|Home | About | Journals | Submit | Contact Us | Français|
The only large randomised controlled trial of extracorporeal life support (ECLS) was conducted in the UK in 1996 and recruited 185 infants with severe respiratory failure (oxygenation index >40); those transferred to an ECLS facility exhibited a markedly reduced risk of death or disability at one year compared with those who remained at their local centre.1 Factors that influence the decision to start ECLS include the potential reversibility of the disease process, anticipated clinical course and a perceived high risk of death.2
Although historically the oxygenation index has been used to guide referral, there are no standardised criteria for initiating transfer for ECLS consideration. We aimed to determine whether any clinical, ventilatory or laboratory surrogates of outcome can assist decision making with regard to suitability for ECLS. A retrospective case note review of the West of Scotland Neonatal Transport Database was carried out for the period from March 2003 to March 2006. At the point of referral extensive epidemiological data were collected: sex, gestational age, birthweight, age, diagnosis, ventilation mode and settings, arterial blood gas parameters, oxygenation index, mean blood pressure, use of inhaled nitric oxide, and number of inotropes used. On arrival at the ECLS centre, based at Glasgow's Yorkhill Hospital, paired data were collected. We analysed the data with SPSS Version 12.0, and bivariate correlation was used to identify factors that significantly influenced survival. Significance was assumed at α=0.05.
A total of 52 infants were referred for consideration of ECLS. Three infants were assessed by the transport team to be unsuitable for transfer and were excluded from further analysis. For the 49 infants transferred (20 girls, 29 boys; see table 11 for characteristics of the infants at the time of referral), the referral diagnoses included persistent pulmonary hypertension of the newborn (n=17), meconium aspiration syndrome (n=16), congenital diaphragmatic hernia (n=7), septicaemia (n=5), respiratory distress syndrome (n=2) and congenital heart disease (n=2). Thirty infants (61%) subsequently received ECLS. Overall, of the 49 infants transferred, 42 (86%) survived to discharge (90% of those who received ECLS, 79% of those who did not receive ECLS). Overall there were no significant differences between the characteristics of the infants in the different diagnostic groups (data not shown). All infants with a diagnosis of meconium aspiration syndrome survived, 69% of whom needed ECLS, highlighting the excellent prognosis for this subset of infants, provided a prompt referral is made. No clinical or laboratory factors at referral significantly predicted survival. In particular, indices of oxygenation at the time of referral did not predict survival (oxygenation index (OI): r=0.04, p=0.76; fractional inspired concentration of oxygen (Fio2): r=0.10, p=0.94; arterial partial pressure of oxygen (Pao2): r=−0.11, p=0.46). The paired counterparts of these indices at arrival at the ECLS centre again did not predict survival (OI: r=−0.004, p=0.98; Fio2: r=0.21, p=0.14; Pao2: r=−0.42, p=0.78).
In summary, we advocate early communication with the ECLS service to minimise mortality. Numerical predictors of illness severity, such as the oxygenation index, are poor determinants of survival following ECLS. The condition of the infant, based on clinical experience, must remain the best indicator of need for transfer.
Competing interests: None.