (90% survival in Sweden and 83% in the UK13 vs. 35% in Japan). This large discrepancy could be explained with lack of satisfactory equipment, therapeutic guidelines, training of staff, and effective systems allowing patient transfer to the dedicated ECMO centres.

Guery and co-investigators described the use of ECMO in two French patients with cardiorespiratory failure secondary to MERS-CoV infection.This has been extended for treatment of refractory hypoxemic respiratory failure during the Saudi MERS-CoV outbreak.

The Saudi Ministry of Health has implemented a national ECMO program since 2014 in three major cities including Jeddah, Al Madinah and Riyadh which have most of the reported infected cases with MERS-CoV. The Saudi ECMO program provides a rapid transportation chain system (Medevac system), adequate number of intensive care beds and ECMO machines, and highly trained perfusionists and staff.

The investigators hypothesized that the early use of ECMO for treatment of severe acute hypoxemic respiratory failure, defined as a ratio of the PaO2 to the fraction of inspired oxygen (PaO2/FiO2 ratio) less than 80 despite optimized ventilator management, in infected patients with MERS-CoV, would be associated with reduced in-hospital mortality rate.

Patient Selection:

The investigators obtained a centralized ethics approval from the Ministry of Health to avoid delays and to facilitate the conduct of this timely important study. Eligible patients or their legal guardians were contacted to request their participation and obtaining of their written consent.

Patients who are 18 years or older who received ECMO support for MERS-CoV associated hypoxemic respiratory failure were included. MERS-CoV infection is defined using the WHO case definition. A positive polymerase chain reaction (PCR) on nasopharyngeal or oropharyngeal swabs, sputum, tracheal aspirate, or bronchial alveolar lavage is sufficient to establish the diagnosis of infection.

Description of Standardized National Protocol

The participating Saudi centers had adopted a standardized protocol based on the evidence-based guidelines for the treatment of acute respiratory distress syndrome (ARDS) associated with the H1N1 virus infection using low-tidal volume, lung-protective mechanical ventilation as the initial strategy. A lung-protective strategy was applied using volume assist-control mode, pressure-controlled synchronized intermittent mandatory ventilation mode or pressure-controlled ventilation mode, with a tidal volume of 6 to 8 mL/kg of predicted body weight and variable FiO2 and the positive end-expiratory pressure (PEEP) to achieve arterial oxygen saturation (SaO2) from 88% to 95% or a partial pressure of oxygen (PaO2) of 55 to 80 mm Hg.Then a full spectrum of ventilator modes, including airway pressure release ventilation, prone ventilation and high frequency ventilation was used.

If, despite and after the above measures, a patient cannot achieve a ratio of the PaO2 to the fraction of inspired oxygen (FiO2) (PaO2/FiO2 ratio) greater than 100 on ''safe'' settings (i.e. FiO2 less than 80%, peak inspiratory (Ppk) and plateau (Ppl) pressures less than 40 and 35 cm H2O, respectively and tidal volume less than 6 to 8 ml/kg), the patient was assessed for eligibility for ECMO support. Veno-venous (VV-ECMO) was used for respiratory support for those with respiratory failure, whereas the veno-arterial (VA-ECMO) was used for those with cardiorespiratory failure.

Once adequate ECMO support was instituted, the ventilator was set to low ''recruitment'' settings. When a patient began to show evidence of pulmonary recovery, the ECMO support was weaned off with gradual reducing blood flow, gas flow, and FIO2 over the membrane, when the PaO2/FiO2 ratio was greater than 200 with an FiO2 of 50% and pressures less than 38 cm H2O.

Selection of Historical Cohort

The control group was identified retrospectively, patient who did not receive ECMO due to lack of access but who fulfill the criteria for initiating treatment will be selected.

Data Collection:

National database was used to identify patients who met our eligibility criteria. Trained research investigators collected the relevant datafor eligible patients. The investigators used pre-designed case report forms (CRF) to abstract data. The investigators collected data on: baseline characteristics including age, sex, height, weight, and ethnicity, as well as the presence of a number of predefined comorbidities, ICU pharmacologic interventions, ventilation data including days of mechanical ventilation, ventilation mode, and mean values of tidal volumes, positive end expiratory pressure (PEEP) levels, FiO2, and PaO2/FiO2 ratio before, during and after the initiation of ECMO support, and the administration of antiviral and antibiotic medications, the type, gas flow (liter/min), blood flow (liter/min/m2) and duration of ECMO, circulatory support, length of ICU and hospital stays, mortality during hospital stay.

In addition, the need for renal replacement therapy; tracheostomy; bacterial co-infection duringICU stay; ventilator-associated pneumonia was recorded.

The investigators will document whether the ECMO treatment was initiatedatthe participating centeror whether the patient was transferred to an ECMO center.

Data on eligible patients was recorded retrospectively during ICU stay. Data on hospital discharge or death will be recorded as well.

Statistical Analysis:

Descriptive data were reported as numbers and percentages for dichotomous variables; and median [interquartile ranges] or and mean (SD) for continuous variables. All outcome data from the Cohort-Controlled groupand ECMO group were compared using independent Student t test, Mann Whitney U test or X2 test as appropriate. P values < 0.05 were considered statistically significant.

To determine the predictors of the need for ECMO and factors associated with death in ECMO-treated patients were entered in a multivariate stepwise backward logistic regression model.

Inclusion Criteria:

• Positive infection with Middle East Respiratory Syndrome virus
• Refractory hypoxemic respiratory failure
• Eligible for use of extracorporeal membrane oxygenation support (ECMO)

Exclusion Criteria:

• Neonates
• Children
• Patients treated with ECMO for primary cardiac failure
• Following heart transplantation
• Following lung transplantation
• Following cardiac surgery
• Patients with an alternative diagnosis who had no virus isolated