The first Epicure study had relatively poor survival rates, especially for the more immature babies. Of course the data were regional data, including all babies born in the UK and Ireland, so they included babies born in places with very low levels of care, many of whom were never transferred to intensive care units. Although I think EpiCure1 was an amazing effort, the results were often used as a way of suggesting that neonatology was very ineffective. Every position statement of learned societies referred to the data, often with no description of the context . In the UK though the data were used to try and improve their regionalization, with what are called managed clinical networks. There were at the time of Epicure1 many very small units treating tiny numbers of extremely preterm babies per year, the data were never presented by level of care of the birth hospital.

The new Epicure study has rectified that (Marlow N, Bennett C, Draper ES, Hennessy EM, Morgan AS, Costeloe KL. Perinatal outcomes for extremely preterm babies in relation to place of birth in England: the EPICure 2 study. Archives of Disease in Childhood – Fetal and Neonatal Edition. 2014.. Open Access)

The data show quite convincingly that being born in a hospital with level 2 services, and then transferred later leads to much poorer survival.

They show that being born in a level 3 hospital leads to better survival, and without more morbidities, so survival free of morbidity is much better (morbidity being defined as one or more of retinopathy of prematurity requiring retinal surgery, moderate or severe bronchopulmonary dysplasia, a severe brain injury (haemorrhagic parenchymal infarct, cystic changes or hydrocephalus on cerebral ultrasound) or surgical necrotising enterocolitis. They also showed that among level 3 hospitals, those which have higher patient volume have substantially better survival.

They also show, depressingly, that

Despite national policy, only 56% of births between 22 and 26 weeks of gestation occurred in maternity services with a level 3 neonatal facility

The proportion can never be 100% of course, but adequate regionalization can make that figure much more than 56%. The exact proportion will depend upon many factors of health care organization, and geography, but aiming for 90% is reasonable, and would save many lives.