Long-term neurodevelopmental outcomes in extremely preterm infants born at 22–26 weeks gestation: a follow-up of 2–2.5 years across two Swedish national cohorts from 2004–2007 to 2014–2016
Ulrika Ådén, A Farooqi, Lena Hellström‐Westas, Karin Sävman, Thomas Abrahamsson, Lars Björklund, Magnus Domellöf, Anders Elfvin, Fredrik Ingemansson, Fredrik Serenius, Stellan Håkansson, David Ley, Erik Normann, Petra Um‐Bergström, Karin Källén, Mikael Norman
Abstract
OBJECTIVE: To compare neurodevelopmental outcomes in extremely preterm (EPT) children born across two epochs in Sweden. DESIGN AND SETTING: Nationwide population-based cohorts of infants born at 22-26 weeks' gestation in 2004-2007 (Cohort 1) and 2014-2016 (Cohort 2), comprising 1606 live births. Survivors were assessed at 2-2.5 years' corrected age using the same protocol design. MAIN OUTCOME: The primary outcome was neurodevelopmental impairment (NDI), defined as a composite of moderate-severe cerebral palsy (CP), visual or hearing deficits, or moderate-severe cognitive, language or motor impairment assessed with the Bayley Scales of Infant and Toddler Development, Third Edition (Bayley III). For children not assessed with Bayley-III, NDI was defined as moderate-severe speech delay, general developmental delay or categories of CP, vision and hearing impairment. Outcomes were compared using logistic regression to evaluate differences between cohorts and perinatal and socioeconomic risk factors. RESULTS: Of 1188 eligible survivors, 1062 (89.3%) were assessed (mean gestational age (GA) 24.8 weeks; 54.9% male). The prevalence of moderate-severe NDI at 22, 23, 24, 25 and 26 weeks' gestation was 60% vs 52%, 51% vs 51%, 34% vs 42%, 27% vs 32% and 17% vs 24% in Cohorts 1 and 2, respectively. Overall prevalence did not differ significantly (27% vs 35%; adjusted OR (AOR) 1.2, 95% CI 0.94 to 1.6). Among 724 (68%) children assessed with Bayley III, Cohort 2 had higher rates of cognitive delay (21.6% vs 11.3%; AOR 1.8, 95% CI 1.1 to 3.4) and language delay (40.9% vs 16.1%; AOR 3.3, 95% CI 1.4 to 4.1). Low GA and maternal country of birth outside the Nordic region were the strongest predictors of NDI and cognitive delay, the latter association confined to Cohort 2. CONCLUSION: Although survival of EPT infants in Sweden has improved, long-term neurodevelopmental outcomes have not. The root causes of failed improvements in long-term outcomes for EPT infants are complex and need further clarification.