The data presented by Rattihalli et al1 on the evolution of health outcomes for babies born before 26 weeks gestation in the Trent region show rising survival rates (27–46%) between 1991–1993 and 2001–2003 accompanied by a slightly increased probability of severe disability among survivors (35–39%). These results raise multiple questions: Why were there such large gains in survival but no improvement in impairment rates? Can these results be used to draw conclusions about trends in outcome in other regions of Europe or in the USA? How can these data be used to guide clinical decisions for these infants whose probability of being severely disabled is over 1 in 3?

Trying to contextualise these findings using the available scientific literature is daunting. There is convergence among studies showing increases in survival for very preterm infants throughout the 1980s and early 1990s. Some recent data, however, suggest that survival increases slowed or stopped starting in the late 1990s.2,–,4 Survival among babies of less than 26 weeks gestation from the state of Victoria, Australia, increased between 1991/1992 and 1997 from 33% to 54%, but decreased to 43% in 2005.2 The National Institute of Child Health and Human Development (NICHD) Neonatal Research Network reported stable mortality between 2003 and 2007 among babies of less than 28 weeks gestation.4 In contrast, survival improved for infants of less than 28 weeks gestation in the Parisian region between 1997 and 2003, in the southern region of Sweden between 1995 and 2004 and in Japan between 1990 and 2005.5,–,7

Assessing the trends in longer term health outcomes after preterm birth is even more challenging because studies use different definitions of disability and evaluate children at varying ages. In this context, this study from the Trent …