NEURODEVELOPMENTAL OUTCOMES AT 4–10 YEARS OF AGE IN CHILDREN BORN AT 22–25 WEEKS GESTATIONAL AGE: AN UPDATED META-ANALYSIS
Abstract BACKGROUND Published data on long-term outcomes are an important resource for clinicians to support expectant parents making perinatal care decisions. OBJECTIVES To update our 2013 meta-analysis with data from recently published, high quality cohorts of 22–25 wk GA infants using relevant definitions of neurodevelopmental disability (NDD). DESIGN/METHODS We used a peer-reviewed electronic search and grey search. Two authors independently reviewed cohorts published after May 2012 with the following inclusion criteria: born ≥1995 in a developed nation, assessed for NDD at 4–10 yrs; prospective data collection, <30% loss to follow-up, consistent definitions for moderate-to-severe (M-S) NDD as per those from EPICure; and results reported by GA. We contacted authors for clarification. Random-effects meta-analysis was used to provide pooled proportions at each GA. Weighted regression was used to examine the relationship between GA and NDD within each study. I2 was used to assess heterogeneity. RESULTS From 2481 titles, 20 full-text articles were reviewed and six (added to the nine from 2013) met inclusion criteria. High heterogeneity at higher GA and wide CIs at lower GA persisted. Results showed rates of M-S NDD in school-age survivors were: 42% (95%CI 23,64%; I2 0%) at 22 wks; 41% (95%CI 31,52%; I2 30%) at 23 wks; 32% (95%CI 25,39%; I2 44%) at 24 wks; and 23% (95%CI 18,29%; I2 60%) at 25 wks GA. The new analysis continued to show a significant decrease in likelihood of M-S NDI between each GA wk (8.1% decrease per wk (95%CI -11.8, -4.5%, p<0.001) but not for likelihood of severe NDD (2.7% decrease per wk (95%CI -6.6, 1.3%), I2 45%, p=0.18). CONCLUSION This high quality data can be used by physicians to support parents during the decision-making process for their infant. Heterogeneity and a small sample size at 22 wk GA remains a concern in the interpretation of the data.